Skip to content

What’s the deal with the YIMBYs?

This post is not by Andrew. It is by Phil.

There’s at least one thing people in San Francisco seem to agree on: the rent is too damn high. The median rent is between about $3000 and $3500 per month…for a one-bedroom apartment. High-tech workers and upper-echelon businesspeople can afford a place, but baristas and hair salon workers and teachers and shop clerks etc. etc. have real trouble.

Of course there is plenty of development pressure, and new high-rise apartments are going in that have hundreds of apartments each, typically with a rent of $4000 – $8000 per month. If you let a developer build “market rate” apartments, that’s what they’ll build.

Suppose San Francisco adds 10,000 market-rate units. Some will be one-bedrooms, some two- or three-bedrooms, and some will be occupied by singles, others by couples, etc. But for simplicity let’s just assume that on average each of these new 10,000 units is occupied by a household that earns $100K per year after taxes. Total, the occupants of these apartments have $1,000,000,000 in disposable income. That’s a lot! They’re going to spend some of that money — indeed, a lot of that money — in San Francisco: they will go to coffee shops, get their hair cut, buy things in shops, etc. Serving those extra 10,000 high-income households will require tens of thousands more waiters and shop clerks and car mechanics and plumbers etc etc etc….that is, there will be more jobs for the kinds of people who already have trouble affording a place in San Francisco. Those additional people will need to live somewhere, so there will be increased competition at the lower end of the market, which means higher rents. Most of these people will end up commuting from other cities.

The thing is, it’s not just lower-income people who feel priced out of San Francisco. Tens of thousands of high-income people who would like to live in San Francisco are living in Oakland and Fremont and Berkeley and Orinda because of lower rents in those places. As market rate housing is built in San Francisco, those people move into it. That’s why the ‘market rate’ is so high.

Of course, by the so-called “law of supply and demand”, building more housing does make housing cheaper. It’s easy to see why: those people with their billion dollars of disposable income are adding a lot economic activity in San Francisco, but they’re decreasing the economic activity in the cities they’re leaving, which no longer need so many waiters and barbers and shopkeepers. There is a cascade: some people move from Berkeley and Oakland to San Francisco, which allows replacements to move from Richmond and El Cerrito into Berkeley and Oakland, and so on. Ultimately, rents in San Francisco go up, and rents in some outlying communities go down. Yes, the increased supply of housing lead to decreased housing prices on average but they’ve gone up, not down, in San Francisco itself.

Admittedly it is possible in principle to build so many apartments in San Francisco that they become more affordable. One way for that to happen is if the city becomes a less pleasant place for rich people to live: crowds, traffic, lack of sunlight as big buildings fill the sky, and improved amentities in neighboring areas, could get to the point that people would rather live in a $3000 apartment in Oakland than a $3500 apartment in San Francisco. But it might take a great deal of building, and a complete change in the character of the city, for that to happen. Manhattan is 400% denser than San Francisco but it’s still not a cheap place to live [Note: I originally said ‘50% denser’ but a commenter named Ira pointed out I was looking at the density of New York City, not Manhattan]. Some people say Manhattan would be cheaper if it were easy to build more housing there, but for crying out loud, they already have 1.6 million people living on the island. What’s the theory, that housing for the first hundred thousand people didn’t make rents go down, nor did housing for the second or third or fourth or… or fourteenth or fifteenth or sixteenth tranche of a hundred thousand people, but we have finally reached the peak and the NEXT hundred thousand housing units will make housing cheaper?  Sorry, no. If the ‘market rate’ for newly developed apartments is substantially higher than the median rent of existing apartments, then building more market-rate apartments will make median rents go up, not down.

(It’s worth remembering that there are other things that could make rents go down too, such as a local or national economic collapse. Maybe the next major earthquake will put a damper on things.)

Given all of the above, until recently I found it really perplexing that San Francisco has a movement of people who loudly claim to be pro-renter and to want policies changed so that rents will go down, but who are strongly in favor of building more market-rate housing. (The broader movement is sometimes characterized as YIMBY = Yes In My Back Yard, which is also the name of an organization; there is also the Bay Area Renters Federation, BARF, which is even more explicit that they don’t care what housing gets built, as long as it gets built). Why, I wondered, are these people promoting policies that are so bad for them?

But suddenly it dawned on me, just last week, that the question “why are people in favor of policies that are so bad for them” might have the same answer in this case that it appears to have for a lot of people in national politics: they aren’t trying to do something good for themselves, they are trying to hurt their perceived enemies. (In fact, there’s a New York Times Op-Ed on the day that I’m writing this, May 14, that points out that some people on the Right are currently cheerleaders for behavior that they once would have found objectionable, not because they think it’s good for them or for the country as a whole, but because it appalls the Left).

When someone proposes to build a 400-unit market-rate housing tower in San Francisco, who objects?  Well, in a minor way, the affordable housing community will usually make some noise, but they usually quiet down if the developer guarantees that 10% or so of the units will be ‘affordable’. I presume the negative impact of higher rental costs everywhere in the city is just too diffuse to energize people from distant neighborhoods.  As for the major objections, they come from people immediately surrounding the proposed project. These people complain (usually correctly) that the proposed project will cause more traffic and more noise, that two years of construction will cause major inconveniences, etc., etc., and they show up in numbers at public hearings to try to kill the project or at least get it reduced in size.

And who are these people? They’re people who already have what the YIMBY and BARF members want: a place to live that they like and can afford. These are often people who bought their condos or rented their rent-controlled apartments years ago. They’re older and often wealthier and probably smugger than the college-debt-laden millenials who desperately want rents to come down.

So this is my new theory: the YIMBY and BARF people know that building more market-rate housing in San Francisco will make median rents go up, and that this will be bad for them, but they want to do it anyway because it’s a thumb in the eye of the “already-haves”, those smug people who already have a place they like and are trying to slam the door behind them.

I hope this is the answer, because if it isn’t, I’m still at a loss.

Comments welcome.

By the way, I, Phil, do not live in San Francisco, I live in a quiet neighborhood of single-family homes across the Bay in Berkeley.  I don’t have a dog in this fight. (Remember, this post is not by Andrew, it’s by Phil.)



  1. William A Schlieper says:

    For a new building project to increase median rents, it has to not only be more expensive than the current median rent but also fail to drive deprecation in existing apartments, a combination that only seems possible to me if every single unit is purchased by someone not already living in the city. I can see that happening due to the tech industry but since the demand to live in SF isn’t driven by the apartments, the alternative if building isn’t done is evictions.

    • Phil says:

      William, you say “For a new building project to increase median rents, it has to not only be more expensive than the current median rent but also fail to drive deprecation in existing apartments, a combination that only seems possible to me if every single unit is purchased by someone not already living in the city.”

      I don’t think that’s true at all. Indeed, I think every one of the new apartments could be purchased or rented by someone already living in the city, and rents would still go up. If you build 10,000 new units for rich people and 15,000 people move into them, the population of San Francisco will go up by close to 15,000 rich people. In general will be a combination of people moving into the new units from out of town, and people moving from other places in the city into the new units and thereby freeing up space for people from out of town. This is what I meant by a ‘cascade’ of effects. It is conceivable that every one of the apartments could be occupied by someone who already lives in SF, but the apartments they are moving OUT of are not going to sit vacant, they will be filled by someone.

      I think if there are currently 852,500 people in San Francisco and you build space for 15,000 more people, to first order you will get 15,000 more people in the city, or something very close to that. And if you build expensive space for 15,000 more people, you will get something close to 15,000 more rich people. And then there’s the second-order effect, which is that those 15,000 rich people create a demand for more workers.

      • Ryan Moulton says:

        If you don’t build space for those 15000 rich people, then they take the apartments of 15000 newly priced out people, and live there instead. You can’t keep out rich people by restricting how much you build. They are the heaviest grains of sand, they will always get to the bottom.

        • Michael Watts says:

          Just to point out, in a mixture of fine-grained and coarse-grained sand, the coarse-grained (heavier) sand will rise to the top, because its density is lower. Similarly, the crumbs in a cereal box sink to the bottom, below the individually-heavier whole cereal bits.

      • Michael says:

        There is only a finite supply of high paying jobs in the Bay Area. If more apartments appear on the market, this then the pool of people who want to live in sf are competing over a larger pool of apartments. Under most circumstances, this will force other landlords to charge lower rent than they would if the supply of apartments were smaller.

        There are some ways this might not work out, for example if the new apartments increase the desirability of the city, then the pool of people who want to live in the city might increase more than the pool of housing, but this would be a strange situation since it would mean the apartments are so nice, people just want to live near them, not necessarily in them.

        More housing could also boost the economy in a way that increases the number of high paying jobs, which doesn’t sound so bad.

        • Martha (Smith) says:

          “There is only a finite supply of high paying jobs in the Bay Area”

          Finite is not the same as constant. If the number of high paying jobs is increasing, then the number of people competing for the increasing pool of apartments is increasing. So it’s a matter of which of supply (apartments) or demand (high paid workers) is increasing faster.

      • A says:

        And those workers raise rents? I think you are relying on more questionable assumptions than you acknowledge. Perhaps workers have such a strong effect in your model because wealthy people are extremely reluctant to move into less expensive areas. And you might be assuming that the wealthy influx significantly increases worker wages so that the worker influx outprices existing workers because you also assume a skills mismatch?

      • Steve Sailer says:

        Phil says: “If the ‘market rate’ for newly developed apartments is substantially higher than the median rent of existing apartments, then building more market-rate apartments will make median rents go up, not down.”

        I think Phil is right. But I also think the economists are right too that the law of supply and demand applies to San Francisco housing. They are just talking about somewhat different things.

        What’s likely to happen if development in San Francisco is opened up is that a lot of expensive luxury apartments will be built. This may well make the median rent go up. On the other hand, the rent per quality-adjusted square foot will go down. People in San Francisco will get a little better deal in terms of space and amenities for their extremely high rents.

        • jim says:

          ” rent per quality-adjusted square foot ”

          Perfect! More people are buying Lexus than Toyota or scion, so the price of Lexus is rising, but it doesn’t mean the price of Toyota or Scion is rising, or is rising as much.

          But the comparison isn’t strictly correct because unlike cars, buildings have fixed locations and part of the value is that location. So a “Scion” building can be situated at a “lexus” location – in other words, the cost of lexus apartments is reflected in the value of the land of scion apartment buildings; I’m struggling for the economic term – the future value of the potential Lexus building that could replace the Scion building is incorporated into the current value of the scion building.

          • Steve Sailer says:

            I’m guessing that if a new building with luxury apartments renting for $10,000 per month goes up in a previously less expensive median rent neighborhood, the rents on the dumpier old buildings in the neighborhood might go up on average. They might also go down, due to more supply, but the larger impact might be to raise the fashionableness of the neighborhood.

            I’m surprised there aren’t much in the way of well-knowns studies of these kind of real estate questions are so interesting to people in their daily lives.

  2. Martha (Smith) says:

    The situation is not quite as bad in Austin (where I live), but there are similarities. Here, at least, I’m not so sure that your “answer” fits, which is not to say that I have “the answer”. But my impression is that here, a lot of the impetus for new developments comes from — the developers!. They’ve got a lot of skin in the game, and my impression is that they co-opt arguments (such as that supply-and-demand will lead to lower prices, or that changing building codes will open housing opportunities to minorities) that give them the new developments that put more money in their pockets — and that (at least some) others buy into the arguments. But there also seem to be some people who have religious-like convictions about supply and demand.

    Meanwhile, we are in the situation where low-cost housing is being torn down to build higher-cost housing — increasing “supply” overall (e.g., if two houses are built on a lot that formerly held one) but reducing supply of low-cost housing. And minorities are being pushed out of their old (now “gentrified”) neighborhoods by high prices and high taxes. And the new developments also lead to increased flooding and traffic congestion.

    The high rises are a slightly different matter — but only slightly: They benefit mainly the developers, have “justifications” that seem iffy (but are bought into by many people), increase run-off/flooding problems, and increase traffic congestion) and thereby air pollution.

    • Phil says:

      I was careful not so suggest an answer! All I’m claiming is that if San Francisco builds more market rate housing, the median rents will go up.

      I certainly agree with you that the developers are in favor of building market-rate housing in hot markets, and that this is where a lot of the impetus for new development comes from. That part is easy to understand! What has always perplexed me is that some advocates of low-cost housing also want to build market-rate housing, which is clearly counterproductive. Now I have a theory that explains that!

      • Ryan Moulton says:

        What? Rents will lower somewhere, just not precisely in the place you build it. The goal and the primary effect is still lowering rents.

        • jim says:

          Thank you. That is the correct answer.

        • Corey says:

          Ryan, Phil wrote in the OP: “Ultimately, rents in San Francisco go up, and rents in some outlying communities go down. Yes, the increased supply of housing lead to decreased housing prices on average but they’ve gone up, not down, in San Francisco itself.”

          So Phil claims that people who claim to want low cost housing in San Francisco are acting in a way that thwarts their stated goal, and is trying to figure out why.

          • Phil says:

            Yes, exactly, thank you Corey.

            • Ryan Moulton says:


              The median is not the important statistic. When you build unit A, the statistic to measure is the rent for existing unit B, which must almost certainly go down.

              • Steve Sailer says:

                Phil is right that building a lot of $10,000 per month luxury apartments will likely drive up the median rent in San Francisco. You may well be right that existing apartments might decline slightly in price because the new supply helps better meet demand.

                On the other hand, Phil may be right that improving the housing stock for rich people might draw more people into the city who will in turn attract even more service workers to care for them.

                On the other other hand, a lot of new luxury construction in Manhattan overlooking Central Park sits idle much of the year because it’s often bought by shady foreign billionaires looking for somewhere to launder their dubious cash. I don’t know whether San Francisco is as attractive to global billionaires as NYC, but a lot of construction could jumpstart a market in investment apartments.

          • Ryan Moulton says:

            He misunderstands their goal then. The goal is to lower rents everywhere, not merely in San Francisco. San Francisco is a tiny part of the regional housing market.

    • Austin says:

      Austin does have its share of YIMBY’s. They are AURA Austin, FAN (Friends of Austin Neighborhoods), and the “Friends” subgroups that have formed all over town to undermine neighborhood association authority. They are entitled millennial’s who will even say that the “legacy” residents need to leave. They claim they are working for more affordable housing in Austin, but like anywhere else, the new housing is always more expensive than what gets torn down. In fact, the YIMBYs in Austin are the puppets of the development and real estate industry who fund their astroturf organizations.

    • John Hall says:

      Its incredibly apt. Japan has a much better zoning system than many U.S. cities do.

    • Phil says:

      A Bloomberg article from a year ago says “The average price for apartments in Tokyo and commuter regions has climbed 22 percent during the past three years, according to the Real Estate Economic Institute. The price for an apartment in Tokyo’s main urban area, where more than 9.2 million people live, rose 12 percent on average to 67.3 million yen in 2015. The increases are coming even as average Japanese base wages in November fell 1.4 percent from three years earlier to about 239,250 yen, data from the Labor Ministry show.

      With building costs and land prices rising, it has become increasingly difficult to pass on those charges in suburban developments, so large developers have been building more luxury apartments targeting the wealthy, according to a report last week by analysts Tomoyoshi Omuro and Junichi Sano at Morgan Stanley MUFG Securities Co.”

      So, yes, in a way this is an apt comparison. But in a way it isn’t, because Japan is still in an economic slump — note the falling base wage in Japan — and the Bay Area is not.

      • A says:

        You are now conflating the effects of Abenomics, conveniently ignoring two decades of flat pricing despite adding 1.6 million people since 1990.

        And Tokyo wages have been flat for decades, while U.S. real estate shows significant increases in price to income, and increasing gaps between valuation metrics between locals. Why did Tokyo buck the trend?

        • Phil says:

          Hey, I’m not the one who brought up Tokyo and said it’s an apt comparison! I’m the one pointing out Tokyo is not an apt comparison!

          According to Japan Property Central, “The ratio [of housing costs to salary] was the highest in the Tokyo metropolitan area, with a price-to-income ratio of 9.79 times for new apartments and 7.20 times for second-hand apartments. Kyoto Prefecture had the second highest ratio for new apartments with the average apartment costing 9.78 times the average income for the prefecture.” A quick look at their website didn’t turn up a time series plot of price/salary for second-hand apartments, but they do have it for new apartments, and the ratio has increased from a low of about 7 (in 2000) to its current level of 9.8. That’s a 40% increase.

        • Steve Sailer says:

          Okay, but Tokyo real estate prices were utterly ridiculous in 1990 so measuring from that starting point is like pooh-poohing the post-Ferguson rise in homicide rates because they’re still way down from 1990 (the peak of the crack wars in the U.S.)

          I can recall from back around then that Bank of America sold the five bedroom house in Tokyo it had provided to the manager of its Tokyo branch as a perk, and it made such a ridiculous amount of money off the sale of one pretty nice house that it had to put a big footnote in its annual report to explain that B of A’s global profits for the year were inflated by this one-time transaction.

  3. Jonathan says:

    I think, Phil, that it’s not necessary to attribute motives when a simpler answer is that people aren’t sure what to do. But I frame it differently: why would we expect that SF or Manhattan and a handful of other places would have a broad mix of prices when they’re islands in a larger region and, though these islands need lower-paid employees, why would the market deliver housing for them? In other words, SF isn’t the market; it’s part of a significantly larger market, which you say but don’t emphasize, and it’s kind of silly to expect SF would act like a complete market that houses across the income scales.

    • Phil says:

      Yes, exactly! Perhaps I should have phrased things differently somehow. Rich people consider San Francisco more desirable than outlying areas, so there is big demand for high-income apartments there, and rents are cheaper farther away. If San Francisco were a small island way out in the Pacific, such that everybody who worked there had to be able to find housing there, the economics would be very different. The fact that it is embedded in a larger market is key.

  4. Ryan Moulton says:

    If working class people have to commute from merely Oakland, that will be a dramatic improvement. Right now many commute from Sacramento. The goal is not to locally lower rents, but to globally lower them. This has an enormous number of benefits, from reduced commutes and environmental impact, to allowing more people to benefit from the multiplier effect of the booming local economy.

    • Martha (Smith) says:


      But the question is how to achieve the goal of locally lowering rents?

    • Phil says:

      I don’t see where “reduced commutes and environmental impact” come from. If you build a bunch more market-rate housing in SF, causing median rents to go up via the mechanism I have described, then some working class people will be displaced outwards, and thus have longer commutes. Others who already live outside SF will now meet the increased demand for workers in SF by commuting to SF rather than to lower-paying jobs where they live now.

      Why do you think commute distances would go down?

      • Ryan Moulton says:

        You describe it yourself. Everyone moves closer to the city center, lowering their commute. The people who currently commute from the peninsula will live in SF, the people who currently commute from San Jose will commute from the peninsula, and the people who commute from Sacramento will commute from San Jose. That improves everyone’s quality of life.

        I think there are a couple of key points you are missing in the article.

        1) There are already chronic shortages of services in SF, and most of the bay. For daycares, 150 person waitlists for 50 child facilities are common. Peninsula cities can no longer hire new teachers, because to own property they would have to commute from two hours away. Once people are commuting from that far away, changes to the local housing market cannot possibly price them out any further, nor would any increase in demand for local services. 25% of San Francisco’s police officers live in the city. Many of San Jose’s police officers sleep in cars during the week so they don’t have to make the several hour commute to where they live.

        2) San Francisco is a tiny area of land relative to the average commute distance in the bay. It does not matter in the least whether rents go down specifically there, or 7 miles down the road.

        3) Yimby’s are active politically everywhere in the bay, and want to lower rents everywhere, not merely (or even primarily) in SF.

        • Phil says:

          You say “you describe it yourself. Everyone moves closer to the city center.”

          No no no, this is not at all what I described! I do not think ‘everyone’ moves closer to the city center! Quite the contrary! I think rich people move closer to the center and poor people move farther out. What’s more, the demand for non-rich workers in the city center goes up, so housing pressure within commute distance increases.

          Everplace I can hint of with a YIMBY movement is a place where more market-rate housing will make median rents go up.

          • Ryan Moulton says:

            1. You didn’t propose a mechanism that would make more market rate units displace lower income renters. When Bill Gates walks into a room, the increase in median income does not raise anyone’s wages. Market rate housing is typically replacing underutilized commercial, like car dealerships, not other housing.

            3. You also haven’t made an argument for why the rich people who want to live in San Francisco haven’t already moved there via raising rents and indirectly pricing people out. What qualifies as a luxury apartment in San Francisco would be a dump in Chicago, because Chicago has enough units. 1%ers live in repurposed living rooms in San Francisco with several roommates and lead paint. That would be low income housing in any other city. It’s not the quality of the units that is keeping people out. It’s one continuous market from the high end on down to the bottom. Clearing the high end makes more room at the bottom.

            4. There are yimby movements everywhere. Mountain View, Palo Alto, Redwood City, Sunnyvale, San Jose. You can’t credibly make the argument that if all of them succeed rents will go up.

            5. I think you also misunderstand SFYIMBY’s goals and proposals. SFYIMBY wants to maximize the number of affordable and the number of market rate units. They differ from other groups only in that they care more about the _absolute number_ of affordable units rather than the percentage. SFYIMBY also wants to lower rents in the entire bay area, not merely San Francisco. If low income workers in SF have to commute merely from San Jose rather than Sacramento, everyone should see that as an improvement.

            • Phil says:

              Some of this makes sense to me, some doesn’t.

              I strongly agree with your statement that there is really a continuous market from top to bottom, it’s not “rich people” vs “everyone else.” At any given moment, there are people who are just barely deciding to move into SF rather than a cheaper place elsewhere, and other people who are making the opposite decision. There is always a margin and there are always people on it.

              A goal to have the absolute number of “affordable” housing units in the city go up, that makes sense to me. But I think adding more market-rate housing is counter to that goal, because I think the increase in market-rate housing comes with a decrease in the affordability of housing in SF (but an increase in affordabilty in outlying areas). That’s more or less what my post is about. I think if your goal is to have more affordable housing in SF, then you should oppose building any more market rate housing there.

              I guess I’m not aware of the reach of the YIMBY movement so I should have kept my mouth shut about that.

              I don’t have a strong claim about where I think market rate housing will make local rents go up vs down. I’m confident about San Francisco, much less so about outlying areas. I do think a way to decrease rents in San Francisco would be to build more market rate housing in Oakland and Berkeley and San Jose.

              • Ryan Moulton says:

                >because I think the increase in market-rate housing comes with a decrease in the affordability of housing in SF

                I’m still not clear on your position on this. Consider the price of a bad apartment in SF, currently renting below the average price. A new apartment gets built, and rents for an above average price. Is your claim that the rent charged for the bad apartment will go up?

              • Jameson Quinn says:

                It certainly sounds as if that’s what Phil is claiming. And he’s made two arguments to support this claim. First, that adding market rate housing increases the median price; this is almost trivially true, but irrelevant. Second, that adding more rich people drives up demand; this is probably true, and definitely relevant, but not necessarily decisive.

                Increasing supply with fixed demand lowers prices. Increasing supply and increasing demand could lower them or raise them or neither; it depends. Phil seems to think it will raise them, the YIMBYs think it will lower them; this is a matter of disagreement, not irrationality.

        • Steve Sailer says:

          By the way, back in 2005 in my article “The Dirt Gap” I pointed out that San Francisco is just about the bluest place in America and Dallas-Fort Worth just about the reddest metropolis because of contrasting geography: DFW can expand in 360 degrees making undeveloped acreage not too far from urban centers affordable, while San Francisco is extraordinarily constrained due to its famously extreme geography:

      • Sebastian H says:

        “If you build a bunch more market-rate housing in SF, causing median rents to go up via the mechanism I have described”

        First the mechanism you’ve described doesn’t work the way you have described. It works like in Tokyo, where they build enough to meet the demand and the rents don’t go up. Tokyo is every bit the desirable first world city that SF is, but they keep the rents down by meeting the demand. So you’re already not dealing with the empirically testable reality.

        “then some working class people will be displaced outwards, and thus have longer commutes”

        Second, working class people pretty much don’t live in SF already. So you’re positing displacement of people who have already been expelled from SF.

        The only way to make that make sense is to have the displacement radiate outward. But you’ve already admitted that even if your theory is correct, it would only be correct for a small area like SF, not for the outlying areas.

        Your factual case seems to be based on the idea that selling high cost housing can never lower MEDIAN costs in the nearby areas. You may be having trouble with mean vs. median (although I’m not even sure it is true for mean costs.)

        • Phil says:

          Tokyo seems to me to be a funny comparison, since the Japanese economy is in the doldrums, with declining wages. Even so, Tokyo has something like 2.5 million daily commuters, and (according to the Bloomberg article I quoted above in response to Paul) most new housing there is aimed at the luxury market, and rents have climbed 20% in the past few years.

          • Sebastian H says:

            Yes. As opposed to SF which is like 150% on that period.

            • Phil says:

              So when you said that in Tokyo “rents don’t go up”, you meant OK, they go up 20% compared to salaries but that’s practically not going up?

              Basically I think (1) the economies of Tokyo and the Bay Area are so different that this is not a good comparison, but (2) if you insist on making it, I argue that the fact that Tokyo has an enormous number of commuters and very freely allows building but has nevertheless seen housing prices increase does nothing to bolster the argument that allowing more market-rate building in SF will lead to lower rents there.

              • See ultimately this is the confusion. To a physicist not going up means… They don’t increase. To an economist it means they don’t increase as much as they would have… And so jargon causes all of this… One of the worst cases of mud slinging in the name of lack of communication I’ve seen in a long time.

              • Carlos Ungil says:

                Physicists are not foreign to the concept of partial derivatives.

              • It’s not the ability to understand the concept, it’s inability to communicate what concept is in the mind of each side.

              • Carlos Ungil says:

                Daniel, I think you overstate the miscommunication issue and I really don’t understand how you can (in my opinion) misinterpret Phil’s post as much as you do. Elsewhere you wrote: “If a pre-school teacher can only afford say $1300/mo and the the cheapest thing is $2200 but it would have been $2400 if the market rate apartments hadn’t been built”. This is not compatible with Phil’s argument at all. You may think you’re applying a charitable interpretation, but that would make his whole post completely absurd.

                If he thought that with more market rate housing prices would go down (but not enough) or that the direct effect would be a decline (but could be dwarfed by the overall trend) why would he be perplexed by people who wants rents to go down being in favor of building more market rate housing? What makes these policies “so bad for them” if they are better than the alternative?

                How do you interpret his remark about a new 400-unit market-rate housing tower in San Francisco having the negative impact of higher rental costs everywhere in the city?

              • Just re-read his post and everywhere he says “median rent” replace by “median rent of occupied units”. Just make everything he says about the observed distribution of prices of occupied units. This is the natural thing for a Physicist to think about, like the energy distribution of binding sites. At no point will adding any market rate housing cause more people to occupy units that have rental rates at the low end: $1200/mo etc.

                I think Phil thinks that YIMBY people are appealing politically to the plumbers and store clerks saying “if we build more market rate housing you’ll finally be able to find a place to afford” and he thinks it’s ridiculous, no way is that going to happen…

                How is that so hard to read out of his post? Just replace “median rent” not by “median advertised spot price of available apartments” but with “median rent of actually rented apartments” everywhere in the post.

              • Carlos Ungil says:

                But then what makes these policies “so bad for them”? Why does the fact that some particular statistic goes up matter, if nobody is in a worse situation than before and some people is in a better situation than before?

              • I think the point is that Phil seemed to think the YIMBY movement is appealing to “lower middle class / low income workers” and suggesting that building a bunch of market rate housing will finally allow these lower income people to afford to live in SF and this simply isn’t going to be true any time soon so he felt that it was disingenuous, and in fact his new post suggests exactly that.

      • Dan says:

        Phil, I think it might do you good to start thinking before you post things.

        Tell us, is the net effect of constructing more dense housing in walkable communities that the total number of people living close to the city center is increasing or decreasing? Are more people or fewer people living in neighborhoods where car-free living is possible? Apartments have lower carbon footprints than single-family homes. Is the net number of people living in apartments higher or lower?

    • Jamie says:

      If the goal is to lower rents in the wider market, especially at the lower end of the market, wouldn’t the answer be to build more housing in those outer communities where I assume it would be cheaper / easier to build than in SF? It’s not the answer from a commuting and environmental impact pov but that’s a separate issue to lowering rents.

      Caveat: I am from the UK, have never been to California so don’t know anything about housing there other than following a few YIMBYs on Twitter.

  5. Jason Briggeman says:

    From an author who claims to be so curious about the matter, remarkable how there’s zero reference to or engagement with any actual research on housing markets, never mind there not being any empirical facts in the piece whatsoever (“Remember, this post is not by Andrew…”). All there is, is a half-baked, half-worked-out model with all the complexity of a game of musical chairs (of which the tossed-off first comment under the post easily exposes some major flaws). And not only does the author not know there is a YIMBY movement in Berkeley (maybe he learns about local issues from New York Times op-eds?), he isn’t able to conceive that the matter could be relevant to Berkeley, or to anyone who owns a single-family home in a quiet neighborhood (!?). Perhaps it is just very difficult to understand these matters if the value of one’s house depends upon one’s not understanding them…and see, there it is, there’s the thumb in the eye. Why am I so motivated to hurt others?

  6. Mike says:

    This bizarre post has the ring of a physicist’s logic (

    For a brief, non-refereed but evidence-based study of this situation, see the 2015 California LAO report (

    • Phil says:

      Hah, yes, I do have a PhD in physics, so you got me! That doesn’t mean I’m wrong, though!

      That ‘report’ has even fewer facts than my blog post! And as far as I can tell, none of its references back up its key assertion!

      One thing I should have been more clear on: I’m not saying market-rate housing increases rents; I’m saying that in a market like the Bay Area’s right now, market-rate housing in San Francisco increases rents in San Francisco. Building more housing does decrease rents on average.

      Eh, I guess I did say that, but somehow people are missing it, and if intelligent readers miss something, you have to blame the writer.

      • Ryan Moulton says:

        I think people are mentally substituting “yimby party’s goals” for “decreasing median rent in San Francisco” because “decreasing median rent in San Francisco” doesn’t make any sense as a goal due to simpson’s paradox.

        As you correctly point out, reducing the rent at every tier of housing can still increase the median rent if the supply at the high end goes up more than the supply at the bottom end. That does not mean that the situation hasn’t improved.

        I think a better conclusion to draw from your argument is that median rent is a bad statistic when the composition of units is changing drastically.

  7. afinetheorem says:

    The reason is that there is an enormous theoretical and empirical literature showing that increasing supply does in fact decrease prices at the relevant elasticities. There surely is some of the effect Phil mentions, but it is trivial compared to the effect of having more housing. Indeed, in cities that are not supply constrained, there is close to *zero* link between housing costs and changes in income! A public policy PhD student has arguments that may convince someone who thinks us economists are all deluded on this fact which has been established time and time again, however: . For an anecdote related to Phil’s post, a recent (and very rare) increase in housing supply in SF over the past couple years has *completely* stopped rent increases:

    • Phil says:

      I certainly agree that building more housing decreases housing prices. See the fifth paragraph of the post!

      • Dan says:

        So then why do you go off on a rant about how YIMBYs are just spiteful millennials jealous of your single-family home, instead of supposing that maybe, just maybe, they want to decrease housing prices?

        It’s hard to see some other motive for your rant than intellectual dishonesty.

  8. Andrew Talbot says:

    I’m involved with one of these groups, East Bay Forward.

    Obviously I wouldn’t attribute my opinions to spite, but you’re free to any interpretation as you please.

    A good majority of those new condos and such sell/rent for less than some of the older stock. It’s likely that any demand effects are local. If Berkeley builds better housing people will choose it over Oakland. But they’re choosing the Bay Area either way.

    Regionally, demand seems pretty inelastic. Rents have gone up lots lots lots in the past few years. People are still coming.

    I too live in Berkeley among mostly single family homes. It’s pretty awesome. Many many others are priced out of that awesome. They want to live here, I want them to live here.

    The best path I, and many others, see toward helping them share in our community is to help build more housing, market rate or otherwise.

    • Andrew Talbot says:

      I see you make the point about the regional effect driving up prices IN San Francisco yourself. So yeah, definitely agree that’s a very reasonable concern.

      But, as you say, San Francisco is only 50% as dense as Manhattan. And Manhattan’s a pretty cool place!

      So maybe we should try at least a bit more density?

      • Phil says:

        I’m definitely not claiming that building more market-rate housing in San Francisco would be bad. I only claim it will make median rents go up, and cause farther displacement of lower-income people. I think that’s generally bad but not everybody agrees. Also, even I agree there are positives to go with the negatives in such a scenario.

        A key point is: how high is ‘market rate’? New housing in San Francisco is built for the very wealthy, and I have described in my post why I think that leads to higher rents in general. But if what you say is true, that in the East Bay the new housing costs less than existing (or indeed, even if it’s in the ballpark) that should drive down the local cost of housing.

        • Carlos Ungil says:

          > I only claim it will make median rents go up

          Median rents of the existing units, or median rents including the new units? If the median rent of the existing offer doesn’t move (it’s unlikely to go down due to the incremental offer, because prices are kept artificially low) the median rent of the whole market will mechanically go up (as the new units will be at the market rate, well above the current median). It doesn’t seem a very interesting claim.

          • Phil says:

            I think the median rent in the city goes up if more market-rate housing is built, unless the regional economy slumps.

            • Carlos Ungil says:

              Is this a bad thing? As other commenters pointed out, the goal is not to lower prices, it is to make people better off.

              The effect could be a Pareto improvement (nobody gets worse off): people who lived in SF before could maintain their rents unchanged (unless they voluntarily decided to move) and people who didn’t live in SF before could live in SF (also voluntarily). Who’s harmed in that case? Why does it matter than the median rent rises.

              It’s a bit like saying that opening a retirement home in a neighborhood is bad because mortality rates will go up.

              • Phil says:

                I’m not saying whether it’s a bad thing or a good thing. People who want to live in San Francisco but can’t afford to do so think it’s a bad thing, and yet many of them advocate building more housing for high-income people.

              • Dan says:

                Wait, didn’t you admit before that building more in San Francisco would lower housing prices in the rest of the region? And you still can’t understand why “People who want to live in San Francisco but can’t afford to do so”–i.e., people who live in the rest of the region–would want this situation to take place, other than out of some Trumpian spite?

      • Anon— says:

        Manhattan is much denser than that. SF isn’t even as dense as Brooklyn.

  9. Aside from the fact that this post assumes offensive and incorrect motivations about myself and my friends, this post also makes the same faulty assumption as SF’s nexus study–that new projects are inhabited entirely by new people. (See pg 3 of the nexus study Clearly this is not the case.

    Also, it’s frankly ridiculous to say “I live in a quiet neighborhood of single-family homes across the Bay in Berkeley. I don’t have a dog in this fight.” If you own a single family home anywhere in the Bay Area, you metaphorically own numerous government-subsidized (thanks to Prop 13 and the inability to raise property taxes) vicious attack dogs propagating generational warfare against the young. Single family home zoning in the Bay Area has explicitly racist origins (See the Oakland Magazine article: and restrictive zoning like that of the Berkeley Hills is worse for the environment since residents drive more and live in larger houses that take more energy to heat and cool. Even worse is if people can’t live in Berkeley and are forced to commute from Stockton.

    The best way to make housing affordable is to build a lot of it. Yes, new construction is expensive, but the buildings will eventually get old. It would have been better to build all the housing thirty years ago, but the second best time is now.

    • Phil says:

      I definitely do not assume new housing is inhabited only by new people! I do assume that if you build housing for anothe 15,000 rich people in San Francisco, the total number of rich people in San Francisco will go up by about 15,000.

      • Steven Berry says:

        This is where your accounting is totally wrong, you have the wrong counterfactual. You assume, without logic or evidence, that if you build 15,000 new apartments for rich people then the quantity of rich people living in SF increases by 15,000. This is the thinking of a physicist who doesn’t realize that human beings re-optimize in the face of regulation. If you don’t build new housing, those rich folks are allowed to simply buy the existing housing of poor folks and renovate it to whatever standard they wish. Look at Greenwich Village in NYC. New construction is largely banned there. It used to be inhabited by relatively poor people. Rich people developed a desire to live there and they did so almost entirely by purchasing the properties formerly inhabited by poorer people. This raised prices and drove all poor people out of the neighborhood. This is *exactly* the same process by which almost all poor people have been driven out of SF. Refusing to build housing to accommodate the new demand drives out poor folks.

        You can test your theory if you want. Does the increased number of rich people living in SF exactly equal the (very limited) number of new apartments constructed? No! It is much higher than that, because rich households outbid poor households for the existing housing stock. Now look at Dallas, which is growing very rapidly. There, new housing construction largely absorbs the increased demand and prices do not rise much.

        You are somehow assuming that supply creates (on a one-for-one basis) demand, which is a the sort of error that smart folks often made prior to the development of the supply and demand model in the mid to late 19th century. Ironically, those 19th folks were greatly influence by physical notions of equilibrium, and it is *exactly* the equilibrium response that you are ignoring. When you turn off supply, you don’t turn off demand. Then, to top it off, you assert that experts who disagree with you require some sort of bizarre politco-psychological diagnosis to explain why they don’t agree with your simple accounting errors.

        • Phil says:

          I haven’t seen any experts who disagree with me. And I’ve looked.

          As I pointed out in my post, Manhattan has housing for 1.6 million people and is 50% denser than San Francisco, but rents there are still high. Some people claim that they’d decrease if not for restrictions on building new housing, but, as I pointed out in my post, that assumes that we are at some magical point in time: the last 100,000 units didn’t make prices go down, nor the 100K units before that, or before that…and yet, somehow, the NEXT 100,000 units will make prices drop.

          I know high-income people who had a hard time choosing between San Francisco and the East Bay. There is a strong sense in which they would prefer to live in San Francisco, but they aren’t _quite_ willing to pay the rents there so they live in the East Bay. This is my evidence (and logic!) for saying that if the amount of high-income housing in SF goes up, more rich people will move to SF.

          As I mentioned in my post, I do agree that it’s possible to build so much housing that prices will go down. But that amount might be very very high.

          • Steven Berry says:

            What?? You can’t find any experts who disagree that supply and demand applies to housing? You’re not looking very hard, my friend. In addition to the references elsewhere in the comments, you might start with

            Start with the nation’s top urban economist, Ed Glaeser at Harvard. One of many examples is


            And: you didn’t address my point that you begin with an accounting error. A new 15,000 unit luxury apartment will *not* increase the number of rich households by 15,000.

            • Phil says:

              Of course supply and demand applies to housing! I said that in the fifth paragraph of my post! If you build more housing, the average cost of housing goes down. Totally agree. The point I am making is that the place where you build the housing isn’t necessarily the place where the cost goes down.

              I haven’t seen any economics papers that discuss the spatial variation in housing costs at the relevant spatial scale. Sure, people like Glaeser say “look at Las Vegas compared to San Francisco”, but Las Vegas stands alone whereas San Francisco is part of a megalopolis. People can (and do) commute between San Francisco and nearby cities.

              I’m not saying such studies don’t exist, but I haven’t been able to find them.

              As for my “accounting error”, I still don’t see it. You’re saying “rich people will move to San Francisco whether new housing is built for them or not”, but I’m not disputing that. I’m asserting that more rich people will move there if more market rate housing is built than will move there if no new market rate housing is built.

              I think you’re saying that if housing for 15,000 new rich people is built, that it will all be occupied but this will also decrease the number of rich people who buy existing housing, so the net increase will be less than 15,000. (By the way, just for convenience I’m talking about 10,000 new apartments = housing for 15,000 people, but that’s just for the sake of argument.) This is the “rich people will move there anyway” meme that others have mentioned too. How big an effect do you think this is? Could you give me a number? We have two realities, one in which 10,000 new market-rate apartments are built, and another in which they aren’t. I claim that the city in Reality A has a population of wealthy people that is about 10,000 * (number of residents per apartment) higher than the city in Reality B; you claim this isn’t close to true. OK, how wrong am I, and can you point me to a reference?

              Here is a blog post by economics professor Nick Rowe that gives an economist-friendly explanation for the type of behavior I posit (although he does say ‘take it with a grain of salt’, without saying why the salt is needed). Thanks to commenter “Sam” for pointing this out.

              • Dan says:

                Phil, when are you going to take a rest from battling this silly strawman? YIMBYs are all about how decisions should be made at a local level. Perhaps YIMBY is a bit of a misnomer–anti-NIMBY might be better, because the whole point is that the issue is too important on a regional level to allow policy to be dictated by people opposed to building at the local level (i.e. in their backyards).

              • Steven Berry says:

                The Rowe post says “take with a grain of salt” because he is making extreme theoretical assumptions to show that it is theoretically possible to imagine that second-order effects overwhelm first-order effects, over some range of the data. So “strategic super-complementarity” in population is not theoretically impossible.

                You are in Berkeley, read UC-B professor Enrico Moretti on cities. He is the best urban economist at Berkeley and one of the very best in the word. You really think you are better at this then he is? Key to his evidence is that the complementarity is largely in employment, not population. Once the firms themselves have agglomerated and generated the employment demand (as has clearly already happened in the SF area) restricting housing will drive prices up. Again, you keep trying to live in a world where housing demand in SF would stay constant if we just didn’t build any new housing.

                Your intuitive model that prices will rise in SF but fall in Oakland violates very straightforward model of demand substitution. SF and Oakland are substitute goods in housing. An exogenous rise in one will drive up the price in the other, just as a rise in the price of apples is likely to increase demand for (and prices of) other fruits in the supermarket.

                You are acting exactly like every climate denier. “I haven’t read all the papers, so I don’t believe the experts.” “I don’t believe the result, because I don’t understand it.” “Here is one paper that seems to say something contrary, so it is OK for me to ignore all the weight of other evidence.”

                Above all: “my argument is collapsing, so I will move the goalposts.” You start with “everyone who disagrees with me is operating in obvious bad faith” to “you haven’t cited a sufficient number of academic papers , that I have personally read and understood, to prove that I am 100% obviously wrong.”

                Pretty soon you will notice that empirical papers have standard errors and then you can get a column in the NYT telling us why “scientific uncertainty” means that we might as well believe that demand curves slope up.

              • Matt says:

                Honestly, one very simple thing that has not been made clear, and seems to have been partly the reason for Steve’s frustration is: Are you talking about a shift in the demand curve or a shift in the supply curve for housing? When we observe housing construction, we don’t know if it’s a shift in the supply or demand curve (econ 101). Which are you referring to? Perhaps in this case it’s neither, it’s just removing a constraint which allows the market to reach equilibrium.

              • Steven Berry says:

                Oh, and yes your elementary error is thinking that 10,000 new apartments will cause the population to increase by 10,000, while simultaneously increasing prices. I gave you literally hundreds of references: the supply and demand chapter of every introductory microeconomic textbook written since the end of 19th century. It is clearly not a model that you are familiar with, which is why it is so bizarre that you continue to think yourself better than experts in the field.

              • Phil says:

                I’m not sure where this comment will show up, since we are at the limit of nesting that this blog software allows. This comment is a response to Steven Berry’s comment that begins “The Rowe post says…”

                I am certainly willing to learn from my betters. Let me come back to this point, after first addressing something else Berry says.

                Berry says SF and Oakland are substitute goods in housing, just as apples and pears are in the supermarket…this seems to me to leave out a key fact, which is that people who live in SF spend more of their money in San Francisco than do people who live in Oakland. I don’t even see how to map that onto the apples-and-other-fruits system. Or, rather, I can, but it’s a strain. Tell you what, I will strain a little in order to try to make this comparison map onto what I think happens in the Bay Area housing market.

                Suppose there are two supermarkets in town, Whole Fruits and Safebuy. They both sell apples. For some reason people prefer the apples at Whole Fruits (maybe they’re organic and this matters to some people) but all of the food at Whole Fruits is rather expensive so mostly rich people shop there. Whole Fruits can only get its hands on a fixed number of apples, and it prices its apples so that it barely sells every single one before they go bad. Wealthy apple-lovers who would kinda like to buy Whole Fruits apples, but aren’t willing to effectively outbid the other Whole Fruits shoppers, go to Safebuy instead…and since they’re there anyway, that’s where they do most of their shopping.

                But now Whole Fruits finds another supplier for its delicious apples. What happens? Some of those wealthy apple-loving Safebuy buyers switch to shopping at Whole Fruits, and rather than buying just their apples there, they do their other shopping there too. Whole Fruits gains some wealthy customers, and Safebuy loses them. Whole Fruits can raise the prices a little bit, and Safebuy might have to drop theirs. Providing more of something — Whole Fruits apples — has led to higher costs at Whole Fruits and lower costs at Safebuy.

                Is that a crazy model? (It sure seems to work in the actual supermarket business!)

                To return to being willing to learn from my betters: I really am. As many of you point out, I am not an economist. There is surely a ton of relevant work that I am totally ignorant of. Some of you have suggested papers or researchers to look into; thank you for taking the time. I will read some of these.

                Before writing this post, I had done various google scholar searches for terms like [economics housing prices] and so on, and read a selection of what I found. Much of the emiprical work was written by two-handed economists — “on the one hand, cities with these characteristics behaved like so-and-so, but on the other hand this might be due to factor X rather than factor Y.” Which is fine, actually good: if there are a bunch of factors at work and it’s hard to untangle them, it’s good to acknowledge that.

                When it comes to the spatial distribution of housing prices within a single metropolitan area — a single area across which people are willing to do a daily commute — I did not find much. You can say “you idiot, how could you not have found the work of So-and-so”, and I don’t know but I didn’t. Thank you, Steven Berry, for suggesting that I read the work of Moretti, I will start there since his work looks very relevant.

                But I note that a lot of economic models include assumptions such as “we assume that workers have homogeneous tastes over locations and are perfectly mobile across locations,” to give a real example from one of Moretti’s papers. This particular assumption is clearly not true: there are people here in the Bay Area who can afford to live in a big house with a big yard in a tony Bay-Area suburb like Moraga, or to live in a small house in the Sunset District of San Francisco, and these people choose to do one or the other. The suggestion that the people who live in Moraga and the ones who live in the Sunset would be just as happy switching places — that the tastes of the people in Moraga are the same as those in the Sunset District — is simply false. I mean, come on. I’m not claiming, at all, that this means the rexults of such a model are necessarily irrelevant to the real world — perhaps the fact that some people really want to live in San Francisco and some really don’t is not a big deal and the model still works well enough. But perhaps not, right? How important is the assumption of perfectly mobile workers with homogenous tastes?

                I would like to thank all of the commenters for taking the time to weigh in, and this is especially true of those who provided genuine advice in spite of being thoroughly disgusted with me, since those people really went out of their way.

                I will do some reading and some thinking, and post a follow-up post in a month or two.

              • Matt says:

                Phil! Re: the supermarket example. Why are people willing to pay more at the Whole Fruits when they get more apples? Before they had these apples, they weren’t willing to pay the lower price (i’m calling it the lower price because prices have apparently increased in your story). This example just gets back to Berry’s don’t understand supply and demand. The demand curve here represents the willingness to pay for apples (it slopes downward in all reasonable worlds, meaning that more people are willing to pay for apples at lower prices). Suppose the price at Whole Fruits was 8$ before they got the increase in apples. This means that the people not shopping at Whole Fruits at this time are not willing to buy apples FOR MORE THAN 8$!!!!! If they were willing, the price would be bid up and the equilibrium price would not be 8$. When Whole Fruits gets more apples, consumer preferences have not changed! How the hell could the price possibly rise?

              • Phil says:

                Response to Matt, asking in the Whole Fruits story I made up, “how the hell can the price of apples possibly rise?” It can’t, but I think that’s because of an imperfection in the analogy to apartments, rather than an inherent principle.

                I think that in the world of San Francisco apartments, having more rich people in the city increases the demand for apartments in the city, which drives prices farther upward. I can’t really see how to add that feature to the Whole Fruits vs Safebuy analogy, at least not in a way that seems remotely plausible in the world of supermarket shopping.

              • Whole Fruits puts out apples on Monday morning at 8 AM for $8/lb and they’re all gone by Monday morning at 9:30. The whole fruits manager has a lot on his mind and isn’t going out of his way to set the price of apples accurately, and besides he has no idea how many more people there are willing to buy at higher prices. IT’s a market mess because the quantity is very limited and the information about demand for special high quality apples doesn’t percolate through the market.

                People who can’t shop mondays aren’t willing to go to the store manager and say “gee if you would reserve some apples for me on Tuesday for $36/lb I’d buy them” so they shop at safebuy, and Whole Fruits never really discovers the extent of the latent demand.

                Now Whole Fruits figures out how to get 2 shipments one on monday and one on tuesday, they discover that they can sell all the monday apples and all the tuesday apples. WHOA the demand is way more than we thought! Let’s raise the price to $9. So they do, and now they’re supplying twice the volume at a higher price.

                Is this because the law of supply and demand implies somehow an upward sloping demand curve? No. It’s because *THE MARKET IS ILLIQUID AND EQUILIBRIUM THEORY DOESN’T APPLY*.

              • Note also that it’s totally the norm for SF apartments to get listed on craigslist for some asking price and then multiple people show up with bids HIGHER than the ask, and the whole thing clears at an unobserved higher price… So non-equilibrium market conditions are *standard operating procedure* in SF.

                treating SF as an equilibrium downward sloping demand, upward sloping supply crossing at a well observed spot price… it’s the spherical racehorse in free space.

              • Matt says:

                Daniel: man, stop shifting the goalposts. It’s clear that Phil did not have this in mind in the original post. You keep making up these stories that we have no empirical evidence for. While sometimes I find your comments useful, I find it annoying that I have never EVER seen you make a concession to somebody’s arguments. You always have a “but” to put in, and then go off on a tangent (like earlier where you brought in the idea about rent-controlled apartments being freed up”. Again, explain to me how Phil’s comment about Manhattan is consistent with anyone who has a basic understanding of S&D framework.

                With regards to your actual comment… sure there are market frictions, but a lot of times simple S&D framework is a good approximation. I think you are underestimating the amount of time/energy put into pricing at supermarkets.

              • Matt says:

                Daniel again: Plus, if your going to explain things in this sloppy manner, you can pretty much explain anything.

                The Whole Fruits manager puts out 20 apples at 8$/lb on Monday morning. By Wednesday, only 10 of the apples have been sold. The manager shrugs his shoulders and says “well I guess people just don’t like my apples!” and he decides to just leave the prices as his..after all, he’s not willing to figure out if there would be more demand at 7$/lb.

                Then, the next week he gets more 20 apples and its big news in the city, so a lot of people show up. But, he finds that he still only sells 10 (out of 40 apples!). Hmm, says the manager, maybe I’ll decrease the price! So he does, and he is able to sell all the apples for 5$! Wow! Increased supply has decreased price. What a shocker! By this example, I have proven that housing prices in SF will decrease due to supply increase.

              • Matt: You say “It’s clear that Phil did not have this in mind in the original post”

                It may be clear to you, but it was pretty clear to me that this is exactly the kind of thing he had in mind. And that’s my point, no one is engaging the contents of Phil’s actual post.

                Of course, when prices are out of equilibrium you could get all sorts of behavior. I’m not claiming that I know how people price fruit, that’s just helping Phil with his analogy. I’m simply pointing out that In SF and Manhattan it’s VERY CLEAR that the prices *ARE* out of equilibrium. That’s exactly what Rent Control does. That’s the PURPOSE of rent control. And so since Phil exclusively is discussing rent controlled regions, the fact that his model for what happens doesn’t follow the predictions of equilibrium price theory is no evidence of anything.

                Elsewhere, Phil explicitly talks about being interested in the time evolution of the distribution of rents of occupied buildings:


                and yet, here we are discussing how whatever you interpreted him to be discussing is actually about the time evolution of the spot price of unoccupied buildings…

                I’m not switching the goalposts! Please address how Phil’s model of the time evolution of the distribution of the prices of occupied buildings is wrong and how the building of market rate housing on the far right of the spectrum will in fact extend the distribution of the prices of occupied buildings farther left as well… I’d like to learn.

              • Matt says:

                Daniel: Honestly, I don’t think there is much point in pursuing this further, it is not possible to talk about these ideas without a concrete model, which you clearly don’t have.

                But, where is this evidence that Manhattan and SF’s prices are not in equilibrium? I agree that rent control will prevent equilibrium. But literally all I’ve seen Phil write about Manhattan is that the first 1.5 million people moving there didn’t decreases prices, therefore why would the next 100,000 decrease prices? This is just a stupid question to ask. Things could very well be in equilibrium or approaching it in Manhattan, demand is just continually rising.

              • Here’s a specific model for the time evolution of the distribution of prices of occupied units in SF:


                Can you describe some aspects of what I suggest there that are obviously wrong? I don’t claim that it’s all definitely right, but I don’t see anything that isn’t at least quite plausible in heavily rent controlled SF.

              • Matt says:

                Daniel: You are just giving an unbelievably (truly, unbelievably) charitable interpretation to Phil’s original post. He doesn’t even mention rent control in the post, and then in the model you propose literally the entire mechanism is about rent control. Look, the model you wrote down is possible – I don’t see any evidence for it, but it’s possible (a new market-rate apartment causes some dude to move from his rent controlled apartment (although really, how many ppl are leaving rent-controlled apartments?) which allows that unit to move up to the market rate). But this is just not what Phil had in mind. Take this quote from Phil:

                “3. Now instantaneously add a bump in the distribution in SF, representing 10,000 new units that are up in the high end somewhere. What does this bump do to the equilibrium distributions, if we wait?

                I think some rich people move out of the “outside” area (and take their money with them), and that this causes distribution #2 to shift downward.
                I think they move into SF (and take their money with them), so that distribution #1 shifts upward.”

                How the hell is distribution #1 shifting upward? Again, you would say its possible because of this rent-controlling cascade mechanism, but its really clear to me that Phil is forgetting that these rich people didn’t move in beforehand because they COULDN’T AFFORD IT. They certainly aren’t going to move in after when prices are higher (as you claim they will be). Please don’t respond to this with a comment about market frictions, because that can go both ways.

              • Matt: mechanically by moving out of area 2, and occupying the new high rent sites in distribution 1 the distribution of occupied houses in area 1 shifts upwards.

                It’s like you have a bag of numbers and you throw some numbers into the bag that are bigger than typical for bag 1, and so now the distribution of the numbers in the bag is more mass above the median. It’s a mathematical fact that is indisputable.

                Now, further, secondary effects are that definitely some of the renters in area 1 (SF) leave their apartments to take the new fancy digs, and so there’s shuffling *within* area 1, and *every time* you shuffle someone within area 1 the rent on that apartment goes up because rent control. So, any affect that an economist wants to posit in which adding additional housing changes the prices of existing housing doesn’t actually result in observed rents on any individual occupied units to fall vs the price they were rented at before the new buildings were built.

                I think that’s the entirety of what phil’s original post was about. Literally re-read his post and everywhere he says “median rent” replace it with “median rent of occupied units”

              • Matt says:

                Daniel: You say:

                “Matt: mechanically by moving out of area 2, and occupying the new high rent sites in distribution 1 the distribution of occupied houses in area 1 shifts upwards.”

                Yeah, my point is these people aren’t moving in. If they didn’t want to move in beforehand, then they certainly aren’t going to want to move in afterwards (if these homes are even higher in price). But, you say, these new units will be occupied. Yeah, sure they will be, but not by people who weren’t already in SF. Look, the fact that you won’t consider the possibility that an exogenous shift in supply could decrease prices is mind-boggling. You can propose all these fancy mechanisms, but the first-order effect is a decrease in price, it has to be. The reason were not gonna observe a decrease in price is because demand is not constant, and its not an exogenous shift in supply, it’s a response to demand. But if it were, there would be a decrease absent this rent control thing your pushing or some other crazy second order effects.

              • Matt, again I totally 100 percent agree with you that *in a market equilibrium situation an increase in supply will lead to a reduced spot price all else equal, and will lead to a less severe increase in spot price if demand is also rising*

                I just don’t think that is relevant to what phil was actually talking about. He was talking about *what would happen to the distribution of the rental rates of occupied housing in SF some reasonable number of months after opening the doors on some expensive new units* and *would it increase the number of people occupying units at the “affordable” end of things*

            • Matt says:

              Phil: Okay glad you agree. I think Berry has already spelled this out though – for prices to actually increase from an exogenous shift in supply, would require some really crazy second order effects! From reading your comments, I think its clear that you are not understanding that these “rich people” you reference, who aren’t living in San Fran before the housing gets built, have an effect on prices before the supply shift. In a simple S&D model, before the new housing gets built, we know that these people have a willingness to pay for SF housing that is lower than the market rate (or else they would be in SF, and the market rate would be higher because they would outbid the marginal guy). So when more housing gets built, the first-order effect is that it has to be at a lower price for these rich people to move in, because we know that they are not willing to pay the market-rate. If the new housing is at a higher price, nobody would move in, we already know they aren’t willing to pay. For second-order effects to somehow counteract that some really crazy shit has to happen.

              Now, where you seem to be getting confused, judging by your reference to Manhattan, is that in reality demand has been steadily rising while new housing gets built (in fact, this is likely the reason new housing is being built). So when we observe a positive correlation between housing supply and prices, this is not evidence for Nick Rowe’s crazy mechanism – this is just Econ 101 – demand for housing shifted outwards in SF and more houses were built, leading to higher prices and more housing.

              I would not suggest going to read academic papers yet, as I don’t think you’ve grappled enough with the basics.

          • Garret Christensen says:

            I too am in shock. Just to name your fellow UC BERKELEY professors alone, there are numerous.

            John Quigley (

            Carol Galante (

            Enrico Moretti ( “People are marching against Google buses when they should be marching for more housing permits.”)

            Find me a serious quantitative policy analysis professor (not a physicist) at UC Berkeley who says we shouldn’t build a lot more housing here and I’ll deliver a six-pack of your favorite beer and apology to your office.

          • Danny says:

            >I haven’t seen any experts who disagree with me. And I’ve looked.

            This has to be one of the stupidest things ever said on this blog. Let me know if you need helping doing a google search a 7 year old could probably complete.

        • Garret Christensen says:

          I was going to reply with exactly the same thing. The choice we face is: (A) the relatively-wealthy (or their investors) can buy the existing low-end housing stock, evict the low-income tenants, and renovate OR (B) we build a lot of new apartments and the relatively-wealthy to move into those and take pressure off the remaining low-end housing stock.

  10. Lewis Lehe says:

    “Serving those extra 10,000 high-income households will require tens of thousands more waiters and shop clerks and car mechanics and plumbers etc etc etc….that is, there will be more jobs for the kinds of people who already have trouble affording a place in San Francisco.”

    Try to model this. Currently, there are jobs and housing available in SF, so everyone who wants to move there at current rents and wages probably has or is in the process of doing so. Therefore, your hypothesized service workers would only move to San Francisco if they attained more utility from doing so. No one is forced to move to San Francisco simply because new tech workers have created more demand. Service worker may enjoy more utility there because the new housing makes rents fall but sustains current nominal wages (the YIMBY theory), or because the new housing raises rents (your premise) by less than it raises nominal wages.

    So perhaps, according to your theory, rents might not fall, but only if wages rose very quickly. That doesn’t seem like a strong argument against YIMBYism. It is important to the country and to workers’ that wages rise. When wages rise, the government collects more in tax revenue. The gains to workers from a rise in utility is tautological. This idea should not be surprising. Urbanization accompanies increases in national wealth and wages.

    Are you an economist? I found it interesting that there are no references to economic models or literature, only intuition that doesn’t seem rigorous in the way that economists’ intuition is. The majority of economists hold that coastal housing markets are expensive because of supply constraints. Some economists differ, however: for example, Peter Navarro.

    • Phil says:

      My claim is not that more service workers would move to San Francisco, it’s that there would be more service jobs (and other jobs) in San Francisco. I think most of these would be filled by commuters.

      Whether this scenario I good or bad is a separate question that I don’t want to get involved with.

      • Lewis Lehe says:

        Okay then suppose the additional jobs are filled by commuters. What would entice additional service workers to commute to San Francisco? Only higher wages. So the scenario you’ve outlined is one where rents fall elsewhere in the Bay (your claim), and service workers’ wages rise in SF. So service workers are paying lower rent where they live and earning more money. Moreover, only higher wages could possibly drive an ancillary increase in San Francisco rents, which you claim would happen. This sounds like a good outcome. It isn’t surprising that YIMBY’s advocate for a policy that results in higher wages and a lower cost of living for service workers. Many YIMBY’s in fact live in the east bay. See East Bay Forward, for example. The psychological theorizing is unnecessary, and the policy is welfare-improving under your own admission that rents fall elsewhere and the fact wages must rise in SF.

        The only downside could be increased congestion. In that case maybe you should have written: Why do YIMBY’s care more about economic growth than traffic congestion? That is another issue. I would add that SF congestion is solvable. The city is considering congestion pricing on downtown streets, and congestion on the bay bridge could be eliminated or reduced to any chosen level by a sufficient increase in the toll. The city is also spending more on transit, e.g. the central subway. While the central subway will not eliminate congestion for those that drive, it will allow more people to avoid it. Moreover, the added tax revenue from higher wages, higher property values (your claim), a larger housing stock and more tech workers could be spent on transit and road capacity. One possible project would be a tunnel to take Caltrain downtown. This would cost about $3.9 billion ( Amortized at 5% this would be $200 million per year. Given federal and regional matching funding, gains in taxes from higher property taxes, and more fares, such a project would probably be affordable for SF.

        • Phil says:

          if BARFs and YIMBYs were arguing ‘we know that building more market-rate housing here will make housing prices here increase, but we want to do it anyway because it’s good for the overall welfare,’ that would make perfect sense to me. But that is not the argument I see them making. The opposite, in fact.

          • Lewis Lehe says:

            They do not acknowledge that building more housing results in SF would result higher SF rents because there is no evidence of any such thing. Do you have evidence? I have never seen any and doubt it would be possible to establish with any rigor, much less so obviously that a bunch of non-econometricians can be assumed to, privately in their heart of hearts, “know” it to be true. As you probably do know, writing on this blog, it is impossible to establish the slopes of supply and demand curves through OLS. One would need instrumental variables or some other technique which has no appeal to casual intuition and cannot be observed casually. And to identify an effect like that for a *particular* city, a small one at that, would be extremely difficult. The fact that the amount of new construction in the city of SF is so small (about 0-.7% growth in stock per year) and has remained in such a narrow band for many decades would complicate things further, given that the extant housing stock deteriorates at a rate of comparable magnitude. I have only seen one attempt to do some statistics on rents and demand for the city proper, and it reaches conclusions opposite to yours, although it is not rigorous.

            As for regional statistics, evidence suggests supply constraints are responsible for higher home prices. Joseph Gyourko has long done research on this, and I think he is coming out with a book about it soon with Ed Glaeser. I still don’t understand why you haven’t referenced any evidence or explained why you oppose what could easily be called the consensus view among urban economists. A simple google search on the topic returns, for example,

            But in general, I think I may have confused you with my argument. I was accepting the premises of your argument about new housing increasing home prices and pointing out that new housing would, in spite of rising prices, be a significant boon to social welfare, albeit by a channel of rising wages rather than falling prices. In fact, in such a world, the degree of price appreciation would actually be an index of the gain in welfare, as prices would rise monotonically with wages, tax revenues and workers’ utility. There are especially good reasons to favor things that raises total tax revenue for given tax rates (see hendren’s arguments My point about rising prices would be obvious if we were talking about policies other than housing: if a new subway led to higher prices near a stop, we could assume it is providing substantially better accessibility. Few people argue for destroying infrastructure as a way to lower home prices.

            But the truth is that, while I can accept your claim for the sake of a hypothetical, I find the premise of rising prices from new construction dubious. The regional statistical relationships point in the opposite direction, and there are so many factors affecting home prices that the additional demand of auxiliary service workers in your scenario is probably not very important to home prices. Do some basic math on your channel of causation. If someone makes $100k per year (your statistic), they can generate, potentially, up to $65k in after-tax spending (after sales, income and payroll taxes). If they are spending $3k/month in rent, then they have about $30k to spend. Now subtract all the money saved, paid toward loans or spent on goods and services made elsewhere (e.g., food materials, amazon, travel, cars, etc.). After acknowledging all that, my intuition says the “aggregate demand” effect is not large relative to random noise, interest rates, animal spirits, stock prices, and supply-and-demand in the housing market.

            • A says:

              This is the frustrating part of this post. This guy doesn’t see the assumptions packaged into clumsy model, and he is so arrogant that he feels confident enough to impugn opposing views. He can’t even express those opposing views due to his sticky framework.

              • Andrew says:


                I know Phil well and can assure you that he’s not arrogant. He may be ignorant on this one, but he’s trying his best, and I think he’s responding as well as he can to the mixture of arguments that he’s seen.

              • A says:

                My apologies. This is an emotional subject because motive manufacturing is a frequent tactic of NIMBYs. Last week an SF publication argued that YIMBYs are actually racists.

            • Phil says:

              An argument that YIMBYs are racist? That’s a new one on me. But it doesn’t surprise me. I’ve seen the NIMBYs are racists argument, but not the YIMBYs.

              Andrew is right that I’m trying my best!

              People won’t believe this but I actually thought I was phys-splaining a recognized phenomenon, that gentrification begets gentrification. Rich people move to an area, suddenly they all want fancy food and nice cars etc etc so there are more jobs for the people who provide those things, and prices go up and poor people get priced out. Honest to god, I thought this was the standard model for this sort of thing!

              And I’ll say again: I don’t think building more housing leads to higher housing prices. Housing prices go down. I do think that when more rich people move to an area, prices for everything (including housing) go up in that area. But those rich people moved out of some _other_ area, and prices there go down. The net effect is a decrease in housing prices. If I have this wrong, I really want to know where and how, and I mean it!

              • Carlos Ungil says:

                I’m not sure if you’re missing the following point. Let’s say at T=0 there is some amount of housing. At T=1 there is more housing. You say rich people will move into SF at T=1 driving housing prices up. Why didn’t they move into SF already at T=0? Because it was more expensive!!!!!

              • Carlos: NO, they didn’t move at T=0 because *THERE WAS LITERALLY NOWHERE TO GO*. At T=1 there is more housing driving a certain amount of liquidity which causes a turnover that shifts all the prices *towards* their equilibrium price… which is a lot.

                In general, equilibrium never occurs because of illiquidity, lack of information, and imperfect substitution between locations. People don’t just want an apartment in SF, they want one in some specific area that’s close to their work, or has easy commute, or whatever, and so restricted to that subset, there is literally maybe 1 or 2 or 10 apartments available per year for each person in the market and each one has 14 applicants within the first day on the market, and the actual clearing price is not even observed because it’s private to the landlord and the renter. In this environment, each landlord advertises at low price to attract bids, and then accepts whatever the highest bid from the best tenant was.

                By increasing the available housing stock, the liquidity increases, the turnover is much higher for a while, and the information in the market increases, and the observed spot price increases towards its equilibrium level. So, yes perhaps unobserved clearing prices went down, but advertised prices increase towards equilibrium and the occupied apartment’s distribution of monthly rates shifts right.

                It’s a mistake to pretend that SF advertised prices are in equilibrium.

  11. Steven Berry says:

    I take it you have never taken a course in either undergraduate introductory microeconomics or in empirical economics? The direct effect of new apartments is to drive the price of existing housing down. It is hard (though not impossible) to rig a model to get your various second order stories to overcome that. The empirical work on housing supply strongly suggests that increased housing supply drives down housing prices. Just as an illustration, growing cities that allow housing to be built have low housing prices (Dallas) while growing cities that constrain supply (SF) have high prices that drive most poor folks out of the city. I am not saying that market supply solves all housing problems, but the evidence suggests that, in markets with increasing demand, simply restricting supply hurts the poor by leading the rich to bid up the price of the fixed housing supply.

    In short, YIMBYs are following consensus expert opinion among housing economists. Your position is similar to “climate skeptics” and “supply siders” in relying on untutored intuition while ignoring evidence and strong expert consensus.

    I think you are looking at the effects of increased demand, where price and quantity both rise, and then you get confused and attribute the price rise to the quantity change. In fact, restricting quantity in the face of the increased demand will make the prices rise even more, as we see in SF, where NIMBY policies lead to little new construction and the poor are driven out.

    • Phil says:

      Of course building new housing will decrease the average price of housing. I said that in my post.

      • Steven Berry says:

        You said, clearly, that supply restrictions within SF would decrease prices in SF, as opposed to the counterfactual where quantity supplied is allowed to adjust to increased demand. That is, you disagree with the YIMBY principle that prices would be lower if quantity supplied is allowed to adjust as demand increases (relative to the prices that would obtain when supply restrictions are put in place.) Your argument violates any simple model of supply and demand in a differentiated products market. It is true that you could try to write down a pretty sophisticated equilibrium model to get your result, but it would depend on many fine details where second-order effects overwhelm first-order effects. In practice, though, your argument is based on a simple accounting error which I outlined above.

        You are making an error that we teach undergraduates to avoid by the second week of undergraduate economics. You need to distinguish a change in demand (or supply) conditions (the “demand and supply curves”) from an equilibrium change in demand and supply quantities (which are responses to those more fundamental changes in the “curves”). You are making the classic undergraduate mistake of believing that a change in quantity supplied, in response to a shift of the demand curve, causes a further increase in the demand curve. It is easy to do when you are arguing verbally and haven’t written down the equilibrium model. (Indeed, it is tempting to give your essay to undergrads to see if they can identify this classic mistake.) You are treating something as an accounting issue that is in fact an issue of equilibrium outcomes.

        You also refuse to address a large empirical literature that uses a combination of instrumental variable techniques and basic economic modeling, a literature that almost always comes to the opposite conclusion of yours. Now, on a site devoted to statistical modeling, we could have an interesting discussion of the validity of those instrumental variables and so-forth. That would be a worthwhile discussion, but that is not what you did. You went straight for ad-hominem: anyone who disagrees with your amateur noodling must be self-deluded and/or have bad motives. It is particularly weird because the YIMBYs are following the greatest weight of expert opinion, whereas NIMBYs are typically pushing pure self interest, as the (relative) price declines from new supply will limit the capital gains of incumbent home owners. Under your theory, it is the NIMBY home owners who are bizarrely working against self-interest.

        You might start with Matthew Yglesias’s “The Rent is Too Damn High” and then move on to the academic literature.

    • David18 says:

      Scarcity leads to higher pries.

      The high cost of housing according to Microeconomics is a market failure or market inefficiency caused by politics called “rent-seeking.” In this case, it is both zoning density restrictions and overuse of historic landmark status.
      It is a regressive tax where wealthy landowners such as President Trump benefit at the expense of those who rent or are purchasing housing.
      By removing the “rent-seeking” in any situation, but in this particular situation there are enormous social and economic benefits.
      Apartment renters can spend more of their income on goods and services, it stimulates a construction boom, and it addresses perhaps the major cause of inequality by making more affordable housing.
      I live in NYC and it is true here, in SF, LA, Washington DC, Boston, …
      David Ricardo first addressed “rent-seeking” in the mid-19th century related to the “Corn Laws.” He joined Parliament in order to overturn the “rent-seeking.”

  12. That’s an insider-outsider problem, made worse by local democracy (e.g. You expect to need an independent social planner in these situations.

  13. tcheasdfjkl says:

    As an SF YIMBY who generally expects building housing to make rent go down, this is perplexing and generally makes me less likely to take explanations of this nature seriously in the future for political positions that I disagree with. Are you not even going to consider the possibility that people actually disagree with you about facts?

    • Phil says:

      I’m happy to consider any possibility, but this seems…well…OK. Please do me a great favor by explaining how building more market-rate housing in San Francisco could make the median rent in San Francisco go down. I understand how/why it will make the median rent in the greater Bay Area go down, but I don’t see how it could decrease rents in San Francisco or even in the surrounding cities that are considered most desirable.

      • Steven Berry says:

        Indeed, as a favor, I will give the answer as an algorithm:

        1. Go to ebay and buy a used intro microeconomics textbook
        2. Read the chapter on supply and demand
        3. Draw a supply and demand graph, with the vertical axis labeled “Housing in SF” Draw demand so that it is not perfectly inelastic, accommodating the fact that people can move to other towns, etc. (and thereby isolating the effect to SF prices and quantities)
        4. Draw an increase in demand, a rightward shift indicating an increased desire of the rich to live in SF.
        5. Examine the change in equilibrium price and quantity. Note that the increase in equilibrium quantity is *not* equal to increased demand (the rightward shift of the curve) as some of the demand is “rationed out” by the increased price. Thus, the increase in demand does drive out some of the lower willingness-to-pay consumers (“the poor”.)
        6. Now re-draw the graph, assuming that NIMBYs have passed a law forbidding the increase in quantity supplied. Now the supply curve is vertical, the NIMBY dream.
        7. Contrast the two graphs. Note that the “displacement effect” is much larger in the NIMBY graph! *All* of the accommodation to new demand is now done by price not quantity. The NIMBY price is higher that the YIMBY price. The rich have simply moved into the existing housing stock. Those with lower willingness-to-pay for housing are driven from the market, to a greater degree than in the YIMBY graph (the original graph).
        8. As an advanced exercise, note that those driven from SF will typically move to other nearby towns. This increases the demand in those towns, driving the price up in those towns as well. Thus, the NIMBY policy results in higher prices everywhere, as compared to the YIMBY policy.

        So in that model, the YIMBYs are correct.

        OK, what’s your model? In simple math or graphs (not “blah-di-blah-di-blah”) and at least consistent with some equilibrium model of supply and demand that distinguishes changes in demand and supply (the curves) from changes in equilibrium quantities. As a hint, in all of your stories, you left out the initial change in demand, and you assumed that supply creates its own demand (one for one). That is, you assume that a new apartment building *causes* people to move to SF, as opposed to being *caused by* the increased demand.

        • Carlos Ungil says:

          I agree that the marginal effect of more supply will be lower price. I just want to add that, as some people has pointed out, there are a couple of subtleties that can be confusing when looking at the existing rents.

          1) there is a market for each set of comparable rental units. The average market rate (median or mean) might go up even if the market rate goes down at every level. Adding volume at the upper levels pushes the average up and can more than offset the global decline.

          2) most rental units are not in the market. At each level, the average rental unit is a weighted average of “old” rents (far from the market rate) and “new” rents (close to the market rate). If market rates go down there may be more activity in the market, allowing some old rents to reset to the market rate. The average rental rate might go up masking the clearly positive trends on the available rental units (higher volume and lower prices).

        • Phil says:

          A commenter named Sam pointed out a discussion that captures the essence of ‘my’ model. I certainly do not claim to have invented this model…indeed, I thought it was conventional wisdom!

        • Phil says:

          Thank you very much for taking the time to write this.

          I have a passing familiarity with microeconomics.

          1. Draw a lognormal distribution with geometric mean 700,000 and GSD 2. This represents the statistical distribution of cost-per-square-foot, normalized to the typical dollar value of a small condo, in San Francisco. (The exact numbers don’t matter). Multiply the whole distribution by N_sf to get the number of units at each dollar value. This represents the equilibrium distribution of housing costs in SF, if we imagine reaching an equilibrium with the current housing stock.
          2. Draw a lognormal distribution with a geometric mean 350,000 and GSD 2. This represents the statistical distribution of cost-per-square-foot in cities neighboring San Francisco. Multiply the whole distribution by N_outside to get the number of units at each dollar value.

          Taken together, 1 and 2 represent the equilibrium if the current housing stock in the whole Bay Area were frozen and held there… the “NIMBY dream.” As we are constantly reminded here, although the Bay Area housing stock is not actually frozen, there are lots of regulations and impediments that make it expensive and time-consuming to build and that place impediments on building. We are surely closer to the NIMBY dream than the frictionless-free-market dream.

          3. Now instantaneously add a bump in the distribution in SF, representing 10,000 new units that are up in the high end somewhere. What does this bump do to the equilibrium distributions, if we wait?

          I think some rich people move out of the “outside” area (and take their money with them), and that this causes distribution #2 to shift downward.
          I think they move into SF (and take their money with them), so that distribution #1 shifts upward.

          In the scenario I have outlined, there is no righward shift indicating “an increased desire of the rich to live in SF”. In my model the desire of the rich to live in SF is constant, and the number of rich people living in SF is limited by the housing stock. If you build it, they will come….and if you don’t, they won’t.

          In your model the desire of the rich to live in SF is at some level now, and your model predicts what will happen if that level increases (your #4). I understand why and how your model predicts that the marginal increase in housing prices will be less if more housing is built than if it isn’t, and I agree it is internally consistent. I’m just not sure that it’s a good model.

          In a nutshell:
          I think we have a current state in which the housing stock in SF is not so far from the NIMBY dream of being fixed; the housing stock in the rest of the area is not so far from the NIMBY dream of being fixed; there are a lot of fairly wealthy people who would like to live in SF ‘if they could afford a decent apartment there’; and we are talking about what happens if we create a new bump of high-end housing in the SF statistical distribution. I have explained in my post exactly what I think will happen and why. You say it’s wrong. Maybe it is. But I still don’t understand why.

          Look, I know this seems like amateur hour to you. I appreciate you taking the time you have already taken. But could you (or one of your students) tell me what is wrong with my way of looking at it? If I am being obtuse, I am not being deliberately obtuse.

          • Phil, I don’t think you seem like amateur hour at all. I think all the economists have piled on without even reading what you say, they are making assumptions that you like most everyday people just don’t understand economics.

            Things that I think are causing people to talk past each other include:

            1) no one has specified scales. If you increase the market housing by a factor of 3,000 times as much as currently there (so around 3 apartments per person in the country?) pretty obviously the prices are going to come down. So, what is the scale of the actual changes we’re discussing? My impression is you’re talking about a factor of 1.01 or so

            2) How about time scales? Equilibrium results are only interesting once equilibrium is achieved. But in SF slow changes in housing are the norm due to rent control locking up apartments for the full 60 year duration of a person’s life, and so you really need to discuss not “at time t = infinity which is probably something like 100 years out” but at some specific time. Suppose we increase the housing stock by 1.01x whatever it is (around 8000 more apartments maybe?) what will happen 1 year after they open the doors? 2 years? How does the short term tech bubble invalidate these concerns?

            3) Definitions: what metrics are we discussing here? When you say average prices will increase, are you talking about the average rent across all occupied units, or are you talking about the average offering price for the unoccupied units? When you talk about “going up” do you mean in dollar terms or in dimensionless terms of rent/income? Do you mean actually go up, or go up faster than the trend?

            I think you are absolutely right about the following things that I’ve inferred you mean:

            1) For the marginal 1.01x increase in market rate housing, the average cost of housing across all occupied units in SF will go up. (but maybe we shouldn’t care about this statistic)

            2) For the marginal 1.01x increase in market rate housing, this will not offset the growth in housing costs sufficiently to cause a noticeable decrease in the dollar cost of an unoccupied unit in the short term all else equal including that the tech bubble continues through past the opening of this new housing. The growth rate of housing costs due to high growth rate in demand is just swamping out this scale of effect.

            3) For the marginal 1.01x increase in market rate housing in SF, there will be a marginal small decrease in the dollar cost of housing in the surrounding areas maybe if the demand in those regions isn’t increasing too fast, or maybe there will simply be a reduced rate of growth in those surrounding areas if the demand there is increasing.

            More to the point, I think to first order we could state your model as:

            dP/dt = K(t) – dN/dt

            on some dimensionless scales. And you are claiming that the forcing function K which represents how rapidly we’re recruiting people to want to come to the SF area and get high priced jobs dramatically outweighs any reasonably achievable building rate dN/dt on this scale. Like, let’s say K = 1 and dN/dt is going to be at best .02 and so absolutely nothing we do to increase dN/dt from 0.00 to 0.02 is going to make dP/dt NEGATIVE

            Hopefully being more mathematically explicit could help people.

            • Also, the decrease in housing prices which may or may not have occurred in the last few months needs to be explained, and there are possible explanations that include:

              1) the threat of new housing is inducing landlords to reduce prices to fill their existing houses

              2) The threat of changes to the Fed interest rates, the new president’s economic ideas, and a long series of problematic issues in the tech industry (Yahoo implosion, huge losses from Uber, etc etc) are reducing the forcing function K over the last few months.

              and there’s no clear answer on this time scale of a few months as to what is going on.

            • Carlos Ungil says:

              Unless I’m mistaken, he’s not claiming that there is a rise in demand and supply cannot increase as fast (so prices will rise despite an increase in supply). In your model, prices will be increasing less at the margin when new capacity is added. He maintains that increasing supply will result in higher prices, relative to the alternative universe where supply doesn’t increase (“the negative impact of higher rental costs everywhere in the city “).

              • See, I think this is right there at the heart of confusion. Phil needs to clarify, because I think Phil is saying simply what I’m saying, which is that observed prices will still go up not down no matter how much housing you build so long as you keep it within the realm of politically feasible building rates.

            • Steven Berry says:

              I do appreciate that you are trying to lay out a model, because it makes your errors more concrete. That is indeed an advantage of being more mathematically explicit.

              You specify a rich parametric model of housing prices, albeit one not fit to data. You treat the supply side as fixed, which as you say given the NIMBY success is maybe first-order correct. Now, you need to add a reasonable demand specification to your model. That is the demand part of supply and demand. The post-construction distribution of housing prices is not something to be “supposed” via numerical illustration, but is to be derived from preferences (demand) and equilibrium conditions. The fact that you don’t have a reasonable demand curve is the problem and makes me question your familiarity with microeconomics. David Lakeland’s suggestion that you be more “mathematical” about a model not derived from supply, preferences and equilibrium will not help. The point of math here is to be able to discuss such equilibria, not to evade such discussion under a blur of ad hoc equations.

              As just one example, you are implicitly forbidding any substitution between the existing housing in SF and the new apartment stock. You are enforcing an odd rule that the new apartments do not compete with other SF housing, but only with outside SF housing. One would have thought that “within SF” housing stocks are actually *better* substitutes than stocks in and out of the city. In that more reasonable model, the addition of new SF housing will drive down the price of existing SF housing and thereby *help* lower income people. This is the essence of the YIMBY argument and their model is much better than yours because they assume that within SF housing stocks are substitutes and their model involves clear notion of equilibrium, whereas you are evading within SF substitution and then simply assuming your answer.

              Furthermore, the *actual* history of SF is that there has been a huge increase in the demand to live in SF. This is due to the movement of the world technology industry to the Bay area. Failing to build housing in the face of the demand increase will cause (and has caused) a huge increase in housing prices, greatly benefiting existing Bay area homeowners (like you?) and giving rise to a huge NIMBY movement (which is, on the other hand, also supported by well-meaning folks who don’t understand supply and demand.) You can “assume to the contrary,” but such assumptions should not drive policy.

              And, if you want a model of within Bay area housing markets that measures cross-neighborhood substitution, read Pat Bayer and co-authors in the Journal of Political Economy, Your exact question is not the focus of the paper, but an implication of their estimates is that neighborhoods are highly substitutable in demand, both within and across city boundaries. Therefore, you should not shut down within SF housing substitution when you discuss your model.

              You could move toward the Bayer et al model by adding a distribution of spatial preferences to your supply model. People could have preferences for pure location, but also for the endogenous amenities of the neighborhood (as in, are my neighbors rich?) There is a large literature on discrete-choice models of imperfect substitution, pioneered by Dan McFadden, also at Berkeley. Your model should address that work.

              Given preferences, there would then be a complicated computation of an equilibrium (maybe not unique) where supply and demand balances in each neighborhood. This is the equilibrium condition that you do not impose. With preferences for endogenous amenities, you could get some Rowe-like effects of complementarities. Over certain ranges, for certain more extreme assumptions, you could generate the kind of “strategic super-complementarities” than Rowe uses (with “grain of salt”) to generate regions of upward sloping demand.

              To move further toward the Bayer model, you would then fit that model to data, allowing for latent unobserved neighborhood characteristics that are correlated (in equilibrium) with both prices and preferences. One could then model the effect of the appearance of the tech industry in SF and ask whether restricting housing supply in the face of that massive demand shock will increase or decrease prices.
              By the way, I have talked to the authors of papers who take that broad approach (although to my knowledge none run that exact counterfactual, partly because it seems to obvious to bother) and they are all on the side of the YIMBYs.

              In any case, the YIMBYs themselves are clearly guided by [1] the standard model of downward sloping demand and [2] the expert consensus. You ought to retract your accusation that this makes them, of necessity, morally evil (which is a separate question from the issue of whether some hypothetical research agenda could someday prove them wrong under some circumstance.)

              • I agree with you Steven that being more mathematical is helpful, because I think a LOT of people are misunderstanding what Phils point is, and just putting words in his mouth. The fact that you misread my name doesn’t help me think that people here are being very careful in their reading and understanding but it’s a common enough mistake so I’m letting it slide ;-)

                In any case, suppose you simply have a distribution F(r,t=0) for the rents in SF right now. Now you add a lump of additional vacant houses to this distribution out in the right tail because they’re luxury apartments, and you observe the distribution later F(r,t=1) where t=1 is long enough to get at least partial equilibriation.

                Now, to the extent that the new available apartments cause people to move out of an SF apartment and into the new fancy digs… the apartment they left will come on the market at higher rent than it was at before *precisely because rent control was holding it artificially down*

                So, suppose we just add one house in the right tail, and someone somewhere in the middle of F(r) leaves their apt and goes into this one house… Now their old apartment increases to market rent and moves right in the distribution, someone moves out of another apartment and into the freed one, again freeing an apartment which moves right in the distribution because it was artificially low due to rent control…. etc etc.

                The net effect is that *one* (or a small number) of new apartments at the right end of the distro can potentially cause a cascade in which N “old” apartments reset to market rate, and eventually one new person from outside SF moves into some other apartment, again at reset-to-market rates.

                Now, reset-to-market rates require *a lot of income* because even tiny one beds go for $2800 or whatnot (I quote craigslist elsewhere) and so after equilibriation we have say one new rich person (a person who can afford 1 bed for $2800/mo *is* rich in my book) in SF and several apartments whose rents increased, thereby driving the whole F(r) to the right (even if some people moved out of more expensive apartments into newly freed slightly cheaper ones, each freed apartment moved right relative to where it was before).

                Now, an alternative is that after a while this doesn’t happen anymore, people can’t afford to leave their rent controlled apartments, and so we have the “liquid” apartments which are driven WAY to the right in a lump, and we have “all the lifer-tenants” who are in a lump on the left never going to leave their apartment until they die.

                And that’s about the sum of it. In the presence of a large latent demand outside SF and caused by tech boom and soforth, every bit of liquidity supplied to this market induces a shift of the liquid batch of apartments to the right in rent. In general, it also causes more apartments to “stick” because eventually even pretty well off people can’t afford to leave their rent controlled apartment.

                Now, this is a DYNAMIC model *OF THE HOUSING STOCK RENTAL DISTRIBUTION* whereas your “downward sloping equilibrium” stuff is a STATIC model of the LONG TIME distribution of prices of UNOCCUPIED apartments in the absence of a constant long duration increase in demand.

                Your downward sloping equilibrium model may be of interest for the prices of apartments in say 2025 long after the tech bubble forcing function is gone, but it simply DOES NOT address what I think Phils point is, which is that every little bit of liquidity poured in on the right dynamically shifts apartment prices upwards through “unbinding” some people from their rent controlled digs and then allowing more people to come in from outside SF and eventually fill in again an un-bound apartment which resets to market rate.

              • Phil says:

                Thank you again, Steven, for taking this all seriously and for helping me understand things better. I will certainly read Bayer et al., and some other work that commenters have suggested, and I will think about what you have written, and I will try to come up with an explicit demand curve and see what happens.

                I definitely don’t think of YIMBYs as “morally evil”…nor, for that matter, do I think of NIMBYs as “morally evil.”

                I will post a follow-up to this post in a couple of months, after I have done some more reading and have thought about some things.

            • Phil says:

              Thanks for your implicit support, Daniel.

              There are so many comments on this post that I haven’t had time to read all of them with care. I’m planning on doing a follow-up post in a couple of months, when I have had time to do some reading and some thinking.

              Thanks to you and to everyone else who put some effort into thoughtful responses.

          • Matt says:

            “In the scenario I have outlined, there is no righward shift indicating “an increased desire of the rich to live in SF”. In my model the desire of the rich to live in SF is constant, and the number of rich people living in SF is limited by the housing stock. If you build it, they will come….and if you don’t, they won’t.”

            Phil, clearly prices have to be higher in the situation where the housing supply is fixed! As you’ve said, these rich people’s preferences for San Fran has not how could prices not be higher when supply is fixed? Before they build the new housing, these rich people were still competing for the houses, they just got outbid by the people who got the houses. You have this very strange idea that somehow the presence of all these rich people “waiting for houses to be built in SF” aren’t having an impact on the current (before these houses are built) prices. They absolutely are! Honestly, forgetting about the crazy upward sloping demand ideas that can explain what you want, I think what you should look up is just Econ 101 effect of having a quota below equilibrium quantity. Instead of thinking of more houses being built as a shift in supply, think of it as removing the quota, and allowing equilibrium to be reached.

          • tcheasdfjkl says:

            Thank you for laying this out. I think I found the disagreement:

            “In my model the desire of the rich to live in SF is constant, and the number of rich people living in SF is limited by the housing stock. If you build it, they will come….and if you don’t, they won’t.”

            Empirically, we have not really been building it, and yet they (uh, “we”, depending on definitions?) have been coming anyway.

            The thing about being rich is that you can outbid poorer people for *existing* apartments. Newcomers aren’t limited to new housing – there’s also old housing that someone else has moved out of. This is why we have displacement and why landlords of rent-controlled buildings have a huge incentive to find an excuse to evict their longtime tenants.

            I agree with you that it’s quite possible that new nice market-rate units might create some demand, too. I just doubt that they would create demand 1:1 with the units – I think that mostly people moving into the area, if they’re wealthy enough to outbid existing renters, are going to move regardless of whether you build stuff or not. (You say that maybe otherwise they’d move into other parts of the Bay Area – but the Bay Area is an interconnected market, if Oakland becomes more desirable it will then go more expensive and then SF will again become more desirable by comparison.)

            But my broader point was – a belief doesn’t have to make sense to you for people to sincerely hold it. I really hope that based on this thread you have been convinced that people *do* sincerely hold this belief, whatever you think of its validity.

            (Interestingly, in YIMBY circles I sometimes see a parallel argument: either “all this anti-gentrification rhetoric by the NIMBYs is just a smokescreen, they just want to protect their property values” or “they just hate techies so much they want to hurt them regardless of whether it hurts the NIMBYs’ own interests”, which both cash out to “my belief is so obviously correct that the opposition is clearly acting in bad faith”, which is similar to what you’re saying. In that context too, I’ve been pushing back against this narrative and trying to encourage a presumption of good faith. People really can have strong disagreements about facts!)

            • tcheasdfjkl says:

              After posting this comment I saw Sonja Trauss’s comment citing a city study showing that 84% of people moving into new housing already lived in SF, while >90% of people moving into SF move into existing housing. This bolsters my central point that the *primary* effect of building new housing is housing people who would be living in SF anyway (which straightforwardly brings down rent), with only a comparatively small secondary effect of *maybe* inducing some marginal tenants to move to SF.

              In light of this, my third-to-last paragraph above should be a lot less tentative. (Though I am not an economist or in any way an expert, so it’s plenty likely that my analysis is still wrong. But not wrong enough to make me not be a YIMBY, because creating opportunities for more people to live here is just plain valuable on its own. (I do care some non-zero amount about neighborhood character, but I think there’s a LOT of room for building in SF before it really damages neighborhood character.))

  14. Fred says:

    “Why, I wondered, are these people promoting policies that are so bad for them?” This is really patronizing, i.e. “everyone who disagrees with me is an idiot!!” Your main definition of ‘bad’ appears to be something like increasing rents paid on average. Under your proposed mechanism, however, this increase is solely because people will switch from mortgage to in the East Bay to renting in SF. How does this increase rents in existing properties?

    • It doesn’t increase rents on existing properties, but it does increase other prices such as food, transport, entertainment, clothing, whatever. Everything you buy in SF becomes more expensive because most of the people living there are ultimately sucking money out of other people’s 401k and putting it into their own luxury goods. Eventually it’s too expensive and you move away, and THAT does increase the price of your property because you used to be paying rent-controlled prices and as soon as you move out your property goes to market rent, which is btw very very expensive.

      All that works as follows:

      1) Fed prints money, gives to finance industry in exchange for govt bonds they would have paid out on tens of years from now. Something like to the tune of around TWO TO FOUR TRILLION DOLLARS over the last 8 years.

      2) finance industry gives to tech industry at ridiculously low interest rates

      3) Tech industry builds some juice bag squeezing robots = useless work employing hundreds of computer programmers (and hookers, cocaine dealers, foosball manufacturers, you think I’m joking but I’m not)

      4) Juice bag squeezing robot company is sold to BIG TECH COMPANY owned by … you guessed it your 401k

      5) Now Witness the Power of this Fully Operational Battlestation… (ie. Your home town explodes)

      6) Later on, your 401k implodes and you’re left holding the bill while ex-Tech company employees own all the real estate within 40 miles of anywhere desirable thanks to all the useful society improving work they’ve done (NOT).

      Mean time, those who don’t have a money printing machine pay $9 for a coffee and $33 for a sandwich and $1500 for a dental cleaning that would cost $200 anywhere else, but they have to actually WORK for it.

      for SF, Monetary policy is at least as big a problem as rent control / housing policy.

    • Phil says:

      I don’t have a definition of “bad”, for purposes of this post. I merely claim that building more market rate housing in San Francisco will lead to an increase in rents in San Francisco.

      (Of course this is an all-else-equal sort of argument. Perhaps other cities will build so much housing that prices in the whole region will come down; perhaps the economy will slump so that housing in the whole region will come down.)

      Unlike what Daniel says, I _do_ think it’s not just the overall cost of living that goes up, I think rents go up, if more market-rate housing is built in San Francisco. I think that if a lot more rich people move into San Francisco, they create higher demand for goods and services, which means more jobs. The people who will fill those additional jobs need to live somewhere. Most will commute from outlying areas, but some will be willing to spend extra to live in San Francisco, and therefore are in competition with people who already live in San Francisco.

  15. Paul says:

    In general, two apartments near each other are substitutes—someone who would have bought apartment A can instead by apartment B. The (N+1)st person’s willingness to pay is lower than the Nth person’s. Etc. Building an expensive apartment in a region will (1) increase average rents, and (2) suppress the rent of nearby properties. This helps renters and hurts property-owners. No one cares that it raises the average rent. This is the usual first-order economic analysis.

    You point out one compensating factor—if you increase the number of people living in SF, you make more jobs in SF, and so more people will move into SF. You don’t make an effort to compare the effects, so let me take a stab at it.

    By what mechanism does more demand for services lead more people to live in SF? The usual answer is: it becomes more attractive to live in SF, e.g. because there are now more or higher-wage jobs.

    So to an economist your argument sounds like “this policy will increase the benefits of living in SF, so more people will move there, so rents will go up, which will hurt renters in SF.” At this point I feel like the paradox resolves itself. In econ 101 land, I don’t think there is any way that a channel like this can hurt the marginal renter.

    Yes, there are lots of market frictions and reasons that the simple economic story doesn’t apply. But I feel like you have to either engage with the econ 101 story (and give an argument which actually references one of those frictions or subtleties) or cite some evidence or give some argument or *something*.

  16. CC says:

    Your model doesn’t explain Tokyo, where as the YIMBY model does so quite easily.

  17. jim says:

    Hi Phil,

    I’m in Seattle where we have a similar problem and the same type of complaint. But your claim that building more apartments is bad for people in the city results from several mistaken assumptions.

    First, the ratio of WADs (Wealth Apartment Dwellers) to Baristas isn’t a constant. No one has come and work downtown to serve WADs if they don’t want to. They do it because that’s the best opportunity they have. If there were better opportunities in Oakland or wherever, there would be fewer services for WADs downtown, and they’d have to either up the pay for services or move where services are cheaper. Providing rent subsidies to low income workers so WADs can have cheaper coffee is about the worst idea I’ve ever heard. It amounts to a subsidy for the wealthy.

    Second, when new luxury apartments come on the market in the city at higher prices, takers for those new luxury apartments almost certainly include people moving out of less luxurious accommodations elsewhere in the city, which means there will be some relief from new dwellings. And the higher prices in newer accommodations is partly a function of the newer accommodations offering more high end features – in other words, the price is higher but the product is better.

    Third, as long as the tech jobs are concentrated in five or six cities around the country (San Fran, Boston, Austin, Seattle), well, the flow will continue. But when workers get priced out, the companies will be forced to distribute their offices more widely.

    Forth, with all this high priced development bursting across tech cities, those cities should be absolutely rolling in cash. But they aren’t because instead of spending it on economic development that pays dividends back, they throw good money after bad trying to freshen the ocean. It’ll never work.

    Cities should focus on building robust transport systems that enhance people’s ability to move and contribute to the economy instead of providing subsidies that have no multiplier effect and, in the end, just make life easier for wealthy people.

    • Phil says:

      Thanks for your comment.

      I was careful not to suggest anything like subsidies to low-income workers. I am not proposing a solution. I’m not even saying there’s a problem, exactly. What I’m saying is that buildng more market-rate housing in San Francisco will tend to make rents in San Francisco go up, which is why I have always found it perplexing that some people who say they want rents in San Francisco to come down are vocal proponents of more market-rate housing.

  18. Sonja trauss says:

    “But for simplicity let’s just assume that on average each of these new 10,000 units is occupied by a household that earns $100K per year after taxes. Total, the occupants of these apartments have $1,000,000,000 in disposable income. ”

    Phil. Do you think those 100 new people would not have moved to San Francisco, if the new housing hadn’t been built?

    Also, do you think if some $100,000 earner is living in, say, Berkeley, and working in San Francisco, they aren’t already spending disposable income in SF?

    • Phil says:

      I think if you build room for an additional 15,000 rich people in San Francisco, the population of rich people in San Francisco will go up by about 15,000. I think there are rich people currently living outside the city who are on the margin of moving into the city (indeed, I know people like that).

      And sure, people who live in Berkeley and work in San Francisco already spend some money in San Francisco. If they move there they will spend a lot more. And of course there are people who live in Berkeley and work in San Jose, but would prefer to live in San Francisco and work in San Jose. And so on.

      • sonja trauss says:

        Why do you think the population of rich people will go up? In other words, do you have any notion of what percentage of the new buildings are occupied by people both (1) brand new to the city and (2) who would not have moved there, but for the new housing?

        In order for your theory to be right, you have to be sure that the disposable income you’re talking about is actually conditional on the new housing, and not conditional on something else, like new job creation.

        Also, I wouldn’t wave off the amount of disposable income already spent in SF by people who work in SF, but live in the suburbs. People buy coffee in SF during lunch break, pay for parking near work, join a gym near work, drop off their dry cleaning near work, eat lunch or dinner near work, go get a drink after work, go to the a play or movie or concert.

        Your argument is based on the idea that 100% of the disposable income is definitely brand new, but now you’re saying some percentage of these potential new residents probably do spend some percentage of their income in SF already. What are those percentages? Do 90% of potential new residents already spend 50% of their disposable income in SF? Or do 20% already spend 5% of their income in SF? It makes a big difference.

        And as I mentioned before, you have no notion at all of how many of the people in the new housing already live in SF or would have moved here anyway. The SF Controllers office actually did go get the answer to that first question, it’s here, on page 28:

        84% of the people that move into new housing already lived in San Francisco.

        Eighty. Four. Percent. So your 10,000 new households is actually 8,400 households already living and spending money and creating jobs in San Francisco. And the 10,000 new apartments have the effect of emptying (up to) 8,400 existing apartments all over the city, some of them will be rent controlled. (I wrote “up to” because some people might be moving out of their parents’ house, or out of crowded roommate situations and not be creating a new vacancy at all, but instead just improving their own welfare and the happiness of the people they used to live with.) Since existing housing is generally cheaper than new housing, it is not safe to assume, in the least, that the 10,000 new apartments will result in 10,000 new “rich” households. The most you can assume is that the 10,000 new apartments result in 1,600 new rich households.

        But in thinking about these 1,600 new rich households you really have to do some work to show that their move to SF is conditional on the new housing, AND that they weren’t spending any money in SF before the move. As you said, you have friends who work in SF and live in the suburbs, who might be induced to move to SF if the right new apartment house came along. If you know someone like that, how much money do they already spend in SF? You can ask them. Then ask at least 30 more people like them so you can get some figures that might be statistically significant.

        And how many of this remaining 1,600 are people that meet that description? 50? 200? 1,000? All of them?

        The same Controller’s report linked to above, also on page 28, says that 97% of new high income SF residents (and 99% of new SF residents of any income level) move into existing housing. This is pretty obvious, because in any given city not recently destroyed by fire, war or weather almost all of the housing available won’t be new. I mention it here because it makes me think that the housing preferences of rich people are not rigid. They will deign to move into existing housing, if new housing is not available or appealing.

        No one moves from one metro area to another solely because they see that a new building has been built. People move for a reason – they have a job, they got into an academic program, their kid has a baby and they’re retired & want to move close. Once a person decides to move, he sets out to find a place to live. That’s the situation of the last portion of households in your original 10,000.

        For a lot of people who decide to move to SF, actually, they look for a place and then realize they can’t move to SF, because everything they thought they could afford (for instance a 1 bedroom apartment in a modest neighborhood) is already occupied, or rather, the landlord reasonably and correctly anticipates it will be occupied by someone who makes quite a bit more money than our hopeful San Franciscan. This is how high income people destroy otherwise affordable housing. They don’t knock it down, they live in it, and by living in it they prevent a lower income person from living in it.

        But let’s return to our high income household who has decided to live in SF. One of the up to 1,600 households that are totally new to the area. These people are coming to SF no matter what. If the city doesn’t build new housing for them they will move into existing housing, thereby ensuring that no one else can live there. If the city does build new housing, then it can soak up the new high income household, and keep the possibility open that the existing modest housing in modest neighborhoods can continue to be available for people of modest incomes.

        That’s the YIMBY argument. You might notice that median incomes are neither here nor there. One commenter above mentioned Simpson’s paradox as a reason that change in median rent is not a very good measure of whether affordability has improved. Another reason is that the whole point of the median is that 50% of the housing is priced below it. The median can be unchanged, but if the overall amount of housing doubles, so too does the amount of housing priced below the median. That matters. Raw numbers matter. Manhattan alone houses as many households making less than $66,000/ year as SF has households (of any income)! A bigger city will house more low income people because it will house more people. More housing is more housing.

        • Phil says:

          Thanks for your comment.

          84% of the people who move into new housing already live in SF. Excellent.

          But when those people move into new housing, they are moving out of existing housing. That housing does not stay vacant. Who moves in? Well, about 84% of the people who move into that newly vacated housing already live in SF, with the other 16% coming from outside. And what about that 84% of people in SF who moved into THAT housing… what happens to the places they used to rent? Well, 84% of them will be rented by people who already live in SF, but the other 16% will move in from outside. And so on.

          Of course this is one of the main ways cities grow in population. (They also grow from people packing themselves more densely into existing units). The Census Bureau says San Francisco has about 70,000 more residents now than it did in 2010; that’s more than an 8% increase.

          As for the argument that high-income people will move to SF no matter what, sure, some of them will. But there’s always a margin, by which I mean people who are just barely choosing not to move into the city (because it’s just a bit too expensive), and people who are just barely choosing to move in (because it’s just barely worth it to them).

          As for “more housing is more housing”, and being in favor of adding more housing even if it doesn’t drive rents down, let me make sure I understand this point. I think you’re saying that it’s OK if the median rent goes up, as long as the absolute number of apartments that cost below $X also goes up (where $X is some number we consider affordable). That makes sense to me as a goal. I just am extremely skeptical that adding more market-rate housing is a way to achieve it: I think that when you move a lot more income into the city, you increase the demand for workers and thereby increase the demand for housing, and…well, you’ve seen my post.

          But some commenters have said that OK, yes, you increase the demand for housing and thus increase rents, but you also increase the demand for workers and thus increase salaries, and if the salaries go up more than the rents do then low-income workers are better off, rent-wise. And I agree, _if_ salaries go up more than rents to, then low-income workers are better off. I dunno.

          • sonja trauss says:

            I think the biggest mistake in your post is that you got the idea somewhere that the main focus of the YIMBY movement is one particular statistic about one particular statistic: the median of the rent. This is not true.

            The second mistake is that you keep conflating “increase in population”, with “increase in high income population.” Sure, you say above, 84% of the people in the new housing already lived in SF, but _someone_ moved into their now vacated apartment.

            Who do you think that someone is?

            Now that you know how many households are moving into existing housing, you need to go back and revise your assumptions about (1) what their income is and (2) how much of it they consider disposable. Both of the new estimates should be lower. That then lowers the number of new jobs you imagine will be created as a result of the new housing.

            As a demographic and political matter, the people who want to/ expect to move into the apartments vacated due to the creation of the new housing are ….. YIMBYs. (And actually some of us expect to move directly into the new housing, but not that many.) We are a constituency all similarly situated – affected by the same forces, all with at least one shared goal. So we are organizing to effect political changes that will in turn improve our material well being. We’re not stupid.

            I think you should delete your pseudo-sociological trolling towards the end of your essay. It’s based on wrong assumptions on your part. The worst thing about it is that it’s embarrassing for this blog. Another bad thing about it is that it’s not nice.

            If you disagree with the YIMBY goal, or if you think that our political success will cause you, or members of your class to be harmed, then just say that. That’s fine, that it part of civil discourse and ordinary politics.

            The people in the margin you speak of, almost by definition, are people who are already spending money in SF. I addressed how that should affect your estimate of disposable income in my post above.

            • I think this is the most straightforward post I’ve seen here. Half the economists here point out that his model is correct (people will move out of surrounding areas into SF and then free up some slightly less desirable stuff outside SF and reduce their commutes etc and this is the point) and half tell him that he’s completely wrong and that prices inside SF will in fact come down a fact that the other half of the economists have already conceded won’t happen.

              But here we have the real meat

              YIMBY is NOT all about decreasing the average rent, and so there’s no paradox to “explain”, nor is YIMBY about decreasing the average offering price of unoccupied units. YIMBY is about building more luxury apartments so that the ultra rich who make more than 99.5% of the US population will move out of older apartments which are the only things available, thereby freeing up some of the older apartments for the merely “quite rich” who make more dollars than 98% of the US population to move into.

              It certainly is not about making the price of SF housing come down, it’s about increasing the supply of a desirable amenity for the rich (meaning, people who make in the top say 10% of the US income distribution).

              Note I say “the rich” because my comparison is to US average. People in SF may not think of their $150,000 tech job as “rich” because they can barely scrape by on it… But people in Sacramento make $62k and they shouldn’t have much sympathy for this point of view because they’d like to live in a nice apartment in SF also.

              Problem solved! There’s nothing to explain. YIMBYs are political-rent-seeking privilege for their personal welfare, NIMBYs are political-rent-seeking for their personal welfare, we can all go back to discussing something controversial like the meaning of p values ;-)

              • Sonja trauss says:

                People in Sacramento do have quite a bit of sympathy for this view, because people in SF who make $80,000/ yr who are displaced from SF by people who make $130,000 are moving to Sacramento and are thereby displacing people who are making $64,000/ year. There is a YIMBY chapter in Sacramento now.

                That’s the thing about displacement, it ripples all down the chain.

              • My feeling on this is we shouldn’t place artificial barriers to building houses, and we DEFINITELY shouldn’t have rent control. Beyond that, the market will provide houses relatively efficiently. If political activism wants a good target, they should target eliminating rent control.

                I’m not aware of Sac. having rent control, but if it does it wouldn’t surprise me to hear Sac people have issues too.

                Displacement on the other hand, in the absence of the BS of rent control, is not a problem per se, it’s efficient allocation of people into places.

                ALL of this problem stems from the widely known (within economics) and well understood consequences of rent control laws.

  19. dave says:

    I agree with your theory around the YIMBY movement’s underlying motivation to bring down “the enemy.” Just attend any town halls or hearings they attend and watch their behavior. They are belligerent, rude, and inflammatory. Sonja Trauss, a public figurehead of the YIMBY movement in SF, is a self-professed anarchist. Some the events where they raise a ruckus don’t even directly benefit their “cause,” but rather provide them another opportunity to provoke the “haves.”

    • Foster Boondoggle says:

      C’mon, this is pointlessly insulting nonsense. Trauss and her associates have advocated very clearly for infill density in transit corridors. They don’t just show up to be disruptive. (You want disruptive, try listening to all theories of the community speakers at Berkeley city council meeting. That’ll cost you a few synapses.) And SFBARF has pushed for it in places that don’t fit Phil’s theory, like mixed-rate apartments in Lafayette, which has so far blocked a proposed project near the BART station. See here: Note that this example directly contradicts Phil’s premise for the YIMBY goals & motivations.

      • Phil says:

        dave, I’ve seen lots of unyielding, inflammatory NIMBYs too, so if some YIMBYs behave that way perhaps it just evens things out.

        Sonja has made some productive comments on this thread and I hope she continues to do so. I have never seen her in other settings so I can’t comment on what she does there.

  20. Tim R says:

    “Well, in a minor way, the affordable housing community will usually make some noise, but they usually quiet down if the developer guarantees that 10% or so of the units will be ‘affordable’.”

    This is inaccurate, and illustrative of the level of understanding the author possesses around housing production in SF. The BASELINE REQUIREMENT in SF has been 15% inclusionary for 55% AMI earners until 2012 when the city charter was amended to be 12%. Then, last November, voters approved a ballot measure to amend the city charter again with Prop C, making 25% the current minimum across a broader range of AMI earners. There is now trailing legislation (Safai&Tang&Breed/Peskin&Kim) which will likely adjust Prop C downward to ~18%-20% after the Controller’s report stated 25% was not feasible. The implication in this article is that a developer can choose to provide inclusionary units, and at a rate as low as 10%. Further when the affordable community does come after a project, they are typically demanding something on the order of a 40% set-aside. (I.e. Jane Kim with SF Giants, Forest City) Further, the notion that a community will “quiet down” for a developer meeting the minimum is also not true. San Francisco has the highest approval barriers of any US city, largely due to Discretionary Review (DR) and the fact that nothing is truly by-right.

    Further, the notion that new Market Rate housing drives avg rents up is simply not supported by data. In fact, in 2016 San Francisco experienced a record # of new MR unit deliveries. Guess what – asking rents on average went DOWN in 2016, and they have gone down in 2017 as well. Take this housing development L7 for example – . Now why would a luxury apartment need to offer 8 weeks of tenancy for free? Take 2 months off of your annual effective rent – that’s a significant discount, is it not? Market rate concessions for new builds have also increased during the same period, further evidencing that increasing market rate supply does have an impact on rents.

    Now I agree building adequate housing is not the short term solution. It takes time. But it is certainly one piece of the solution, and among the only solutions that does not require public subsidy. If we care about housing our workforce, we should support it.

    • I think there’s a confusion between what Phil is talking about and what you’re talking about. Sure *asking* rents went down, but *average rent across all units* went up.

      Whether average rent across all units is the relevant metric or not is another question, but phil wasn’t talking about marginal rent, he was talking about overall in the city there are more people paying higher rents (I think).

      In case it’s not clear what I’m talking about, to explain this in simple numbers, suppose there are 1000 apartments each renting for $1000 per month, and then 1 available apartment renting for $10,000 with no takers. Now I build one apartment and offer it for $7000/mo and someone takes it, and because of my competition the $10,000 rent drops to $7100 and someone takes that.

      Now you can say “see rents dropped from $10k to $7k” or you can say “see the average amount being paid in rent used to be $1000 and now it’s (1000*1000+7000+7100)/1002 = $1012/mo a 1.2% increase

      • Tim R says:

        I have to clarify.

        Asking rent is the generally accepted way to measure rents across time, as it represents the current spot rate for MR housing. A rent controlled unit with a multi year tenant would not be prudent to consider in the calculation, nor would a BMR unit.

        It’s also worth pointing out that there is some skewing in the numerical example provided above. If you change the inputs to something closer to reality (i.e. not a 10x difference in new product rent vs existing) it actually supports the case I make above. Let’s say average rent is $2,000, new unit A is $4,000, then new unit B comes along and units A and B are both reduced to $3,000. Basically it is your scenario above with different numbers.

        The math becomes:
        ($2,000*1000+$4,000) / (1001 units) = $2001.998 / unit
        ($2,000*1000+$3,000+$3,000) / (1002 units) = $2001.996 / unit

        The second scenario shows a decrease! The assumptions make all the difference in such a simplistic model – it’s too easy to produce the result you are seeking.

        • I am not seeking a result, just pointing out that people are confused about what Phil is saying, and maybe if they were more interested in discombobulating the confusion instead of attacking Phil’s physics background they’d make more headway in understanding his point.

          I agree with you that *economists* already make the leap to “all that matters is the spot price of the empty units” but Phil doesn’t take that position, and so they’re talking past each other because they aren’t using words to mean the same thing! It’s very frustrating to look at because it comes off as utter smugness on the part of economists when in fact they obviously don’t understand what Phil is talking about because Phil isn’t using technical Jargon versions of the terms that economists use.

          • Matt says:

            You’ve got to admit that Phil’s comment about the “puzzle” of rising housing prices and rising population in NY demonstrates a fundamental misunderstanding of supply and demand framework. This isn’t a puzzle; the demand curve shifted and we moved along it.

    • Also I assure you that to a physicist “10% or so” is equal to 15%, 12%, even 20%


    • Steve Sailer says:

      “Well, in a minor way, the affordable housing community will usually make some noise, but they usually quiet down if the developer guarantees that 10% or so of the units will be ‘affordable’.”

      How do they decide who gets the gift of a below-market lease?

      • Kevin C says:

        In my neighbor it has been developers family members and people who work for the developer.
        Right out of college at another location, it was me when I wasn’t making much money.
        Not too long ago, it was a family member who only worked 6months a year, on purpose and traveled south america the rest of the year.

      • Tim R says:

        The City actually administers the BMR lottery process, as opposed to the developer. All qualified earners (those under the AMI requirement) who apply are entered into a lottery and selected at random. You can read more about this (for SF, anyway) on the Mayor’s office of Housing website. The developer is required to market the units in a variety of sources for a given period of time before the lottery.

        I’m not sure which specific cases Kevin C is mentioning below, but those individuals would still have needed to go through the lottery process. There are ongoing audits and the repercussions for fraud would be significant, as the BMR guidelines are a covenant recorded into title against the building.


        • Steve Sailer says:

          I bet the lottery process is made complicated to get a better class of applicant.

          From a political science viewpoint, is the idea to create a finite class of definite winners — people who save tens of thousands of dollars per year on rent because they won the lottery — rather than spread small benefits widely, including to out of city residents?

  21. Boris says:

    So San Francisco has built less housing than needed to accommodate population growth for about 30 years, and now prices are through the roof. But somehow, building more housing would not address this problem? We could increase building heights by just one or two stories, and/or allow inlaw or granny flats, and solve the shortage right away, but for some reason all development is demonized as “Manhattan” scaled, and unacceptable.

    By making all housing illegal except for high rise development in SOMA, that’s all we see. Yes, developers are greedy. But limiting what can be built hands them the keys to the kingdom. It’s both idiotic and shortsighted, not to mention deeply conservative, to resist change so much. SF sucks as a result. Congratulations.

    • Phil says:

      As I said in my post, there is some point at which, if you build enough, rents will come down. But you’d have a very different city.

      If SF “sucks,” it’s funny that so many people want to live there.

      One thing that would decrease rents for sure would be to make San Francisco a much less pleasant place to live. But I don’t expect SF residents to sign onto that plan.

  22. Carlos Ungil says:

    A side remark about your example: I don’t think households with $100K in disposable income can afford the $48-96K typical market rates.

    Your model seems to be that ten thousand high-income households will come from outside San Francisco. I would expect a substantial fraction will be people moving within San Francisco, so there will be an effect on the lower layers of the housing market (but maybe there is something about the structure of the market in San Francisco that prevents that from happening).

    You also seem to ignore the effect on wages (or maybe I didn’t read carefully). If the demand for “more waiters and shop clerks and car mechanics and plumbers” cannot be matched by supply constraints, price will go up. It could happen that even if the average rent goes up, it goes down as a fraction of income.

    • Phil says:

      Thanks for your comment.

      I think if you build 10,000 new units in SF, many of them will be filled by people who already live in SF… but not all of them. And what about the places they moved out of? It’s not like they’ll sit vacant; they will be occupied by new renters, many of whom already live in SF…but not all of them. This creates another set of vacancies, many of which will be filled by people who already live in SF…but not all of them. And so on, in a cascade of effects.

      On the whole, I think that if you build space for 15,000 more people in SF, the population of SF will go up by about 15,000. Do you disagree?

      • Carlos Ungil says:

        I agree. It’s just that you made it sound as if “each of these new 10,000 units is occupied by a household that earns $100K per year after taxes” coming from elsewhere and adding in aggregate $1,000,000,000 to the San Francisco economy.

        • Phil says:

          If you move 15,000 more rich people into San Francisco then you do move a ton of new income into the city. How these people distribute themselves among new vs existing housing doesn’t matter much.

          • Carlos Ungil says:

            Why would the 15000 newcomers be rich? What about the cascade of vacancies you mentioned a couple of messages ago?

            • Phil says:

              If a rich person moves out of an apartment in the city, I think it will be taken by another rich person.

              • sonja trauss says:

                Yeah, this is just “something you think.” You need more than that to be making a good point.

                Why do you think that? You guessed it, because it supports the conclusion you want to come to?

              • sonja trauss: it’s not even a question in SF. I go to craigslist and I see for SF apartments currently:

                1bed/1bath Mission: $2650
                1br/1bath inner richmond: $3195
                1br 627 ft^2 potrero hill: $3399
                1br North Beach: $4300
                1br 625sqft SOMA: $3300
                2br Noe Valley: $4595
                2br Nob hill: $5195

                Note that *before taxes* the national median household income was about $56,000 / yr = $4666/mo

                So, these *rents only* cost around 60-100% of median before tax income in the US, for basically 1 bedroom

              • Carlos Ungil says:

                Daniel, how is the median US income relevant? Median income in the Bay area is substantially higher.

              • Carlos: it’s a metric that anchors the question to something other than the bay area. Suppose everyone in the bay area made a bajillion dollars but only 2 bajillion would get you an apartment… they all are rich relative to the rest of us, and could at any time live for a few days in their car and then retire for life and move into a mansion in Boise or Albuquerque or whatnot.

                Of course they aren’t making a bajillion dollars. But if we make their issue entirely scale free by considering only the local salaries, we lose the point of money which is to help measure desirability of stuff relative to alternatives. There ARE alternatives to living and working in SF no matter how much the Tech Bubble might like you to believe otherwise.

              • Carlos Ungil says:

                Daniel, I think the question is about housing in San Francisco (pop: 800k) in the context of the Bay Area (pop: 8mn) and how affordable it is to people living in the area. It will clearly be less affordable to people from Tennessee and even less so to people from Kazakhstan.

              • And yet, there is no political barrier to moving to Tennessee, we are a united states and people are free to move anywhere within them for economic opportunity, and so artificially isolating the SF Bay gives an artificial bias towards the idea that all these people in the SF bay *should* be there and the problem is to just make it affordable for them. This is exactly the thinking that *leads to rent control* which as we already agreed elsewhere was a terrible idea.

                Living in SF bay is a choice, and the high prices in the area are a signal that you should not live in the bay area unless you can make a crap-ton more money than you can in Tennessee. We need to be reminded that maybe one alternative is to simply not move to the SF area. In the absence of rent control distorting the market, that would be the single end of story take-home message, don’t come there unless you can get a good job that lets you afford high rents. That’s the PURPOSE of prices in a market, to provide information about supply and demand.

                But we DO have rent control, and so we have terrible market distortions. So the question becomes should you move out of SF or fight politically? I think the answer is *fight rent control* not *massacre the character of the city by doubling the housing density* nor *give some lucky few subsidized housing*. Nor will minor increases in housing (say 10% total) make a major difference in affordability of SF living given the other market distortions going on (tech boom for example). Jamming a few extra units down the throats of the city planning commission is just a subsidy for the bottom slice of the top layer of american incomes.

                It’s important to consider things in dimensionless form because dollars are not dollars everywhere and every time. So one dimensionless ratio is cost of housing in SF as a fraction of median country-wide household income. This tells you, yes it’s really expensive to move to SF, you need to make a lot more money. Just living in SF is a privilege that not everyone has, and it’s an amenity you need to pay for.

                Next question is (After Tax Income-Housing)/LocalCostOfGoods at each location, this measure of disposable income tells you how much stuff you can buy with the resources you have left over to spend on things other than housing.

                If a person leaves SF and gives up their “great” job and moves to Tennessee very possibly they will have larger disposable income on this measure, and a better quality of life as measured by that. If they *truly* don’t care about the environment of SF, then if that’s true, they’re doing it wrong, they should go to Tennessee (or wherever, assuming such a place exists). On the other hand, if they aren’t even considering this fact and just politically trying to muscle their way into the “haves” who get to sip cappuccino at the Cafe while watching the sailboats on the bay… then perhaps we shouldn’t be very sympathetic to this. Under that idea, forcing a few extra units is basically rich people who aren’t quite rich enough for their tastes buying political favors — rent seeking

                The fact is living in the SF Bay, and particularly SF city, is something extra than just living in a place with a good job that lets you have a fair amount of disposable income. Isolating the problem to the SF Bay artificially biases your analysis.

              • Carlos Ungil says:

                > unless you can make a crap-ton more money than you can in Tennessee.

                You can, that’s why the median US income is irrelevant.

              • “You can, that’s why the median US income is irrelevant.”

                It’s not irrelevant, it’s a component of the metric by which you need to evaluate whether you should move out of SF.

                Not EVERYONE can make more *real income* in SF than if they moved somewhere else (wait-staff, plumbers, roofers, copy-shop clerks?). But, lots of people have subsidized rent through rent control, and so they don’t need to move somewhere else, because they’re getting a free ride.

                In the presence of a large set of people waiting in the wings with one of the big money maker jobs for just a marginally cheaper apartment, adding a bit of market rate housing causes more rich people to move from Oakland or wherever into SF. Phil is NOT wrong about that. Adding subsidized housing just benefits the rich people living in SF by reducing the cost of paying Kinkos shop workers and baristas and things. Adding a LOT of housing decreases the rents, but at least much of the reason is that it would dramatically decrease the amenity that everyone is paying so much for and the demand to live there would potentially decrease.

                To the Kinkos workers and the baristas the median income in the US is very very relevant.

  23. A says:

    Interestingly, bay area rents have been generally declining over the last few months, which some realtors credited to increases in housing units.

    Your reasoning seems obvious to you, but I found it to be ambiguous. Are you anticipating rising prices in response to increased supply due to weighting shifts from older and rent-controlled structures to new market buildings? In that case, YIMBYs might claim victory in lower accessible rents now available to the marginal renter, even as composition effects increase median rents. Or, are you claiming that new renters would have paid higher rents, but didn’t, or couldn’t, and thus didn’t drive prices higher until new units came online? Or, are you stating that tens of thousands of prospective renters are offering bids tightly clustered at just below market rents. Thus new units price at effectively prevailing prices, which creates demand for lower income workers, whose lower incomes translate to higher rents via second order effects?

  24. Some things people should know as background:

    SF, (and Manhattan), both have extensive rent control. There is a lot of housing “available” (in the sense of not taken up by bodies) which isn’t actually available because it’s occupied by people who moved in in say 1995. These people are spending $900 a month for an apartment or small house that would rent today for $3800/mo These people will never in their lifetime move out. I know of several personally, so I’m quite sure this is true.

    Furthermore, there is an extensive black market for housing, people with one of the rent controlled units essentially re-rent rooms at near market rates, thereby making money (they pay $900 for a 3 bed apartment, rent out two rooms at $1500 ea, substantially below market rents, and net $2100). It’s technically illegal but it’s not easy to police.

    In this environment, people who own buildings are naturally inclined to do something about the situation, and they do. There’s a law called the Ellis Act which allows them to get out of the rental market in CA through a process where the whole building becomes something other than a rental, and tenants are evicted.

    Here’s an advocacy site that maps the process (infovis! it’s very advocacy oriented)

    In this environment, the market for housing is heavily distorted, and the services in a region are heavily reliant on the historical rent-controlled people. If there aren’t enough of them to provide plumbing and restaurant wait staff and book store clerks and soforth… then the place winds up against a wall.

    SF and Manhattan both have severe issues regarding commute times. in SF you have to cross a bridge or take a train through an under-bay tube, to get there from any of the “cheaper” places to live (whether it’s Oakland ~ 5mi, or Bay Point ~ 40 mi, or Sacramento ~ 100mi). Also because of the topography there are “arteries” along which you need to travel along valleys, and these tend to clog up. I know of people (friends of friends) who drive for Uber who sleep in their cars in SF so they don’t do an 8 hour a day commute (4 hrs in morning from Sacramento, and 4hrs in evening back).

    A big part of the issue is that unlike say the LA area, there is no nearby lower cost housing markets where people could come from to do service jobs in SF like waiting tables or being a plumber or a nanny or whatever. Even areas like Oakland, Berkeley, Emeryville, El Cerrito, all have had rental spikes, and many of them have similar issues regarding rent-control and associated distortions in housing market.

    As Lindbeck said, only bombing does a quicker job than rent control of destroying cities:

    I’d add to that list of bombing and rent control an additional one: free money printed by the Fed and given to the finance industry who prop up companies that make $400 juice bag squeezing machines

  25. Jorge Guzman says:

    For a blog that prides itself in careful statistical thinking this is a surprisingly sloppy theory on how more housing increases median prices. Obviously adding housing does not need to change the sign of the trend from positive to negative to have an effect, as long as it changes the slope—i.e. even if Manhattan real estate looks more expensive after adding housing, if would have been *even more* expensive if not added. The theory is a snub at a well grounded economic theory—supply and demand—and only replaces it with some handwaving on general equilibrium issues.

    We could apply the same type of thinking to other issues, and it’s easy to see that it can’t be right. For example, We keep adding hospitals and healthcare costs keep going up, for crying out loud, why do you think we should add more? We keep adding schools and costs of of education keep rising, for crying out loud, why should we add more? Most people would say that hospital costs are simply raising too fast, but that the world without adding those hospitals would be even worse than after adding them for people that need care.

    • NatashaRostova says:

      I had to reread it to make sure the argument was saying what I thought it was… Some sort of clever ‘counter-intuitive’ thought experiment on why more housing will increase prices?


      >Serving those extra 10,000 high-income households will require tens of thousands more waiters and shop clerks and car mechanics and plumbers etc etc etc….that is, there will be more jobs for the kinds of people who already have trouble affording a place in San Francisco. Those additional people will need to live somewhere, so there will be increased competition at the lower end of the market, which means higher rents. Most of these people will end up commuting from other cities.

      This is, of course, an empirical question. Will the rich folks moving into more expensive new housing make more room for poorer people to take advantage of more supply, thus pushing down their rents? Or will the increased demand for their services push up their rents? You can’t intuit that from a one paragraph thought experiment. The strong economic priors I would believe (and it seems a few other commenters also believe), would be that increased supply will generally overwhelm any second order effects. Again — if we want to be hyper rigorous it’s an empirical question — but it’s still the strongest expectation based on economic thought and research.

      • Phil says:

        Natasha, thanks for your comment.

        I haven’t exactly scoured the literature but I’ve read what I’ve found and what people have pointed me to, and…well, this is definitely an area that could use a proverbial one-handed economist. When it comes to empirical evidence, people interpret it different ways.

        One thing that seems clear is that in metropolitan areas with high housing prices, restricting the supply of new housing increases the cost of it. No argument here. But when it comes to the spatial distribution of housing prices within a single metropolitan area — by which I mean an area of a scale that people are willing to do a daily commute — I haven’t found anything that even claims to give a definitive answer.

        One thing I keep pointing out, but nobody responds to, is that Manhattan has 1.6 million people on it (and has the apartments to support that many people) yet rents are still very high there. Of course there’s literature on this, and some people say that prices there would be lower if more apartments were built…but where’s the empirical evidence for that claim? It’s hard to build new apartments there, but they do get built — the population is up more than 100,000 since 2010 — but rents have gone up, not down. Empirically, building more apartments has not caused rents there to go down. So, OK, the first 16 tranches of 100,000 didn’t cause rents to go down, but the next apartment that gets built will finally trigger the decline?

        There is a weaker claim that I think is what Steven Berry is saying in this comment thread: that building more apartments in a popular city will not make rents go down, but it will make them go up less than they otherwise would. If this is the claim the YIMBYs and BARFs are making then I think the news coverage I have seen about these movements is not accurately reflecting them: the coverage I have seen suggests that these groups believe the increase in rents will halt or reverse if a lot more housing is built.

        • Steven Berry says:

          By the way, I am making the weaker claim that you suggest: in the face of increasing demand, allowing new construction will result in prices *rising less than they would if construction is prevented.* That is the YIMBY argument and it involves a counterfactual policy. This is the implication of the supply and demand model. Really, you should learn it. Really, before you start writing about it, you should learn it.

          In this S&D context, the counterfactual question isn’t answered by saying in the “in the face of rapidly increasing demand, NYC had a small % increase in apartments, and yet prices still rose.” A better set of anecdotes involves Dallas and other southern cities. In the face of rapidly rising demand, they place fewer restrictions on supply and prices go up only a little, relative to NYC or SF. Another set of anecdotes involves NYC and SF before the zoning restrictions of the 1960s and the increased demand of the 1980s to the present. In the 19th and early 20th century, mass waves of poor immigrants were housed at low cost because there were few housing restrictions. When they wanted to move in, housing was built and prices stayed relatively low. 19th century NIMBYs hated it, of course. All those swarthy dirty people, you know. Since tight supply restrictions were put into effect (followed later by big demand increases), the poor have been continuously driven out of both cities and now they gleam in their purely upper-class splendor. NIMBYs are very happy. Another set of anecdotes involves neighborhoods in NYC with relatively lax zoning restrictions (Yorkville) and relatively tight ones (Greenwich Village). Prices have risen less in the neighborhoods where some new construction was allowed, while gentrification is complete in GV.

          By your argument, all of these anecdotes should be reversed. The massive construction wave in Dallas should be driving prices up and up, whereas the supply restrictions in SF should be holding prices down (that is the implication of the “strategic super-complementary” that you so like.) By your argument, *no rich people live in Greenwich Village, because no new housing was built.* If you want more than anecdotes, learn econometrics and follow the research literature.

          The reason folks bring up Tokyo is that they allow neighborhoods to densify with little restriction. Unlike SF, the middle class can still afford to live in reasonably close in neighborhoods. Prices probably do have to rise a bit to cover the cost of the new denser construction (that just says that supply slopes up, supporting the standard S&D model). In the FT: “Why Tokyo is the land of rising home construction but not prices.”

          I think you are closer to the NIMBY truth when you say that in the face of a massive demand increase, to bring prices down *a lot* you would have to build *a lot* of housing (Tokyo style) and that would change the city in some way that you think you wouldn’t like (although many other folks love the many dense cities around the world.) That is, you are willing to drive all the poor black folks out of SF and tolerate sky-high prices in order to preserve it exactly the way that you like it. Hold the existing housing stock fixed, and simply sell it to rich folks for fancy renovation, like in Greenwich Village and SOHO. Fine, but own up to it. You are arguing for higher prices to support your comfortable upper-class esthetic ideal, at great cost to folks who could otherwise move to SF and take advantage of high wage jobs to lift themselves into the middle class. Instead we leave them to despair and overdose in unproductive places, and wonder how we got Trump. But SF looks really really nice, I will give you that.

          • Phil says:

            Steven, you say “By the way, I am making the weaker claim that you suggest: in the face of increasing demand, allowing new construction will result in prices *rising less than they would if construction is prevented.* ” That is, I think, almost exactly the claim I attributed to you in the comment you’re replying to! I said “There is a weaker claim that I think is what Steven Berry is saying in this comment thread: that building more apartments in a popular city will not make rents go down, but it will make them go up less than they otherwise would.”

            • Carlos Ungil says:

              Phil: There is a weaker claim that I think is what Steven Berry is saying.

              Steven: By the way, I am making the weaker claim that you suggest.

              Phil: That is, I think, almost exactly the claim I attributed to you in the comment you’re replying to!

              I don’t think there is any disagreement over that.

        • Matt says:

          This is epic! The claim about NY is incredible… your just comparing to equilibrium points and saying, oh! I guess the demand curve is upward sloping, because supply shifted (i.e. we observed housing increase, which by your logic apparently must be a supply shift). The point with NY is that demand is shifting!! We’re just moving along supply curve, which clearly will be upward sloping.

  26. John Hall says:

    I found this to be a very unsatisfying article. First of all, your argument is really focused on the motivations of the YIMBYs, which are probably very diverse, rather than their actual arguments. The actual argument is based on Econ 101. Cities like San Francisco adopt regulations that artificially reduce housing supply. As a result, quantity supplied does not grow as fast as it would have otherwise and prices increase. The city then adopts policies to reduce the price of homes, which have the effect of shifting demand right – though maybe not completely offsetting the impact of the previous step. Nevertheless, it propels prices even higher. The YIMBYs argue that remove the market distortions will lower prices and more supply.

    Your whole argument is that median rent prices will increase as a result of new housing. That may be a short-term effect, but I highly doubt it will be a long-term effect. Basic Econ 101 is that shifting supply right raises quantity and reduces prices. While housing has some special issues, they are not sufficient to overturn this in the long-run.

    For your point about motivations, I highly doubt that Republicans in Texas and Alabama care too much about housing policy in San Francisco. They don’t even need to bother supporting YIMBYs in San Francisco since Texas has basically already adopted their policies. I highly doubt they want to HURT San Francisco. They think their policies will make San Francisco more affordable. The YIMBYs are people living in San Francisco who are trying to make it better. Maybe they were originally from somewhere else, but they recognize how distorted the market is and believe there is a better solution. Perhaps you should not impugn their motives.

    That beings said, I think you could make a convincing case that YIMBYs don’t care enough about people who are being priced out of the market and if they make compromises, they make some headway

    • The real problem in SF and surrounding areas is two-fold:

      1) Long Standing Rent control makes it a terrible idea to invest your money in running rental housing businesses and so there are not enough rental houses being supplied (add on to this things like zoning codes etc that make it difficult to build housing)

      2) Huge quantities of cash are being dumped in SF and surrounds by the finance industry through venture capital seeking to turn free money from quantitative easing policies by the Fed into ownership of the next Google or Facebook

      The net result is that what little tiny bit of housing turnover there is is sucked up by people who are in essence printing money (and I mean that quite literally, the money is appearing out of thin air as the Fed buys back bonds hand over fist, the finance industry loans at essentially zero rates to VC types who then sink millions into bullshit schemes involving internet connected fruit juicers and apps that make it easier to acquire sexually transmitted diseases or whatever)

      The thing is, in this combination, it seems unlikely that building more housing will help. As much new housing as comes online will be soaked up by people with a printing press. If they leave a less desirable apartment, that apartment resets to market rent and will also be snapped up by people with printing presses.

      Long term, after the printing press crashes, housing prices may decrease, but the policy of rent control will ensure that as soon as prices crash sufficiently, apartments will be soaked up by lifetime renters as rent control policies actually require real rents to DECLINE through time (in Berkeley for example landlords are allowed to raise the rent 1/2 the increase of the CPI each year, meaning real rent declines). Once you’ve filled the crashed housing with lifer-tenants you’ll be back where you were: no housing for anyone who isn’t one of the privileged few who have lived there forever and so have virtually “free” rent.

      As long as you create stupid distortions of the market through rent control AND terrible monetary policy, you will have truly terrible outcomes.

      • John Hall says:

        I mention rent control as policies that support demand.

        To your monetary policy point, it should shift the demand curve comparably by region. Prices only increase by more in SF because supply is more inelastic than in areas with fewer restrictions on building.

        • Supply of money is not uniform across the country either though. SF is driving prices up because SF is where the pits are that the Fed is dumping the money in, also Seattle, and to a lesser extent maybe a few others like Boston or Austin TX or Atlanta or whatever.

      • Foster Boondoggle says:

        Daniel – Your theory of monetary policy is rather peculiar, though definitely shared by some goldbugs out there. Fed policy is driven by two (competing) mandates: low unemployment and low inflation. Since the financial crisis, broad inflation has been running between small negative values (in the immediate aftermath) and an upper range of 1-2% (recently). These are historically extremely low levels, not seen since the 1950s. Meanwhile, unemployment has only come down to “tolerable” levels in the last few years – causing the Fed to start raising rates off the zero bound. You claim that a primary effect of the low short term rate has been to inflate a new tech bubble, similar to that of the late ’90s. This is certainly a possible side effect, though you have to acknowledge that there’s no necessary relationship between the two — the late ’90s bubble took off at a time of much higher Fed policy rates. But the intention of low policy rates is to goose economic activity *across the board*, and there’s certainly no guarantee that everyone will make “sensible” investment decisions. In the absence of expansionary fiscal policy out of Congress, do you have another idea for how to promote broad-based economic growth? Or, in order to avoid some bad investments, would it have been better to let the deflationary spiral of 2009-11 continue by keeping rates well above the CPI inflation rate? Incidentally, most of that flood of cash pushing up rents is coming from the incumbents — Google, Facebook, Apple, Twitter, etc. A lot of the people working for these companies are recent college graduates. Millennials making enough money to move out of their parents’ house doesn’t seem like something to complain about these days, even if it does turn I80 and 101 into parking lots at commute hours.

        Back on the original theme: to someone who first moved to the Bay Area in the early ’80s, these complaints sound awfully familiar. Back then it was gray-hairs complaining about the DINKs (dual-income, no kids) driving up rents, and the “Manhattanization” of SF. Now those DINKs are the gray-hairs complaining about how new construction will destroy their neighborhoods. Admittedly, it’s hard not to have *some* sympathy for people who put down roots in a neighborhood because of its character (and affordability) only to find that they can no longer park in front of their house and that their once sunny back yard is now fully shaded by the new apartment building on San Pablo Ave. Variants of Phil’s theory seem to have driven recent Berkeley politics, leading to the peculiar alliance between a left that opposes all market-rate development (e.g., Berkeley’s new mayor), and a right that opposes all development full stop because the increased congestion will lengthen the drive from their N. Berkeley home to their health club at the Claremont (secondhand quote from my new council-person).

        • Foster Boondoggle says:

          Also, the Fed is not “dumping money” into SF any more than anywhere else, certainly not by QE, which is mainly about pushing down long-term yields by buying bonds from the Treasury. The Fed has also been buying conforming mortgage loans, of which very few come out of high-priced areas like NYC, SF or Seattle – most mortgages in those areas are above the conforming limit. So to the extent that there’s a regional bias in QE, it’s *against* those high-cost areas. QE pushed down the “market price of risk” for long term investments. That applies to *all* long-term investments, whether its being done by Solar City or Google, or by some dumbasses with a $700 juice machine.

          Or do you feel that the Fed’s setting of rates has somehow changed the risk-return equation so that the comparison between an investment in a risky startup vs. an established firm has become more favorable to the dumbasses? How does that work exactly?

          • It’s more like insider trading / gaming the system than “investment”. You create a startup selling juice bag squeezers, you sell this to an incumbent tech firm and retire. The incumbent tech firm is owned by… everyday people’s 401k and soforth but the people who make the decisions about which worthless tech startups to buy are not the individual investors who learned their lesson in the 90’s they’re… high level execs at incumbent firms who are friends of the guys who work at the startup, or VCs who are friends of the startup guy and who have clout at the incumbent firms.

            The incumbent firms buy the startups “to get the talent” and then disassemble them and shuffle employees around, and eventually implode… vis Yahoo.

            some of the incumbents are stable enough with a core business that they can just weather the storm, Google probably is like this. But why do you think they have “Alphabet?” the reason is… they bought up a billion stupid startups and needed a structure that separated their core business from all the crazy crap they’re doing just as a way to turn free Fed money into assets for friends.

            Here is net commercial lending divided by CPI


            Have you heard of the term “pilot induced oscillations?”

            • Even better:

              which adds stock market capitalization as a fraction of GDP rescaled to the same range so you can see the time correlations, and mortgage lending.

              Money from the Fed is pouring into the stock market through the mechanism I discuss. The current stock market levels are at the same height as the peak of the late 90’s crash as measured by GDP, essentially all of this is driven by massive uprun of commercial lending at cheap rates to finance insiders, tech companies, etc.

              Mortgage lending ran up after the first tech crash in 2000 but held constant or declined after 2010 while from 2010 onward commercial lending surged dramatically.

              • A says:

                An alternate explanation is that lending volumes are marginal transmission mechanisms, but useful indicators of market sentiment. Hence, you see stock margin debt credited with causality, when it moves simultaneously with stocks, because both indicators reflect some deeper changes in expectations and required compensations.

              • A long comparison of different models is of course in order, but for now in this time and place I’ll just state that I think my model is accurate and explains a lot of the observed facts about our economy, including the fact that since 2000 or so GDP growth has been abysmal while enormous quantities of bullshit have occurred all related to finance and nonproductive tech activities: mortgage backed security bullshit, whole ghost towns of 4600 sqft luxury houses in CA central valley ca 2008, internet of things connected juice squeezing machines, WhatsApp sells to Facebook for $22 Billion, App startup “Yo” raises millions of dollars for an app that says “yo” to your friends, startup “Tilt” flames out in a blaze of hookers and blow, Uber loses $2+ Billion dollars last year and on track to lose even more this year… Most Uber drivers make near nil to negative money after the cost of car maintenance, etc etc)

                GDP per working age population continuously compounded rate of change: since 2000 or so it’s been less than 2.5% time-averaged, and yet we have the greatest most amazing computing technology and outsourced cheap labor and ….


                And let’s not forget labor force participation rate at its lowest since 1978


                And white suicide rates up continuously since 2000 approximately doubling for both males and females

                I think its safe to say we seem to be doing it wrong.

        • Foster:

          The bubble of the late 90’s was driven by people spending their own investment money (ie irrational exuberance at the *retail* level, or at least including both investment banks and retail level), ever since that crash we’ve had bubbles driven by people spending Fed money: vis: the mortgage crisis, and now the next tech crisis.

          Alternatives promoting broad economic growth ?? Yes absolutely I have a better idea, instead of printing money and giving it to banks, print money and hand it out to every single person in the country in equal quantities as a Universal Basic Income (basically, deposit it into their individual demand deposits accounts at banks). Then, when it’s spent, it drives investment in the production of goods that are desired by most people, instead of the production of goods that are desired by investment banks (namely shitty tech startups).

          Yes, the Fed *is* dumping money into SF through QE because the flow of money is: FED -> Banks, and Investment Firms -> VC -> Tech Firms -> Tech Employees -> Real Estate and expensive services in Tech Regions -> trickle out to the outlying regions

          Essentially none of the money from the Fed is going to what you’d call “everyday” people and so you aren’t seeing broad across the board CPI type inflation, but you ARE seeing SPIKING inflation directly adjacent to Tech firms.

          One of the major ways that money “comes from” incumbents such as FB and Google is that these companies have been buying up startups left and right for years. They are, in essence, acting as big tech hedge funds.

          • Foster Boondoggle says:

            I think your polemics have gotten ahead of the facts here. First of all, I don’t understand why you think, given that (according to you) retail investors were responsible for the 1990s bubble, those same retail investors would do any better than the “tech hedge funds” if there were a helicopter drop of money in place of the Fed’s QE.

            Secondly, a significant portion of QE went into lowering mortgage rates: the Fed bought (pooled) mortgage loans to drive down the rates, resulting in a large amount of refinancing into loans with rates not seen since the 1950s or before. The difference between the old and new rates is money directly into the pockets of the middle class, though not with the “UBI” type distribution that you might prefer. But it’s clearly *not* just “FED -> Banks” as your theory has it. It’s exactly money to “everyday” people.

            You seem to have a larger economic theory that says that middle class individuals will spend money wisely, while rich individuals will invest money stupidly. While I’m far from a fan of the average American rich person, I fail to see on what basis (other than “the tech bubble produces some stupid companies”, or maybe just general resentment) you maintain this theory. Particularly when I look at Google, I’m rather astonished at how smart the people who’ve built that company’s products seem to be, and how much value they’ve added to our everyday lives.

            • Foster Boondoggle says:

              Also, regarding “pilot induced oscillations”, you’re apparently claiming that the Fed keeps overshooting, driving the economy to boom and bust. Given how slowly policy has moved since the financial crisis (while inflation has been near zero), this seems like a peculiar idea. But your claim is also about something that it’s explicitly *not* Fed policy to target, namely the level of the stock market. The Fed’s job is to keep inflation moderate and unemployment low. Show me how the Yellen Fed has failed at that project (within the limits of what they could do). To the contrary, it seems very clear that – in the absence of help from better fiscal policy – they’ve done an admirable job.

              • Pilot induced oscillations are caused by a forcing function that is out of phase with the thing it’s trying to control, this primarily occurs by being *too slow* to respond or having a *long delay* between observation and the response or just not caring about that measure and letting it do whatever it does when you apply forces to control something else. The Fed has been trying to control Unemployment and CPI and has managed to do those while trashing the economy in many other dimensions. I’m not blaming the Fed, I blame the idea that politically speaking we should task the Fed with “control unemployment and CPI by controlling the flow of money into banking institutions” while ignoring everything else.

            • First of all, yes I am angered by what I believe is a serious problem in our economy, but at the same time, I am actually quite open to being proven wrong. In fact, I’d love to be proven wrong, because obviously the situation as I’ve laid it out is pretty lousy. Nevertheless, at the moment I think it’s true:

              Foster: “First of all, I don’t understand why you think, given that (according to you) retail investors were responsible for the 1990s bubble, those same retail investors would do any better than the “tech hedge funds” if there were a helicopter drop of money in place of the Fed’s QE. ”

              hedge funds don’t consume anything, their purpose is entirely to turn money into more money. Because they don’t consume anything essentially no information flows between hedge funds and the market about what items for consumption should be produced. If the helicopter drop is straight into the hands of Goldman Sachs etc (and according to me it is) then what the market will produce is exactly what GS et al demand, which is funny money in the form of stock certificates for closely held startups to be sold to incumbent tech hedge funds whose company is owned by hundreds of millions of peoples IRAs and 401ks.

              Why is it bad for a small set of people (let’s call it 1000 high level finance and tech execs) to have concentrated power in deciding how to allocate resources? For the same reason it’s bad to have a communist politburo, they are information poor about the needs of 320M Americans. Your assumption seems to be that the helicopter drop of money ought to go directly into investments. But the truth is that neither of us knows where it ought to go. If you’re going to manufacture money, it seems terribly unfair and information-theoretically stupid to simply helicopter drop it into the hands of a few thousand people who created the crisis that required the drop in the first place (finance industry). The truth is I’d personally like a new roof on my house and a new air conditioning system. Someone down the street has elderly parents who need assisted living, a family with twins I know needs day care… how much of that is GS/Wells Fargo/BofA demanding with their shiny new electronic deposits at the Fed?

              The point of a UBI helicopter drop is to drive broad consumption of goods needed and desired by everyone not an elite few. It filters information into the economy globally while at the same time avoiding the terrible feedback loop where money disappears and so holding money and doing nothing becomes a way to get rich. That’s the real justifiable purpose of inflating the currency, avoiding that feedback loop of deflation.

              Regarding Fed -> Banks transfer. This is by definition what the Fed does when it buys treasury bonds, only Banks and investment firms with direct accounts at the Fed receive the largess of the Fed inventing a new parcel of money in an electronic account out of thin air. It’s only after they’ve sold the bond to the Fed that they then have a question of who next should receive it through lending?

              What do the Banks and finance companies do with it? Sure there was a period of refinancing for a year or so. But looking directly at the graphs I showed you, you can see that total outstanding mortgages declined 10 or so percent over the last 9 years, and commercial lending increased to something like 160% of its value in 2010. Rates for loans were already quite low in 2008 before the crash, that was what was driving all that shitty mortgage consumption, so refinancing of loans by people who were solvent and able to meet the lending requirements in 2009 and 2010 I wouldn’t expect to make a major contribution to cash flow, though I could be proven wrong with data. Do you have data on the decline in average mortgage payment + rent by an american family (of course, some people made their mortage go away and now are renting… at high rents!)?

              I actually have the completely microdata of the American Community Survey, and I am working on analyzing it for this purpose.

              I definitely doubt it was the broad improvement you seem to think it was, not like taking the 2 Trillion dollars manufactured in the M1, and helicopter dropping it into each and every citizen’s bank account. Incidentally, that would have been something like $100/mo per person, so a family of say 2 adults and 2 kids would have gotten something like $400/mo for 5 years between 2010 and 2017. Did their mortgage payment or rent level decline by $400/mo? Did their after tax income increase by $400/mo? I doubt it but I don’t yet have the numbers to show you.

              Regarding Pilot Induced Oscillations. Sure, the Fed is blind to lots of stuff, this doesn’t mean they SHOULD be blind to it.

              “inflation near zero” is not all that interesting. Economists seem to think that “inflation” is *necessarily defined* by the CPI. I don’t take that view. I think the CPI measures one important dimensionless ratio in the economy which is relevant to how hard it is to buy a consumer item with a fixed number of dollars (or alternatively, how many dollars it takes to buy a fixed consumer basket, they’re inverses of each other). But other dimensionless ratios shouldn’t be let go unchecked. Suppose I decided as the Fed to “print” a trillion dollars a month into the coffers of banks. Suppose those banks did nothing with it except gift it to the top 1000 executives. The top 1000 executives simply isn’t enough people to make a difference to the broad consumer goods market. They just aren’t going to go out and buy a trillion sandwiches or a billion automobiles. But is this a good outcome? Being blind to what’s going on isn’t a virtue on the Fed’s part.

              Now, the Fed can’t simply perform my UBI project because it’s not within their power. So, fine, let’s change that, because pumping money into the finance industry is what’s getting us into trouble in the first place starting in 1995.

              • Also regarding unemployment. I personally think unemployment is a joke that Economists should be ashamed of. It is *primarily* a measure of willingness to answer a telephone survey that “yes I am actively looking for a job” instead of admitting “no I’ve given up because my local economy is so bad”. The World Bank data shows Sweden’s labor force participation fell from about 70% to about 40% between 1990 and 2000 while their unemployment remained constant at between 5 and 10%, Norway dropped from 65% to 55% between 2000 and 2015 and their unemployment stayed below 5% the whole time:


                Labor force participation rate has no such psychological component, and in the US, it’s lower than its ever been since the advent of “women’s liberation” produced a massive influx of women into the labor force:

                Essentially 13 Million people dropped out of the labor force since 2000 and about 10M of those were since 2010.

                And you can’t explain it by boomer retirement either, because the same pattern is there in the age 25-45 range:

                So, if your goal is drive down unemployment, you could for example do it by making everyone have no job and give up entirely on work and live off the land, and hunt each other for meat, but I don’t think it will be a good outcome. My point is that we can’t just point to standard off the shelf macroeconomic measures like unemployment and the CPI and say “see everything is good” because the economy of the US has MANY MANY dimensions not 2, and even if you restrict yourself to the top 20 or 100 principal components of variation you are still talking about all of them being important. The fact that the Fed only cares about 2 of them because that’s what they’ve been told to do is not evidence that we as a nation are on the right track.

              • Martha (Smith) says:

                +1 to Daniel’s “Also regarding unemployment” comment.

  27. Lowk says:

    I don’t think I’ve come across this “increased housing supply driving up rents” theory before. What is the evidence behind it? Is it proposed as a Bay Area specific phenomenon (if so, why?) or has it been observed in other cities (if so, where?).

    • Imagine you have a hot water faucet pouring into a full bucket. AT the bottom of the bucket is some colder water, at the top hot, they don’t really mix just layer. New hot water pouring into the bucket just goes over the side.

      Now you raise the sides of the bucket by an inch. More hot water pours into the top and stays in the bucket. The water *in the bucket* is now hotter because all the new water that came into the bucket was hot. I think that’s all Phil’s theory is. His point is that new apartments will rent at higher rates, and so the median goes up.

      This isn’t supply and demand, because there’s rent control, so basically you have to wait for people to die in order to free up houses, and when that happens, the new house rents at top market dollar essentially independent of its quality.

      • Aaron G says:

        As I see it, the issue isn’t with “increased housing supply driving up rents”, since the housing supply that is dedicated or allocated for low-cost is not going up. What is happening that there is great demand for housing in desirable neighbourhoods, so the supply isn’t keeping up with that demand, and hence rents can be charged at the market value that is being charged.

        A very similar phenomenon is occurring in Toronto (where I live), which certainly in recent years has seen a relative economic boom (particularly in the financial, the tech start-up, and creative industries), followed by a boom in immigration from countries all over the world. All of these have resulted in a massive increase in demand for housing both within Toronto and the immediate surrounding areas. Even with a major construction boom in high-rise condo units and townhomes, mean (and probably median) home prices have been rising for the past 10-20 years (I believe the mean price for a single detached home in Toronto is now $1,000,000 CDN, and the price of a 2-bedroom condo is now above $500000).

        • Right, structurally what is causing high rents is 1) rent control and anti-building policies and 2) The Fed printing money and giving it to tech companies (essentially).

          Solve those two problems and you solve the problem of rent in SF.

          That being said, at the margin, adding a few additional apartments in SF will add a few additional hyperinflated tech jobs caused by a near infinite pit of money that the Fed is pumping out. And any apartments freed up will reset to market value. Net result is total cost of housing in SF goes up not down.

          BTW even the allocated “low cost” housing in SF is more expensive by far than the cushy deal the lifetime tenants are getting who moved in in 1993

      • Lowk says:

        What you have said makes sense, though it isn’t usually what people mean when they talk about rents rising (usually people mean “rents for an equivalent property go up” not “more of the houses are massive now so it costs more to rent them”). It is also possible that building above-median housing would lower the median rent (in the limit, building an infinite number of luxury apartments would lead to rents on those apartments being driven down to near-cost) – knowing which will happen is an empirical question. But I can believe that, in SF, building small numbers of luxury housing would increase the median rent (though presumably in a way that has little effect either way on the lower end of the market).

        However, that didn’t seem to be Phil’s theory (at least, not all of it). Phi said “there will be increased competition at the lower end of the market, which means higher rents”. This is the part that I have not come across before, and was wonder what evidence there is for it – that building luxury accommodation can actually INCREASE the rent of lower end market properties. I think I get the intuition behind it (driving up demand for service jobs drives up demand for low-end accommodation and thus increases prices), I just wondered if there was any evidence that this effect is large enough to council out the classical effect of increased supply (as everyone “moves up” the accommodation ladder).

        • Part of the issue is that when someone moves out of an SF apartment the rent resets to market rate, now if you build luxury apts some people who are earning a lot but living in low-quality places at rent controlled prices from 4 to 6 years ago will finally move to their dream home. But their 4-6 year old rental will go on the market, and due to incredibly tight supply, the price of this low quality rental will be sky high and probably also rent to a high earner. The skyrocketing rents ensure that *whenever* you have a vacancy the new rent will be WAY more than the old rent-controlled rent.

      • NatashaRostova says:

        I don’t think there is a strong reason to expect that the thermodynamics of water in buckets is a good model/analogy of housing prices.

        • It’s not a model of housing price dynamics as much as it’s a model of averages over many quantities. If you average over all the housing stock, and you add a few additional ones, and they are all high cost, then you raise the overall average “mechanically” as Carlos Ungil mentioned elsewhere.

          There is really no arguing with this, it’s just a fact about numbers.

          Even the argument that some people inside the city will move out of their old apartment into the new one and free up the old one which could then decline in price… when they do that their new one resets to market rate, and it may reset to slightly less than it would have if they hadn’t built the new apartment, and the person in the old one still decided to move out… but the fact of the matter is that DOESN’T HAPPEN. No one moves out of rent controlled apartments in SF of their own free will because they inevitably are receiving an enormous wealth transfer. So when the person moves out of the old apartment into the new apartment, the new person moving into the old apartment is paying way more than the previous rent… again average rents across all apartments go up.

          The bigger question is whether or not we care about average rents.

          If the stated goal of YIMBY people is to increase the supply of low rent housing then we have to understand what they mean by “low rent” and if what they mean by “low rent” is “the same as at current market rates but more of it” then they might accomplish this by building high end housing, but if what they mean is “make it possible to live in SF for less than $180k/yr for a couple with 2 children” then that’s not going to happen until the tech crash occurs.

          • NatashaRostova says:

            I speculate that it’s probably true that the age of a lower-middle/middle class family living in SF proper, and raising a family, is over. The question, I suppose, is whether the “tech middle-class” so to speak, can live in SF and raise a family. Right now they can’t. I left SF for that reason.

            A better housing policy might have meant, on the margin, people like me didn’t leave SF. To the extent that this would have improved (my) welfare, that’s a good thing I guess.

            >until the tech crash occurs.

            May it never happen. Praise Thiel.

    • Phil says:

      Increased housing supply does not drive up rents. If you build more housing, the average housing price goes down. Please see the fifth paragraph of the post.

      • Lowk says:

        Sorry for being unclear, I understand that you are proposing that new housing increases rents in SF specifically (with an overall decreasing in rent across the entire closed system). I just wondered what the evidence for this proposal is.

  28. Phil gets one thing right here- our goal isn’t, or shouldn’t be, simply to lower prices. It is to make people better off.
    In a basic supply-and-demand model, an increase in supply both lowers prices and increases welfare.
    In a supply-creates-its-own-demand model of the type Phil outlines, an increase in supply has an ambiguous effect on prices but actually increases total welfare even more.

  29. gdanning says:

    It is hardly surprising that rents in SF are sky high – the number of housing units in SF has increased only by about 25% since 1970 (from about 310,000 to about 387,000 – see chart here, while the population of California has almost doubled in that time

    Given that demand for housing is highly inelastic, a relatively small increase in the supply of market rate housing should cause rents to decrease quite rapidly. (Of course, that will only happen if the increase in supply is greater than the increase in demand – if demand rises faster than supply, as it apparently has of late, then rents will continue to rise – but not as much as they would if no additional housing had been built).

  30. wow says:

    Phil, this sounds like it was written by a know-it-all freshman physics major. Maybe you should stick to physics and leave economics to, hmmm, economists?

    > Why, I wondered, are these people promoting policies that are so bad for them?

    Maybe it’s because you don’t understand economics? No that’s not possible. It must be those people are mean and spiteful

    • Andrew says:


      This is not a comment on the merits of Phil’s argument, one way or another—it’s been a long time since I’ve lived in the SF area and I’ve not been following what’s happening there—but I do have a problem with your argument. My problem is that your argument “explains too much,” as the saying goes.

      People promote policies that are bad for them all the time! When it comes to my neighborhood in NYC, I’m a yimby, and I do think that lots of the nimbys are promoting policies that are bad for them! You might even think that the nimbys in SF are promoting policies that are bad for them. That’s one reason we have the field of economics in the first place, to help us understand the effects of policies.

      Also, Phil never said anything about people being “mean and spiteful.” That’s all coming from you. Maybe you feel that those people are mean and spiteful—and maybe you’re right, after all, mean and spiteful people do exist—but that’s not coming from Phil at all, so you’re criticizing him for saying something he never said.

      • Well, Phil did say that he thought perhaps the motivation of YIMBY policies is to poke the people who have a great rent controlled apartment in a nice neighborhood in SF in the eye for being such slam-the-door jerks. specifically quoting Phil:

        “the YIMBY and BARF people know that building more market-rate housing in San Francisco will make median rents go up, and that this will be bad for them, but they want to do it anyway because it’s a thumb in the eye of the “already-haves”, those smug people who already have a place they like and are trying to slam the door behind them.”

        That sounds like Phil is attributing spiteful attitudes to YIMBY proponents to me. Or at least, he’s considering it as one possible explanation. “YIMBYs want to make SF worse to spite the rent-controlled fortunate people who are keeping the rest of society out” It’s not that implausible, but I don’t think it’s the answer either.

      • wow says:

        Andrew, I was referring to

        > So this is my new theory: the YIMBY and BARF people know that building more market-rate housing in San Francisco will make median rents go up, and that this will be bad for them, but they want to do it anyway because it’s a thumb in the eye of the “already-haves”

        That seems to me to be an accusation of being spiteful.

        > But suddenly it dawned on me, just last week, that the question “why are people in favor of policies that are so bad for them” might have the same answer in this case that it appears to have for a lot of people in national politics: they aren’t trying to do something good for themselves, they are trying to hurt their perceived enemies.

        That too. Seems pretty spiteful to me.

        Overall, the article is patronizing and insulting.

        Did Phil consider that maybe YIMBYs and BARFs support these policies because they actually expect them to help the less fortunate? Maybe they oppose measures like rent control and onerous zoning because they have been empirically shown to hurt to hurt the poor and middle class. Maybe they oppose them ideologically.

        Speaking of empirical research, where is it in this article? This isn’t some novel concept. As others have already mentioned in the comments, there is a huge literature on it. Instead, all this article includes is a half-baked model from someone with no economics background. I know I’m being somewhat harsh, but this article doesn’t live up to the standard of your blog. I understand it’s a blog, not an academic journal, but this article fails to reach any semblance of quality.

        • Matt says:

          +1. Really annoying to read this, as an economist. So Phil is worried that increased market-rate housing increases the demand for lower-wage workers, which will ultimately make lower-wage workers worse off (!!??). Just because there is more demand for low-wage labor (causing upward pressure on their wages) does not mean it will be met (especially if an inflow of workers will increase renting prices). Basically, if we imagine all lower-wage workers as being identical, I don’t think there is a way in which building market-rate housing can make these lower-wage workers worse off. In your line of reasoning, more rich people moving in increases demand for low-wage worker. This could cause workers to flow in, putting upward pressure on renting prices for this income group….but renting prices will never increase so much as to make these workers worse off. If they did, than low-wage workers wouldn’t find it worthwhile to move in, because higher wages are being offset by terrible rent. Honestly, just look up Rosen-Roback spatial eq’m model. Captures all the intuitive arguments you were making somewhere else in comments about preferences over amenities.

          • Phil says:

            Matt, I’m not making any claims about whether workers are better off or worse off or whatever. I’m claiming that some people who say they want the rent in their area to go down are supporting policies that will make the rent go up instead.

      • Michael Watts says:

        > Also, Phil never said anything about people being “mean and spiteful.” That’s all coming from you.

        Here are Phil’s own words:

        >> they aren’t trying to do something good for themselves, they are trying to hurt their perceived enemies

        >> the YIMBY and BARF people know that building more market-rate housing in San Francisco will make median rents go up, and that this will be bad for them, but they want to do it anyway because it’s a thumb in the eye of the “already-haves”

        I don’t think it’s at all unfair to object to this by characterizing it as “these people are wrong because they are mean and spiteful”.

        • Phil says:

          OK, I’ll own it: I hypothesize that there is an element of spite in the YIMBY/BARF stance. I don’t claim this is proven; I don’t know. But this would explain why some people are supporting policies that run counter to what they claim is their desired outcome.

          As for ‘patronizing and insulting’, I dunno, could be. I’m more interested in whether it’s right or wrong!

          • wow says:

            > I’m more interested in whether it’s right or wrong!

            Are you? You clearly didn’t do any research of what actual economists have written about the subject. You should start there since this isn’t a new topic.

            > But this would explain why some people are supporting policies that run counter to what they claim is their desired outcome.

            That the policies result in the opposite of their stated desires is something you “proved” with an amateur model and no empirical backing. This is in contrast to what actual economists have rigorously developed. So maybe you’re trying to explain a phenomenon that doesn’t exist in the first place.

            • Phil says:

              Could you give me a citation? I’ve looked on Google Scholar and it’s sort of astonishing how little work seems to have been done on this topic, or at least that I can find. There’s a fair amount of stuff on the effect of rent control, and there’s some work at national scale, but I’ve found nearly nothing that is directly relevant.

              • Lewis Lehe says:

                Here is a good summary of negative assessments:
                Here is a more recent summary, more guarded in its conclusions: “In summary, most models and empirical estimates suggest that regulation reduces
                aggregate welfare, on net. And the estimated effects are often economically large.”

              • Phil says:

                Lewis, thanks for passing those references along.

                The first (Ellickson) is rather fact-free, mostly just asserting that land use regulations increase housing costs and that the author thinks that’s bad. People who object to my post being just a bunch of blather should note that this is even worse! But at least it has a bunch of references, some of which I will take a look at.

                The other (Gyourko and Molloy) is much more interesting and more to the point. It even has a concrete assertion about housing prices in the Bay Area: that, if building were unregulated, they’d be about 35% lower than they are now. (The claim is that the cost of regulations increases Bay Area housing prices by ‘upwards of 50%’). But this is a rather general work and doesn’t go into the relevant details that I think are most relevant to my post. The first of those is, what is the marginal effect (on the San Francisco job market and on San Francisco rents) of building another market-rate apartment in San Francisco? And the second is, how many more apartments would San Francisco have to add in order to cause a noticeable downward movement in housing prices? If the answer to the second one is “San Francisco needs twice as much housing as it has now”, I’d find that plausible but I’d also say I’d be surprised if it’s politially tenable. I think even many people in SF who think of themselves as anti-NIMBY would balk at doubling the population of the city in a short time.

                Thanks very much for these links. But I have to say that overall I don’t see anything that argues against my narrow assertion, which is that in the current environment building more market-rate housing in SF will cause rents in SF to go up, not down.

                By the way, another commenter passed along
                this link to a blog post by Nick Rowe which I think is interesting.

          • sonja trauss says:

            Oh absolutely there’s quite a bit of spite motivating our activism. Mainly it’s the spite that anti-housing residents must have for us, for them to be so actively and intensively frustrating the production of housing for us to live in.

            Basically we have, in the Bay Area, homeowners who shut down proposed new home and apartment construction because they don’t want to have to look at the new apartments (protect neighborhood character!) Or because they don’t want parking or traffic to get worse, but they have no empathetic imagination for where the people who would have lived in those apartments will live instead, if they’re not built.

            If any anti-housing homeowner thought about that question for two seconds they would realize what jerks they are being, because by blocking new housing, they are causing a chain of bad outcomes, not only for the people that would have lived in that new housing, if it existed, but also for the people that would have lived in the housing that the potential-new-housing-resident did wind up living in.

            If you want an example, I can take it from my own life. I got into this activism because I moved to the Bay Area to live in SF, but I could only afford to live in West Oakland. I was making $10.50/ hour working at night in a Bakery, so I was a poor person living in a poor area. This was 2011. All around me were people moving to West Oakland from San Francisco, because they had been displaced from San Francisco by rising rents. Like me, they didn’t prefer West Oakland, we all would have rather been in SF. In addition to the welfare loss that my new higher income neighbors felt personally (they’d rather be in SF), there was also the welfare loss to the low income person that this new higher income person _directly replaced_ in the existing West Oakland apartment. The low income former West Oaklanders who had to move out also experienced a welfare loss when they moved to East Oakland, Concord or Antioch.

            It’s not just frustration of personal preference. My immediate neighbor who was displaced by a higher income renter went from commuting from 12th & Peralta to Alta Bates at 30th & Telegraph, to commuting from Concord to Alta Bates! That is an objective reduction in quality of life. Long commutes are bad for your kids, all things being equal. Also, hospitals have no tolerance policies for lateness (my mom was a nurse). Having a longer commute inevitably increases her chances of being late to work and getting whatever the hospital’s version of demerits is.

            Amongst all this, concerned neighbors in SF were valiantly protecting their quality of life by opposing apartment building construction, at a time when a Bay Area wide shortage of housing was completely obvious. Berkeley is not off the hook, just because it’s not the town that initially inspired my activism. Suppressing housing in downtown Berkeley is causing displacement in South Berkeley and North Oakland, and even as far as West Oakland for that matter.

            I don’t why this is so hard to understand. Phil maybe you should do an activity at home. Draw 6 circles on a piece of paper. get 10 paperclips. Mark some number of the paperclips as rich guys. It actually doesn’t matter for this exercise how many rich guys you make. Could be 1, could be all 10. Maybe tie some string around the rich buy paperclips so you don’t forget who they are.

            Now put the paperclips on the paper so that there is one in each circle. The rich guy paperclips get first dibs at the circles, because that’s how the economy works. If there are any circles left over after the rich guy paperclips all get circles, then the regular paperclip(s) can go there.

            You’ll notice, that since you drew 6 circles, but you have 10 paperclips, that there are 4 paperclips left over, not in any circle. Those are the displaced renters. Maybe they live in West Oakland, maybe in Concord, maybe in New Mexico. Another thing you could do, is put a bunch of regular guy paperclips in one circle. That is illegal in the game, and illegal in life, that’s overcrowding, but people and paperclips get desperate. Whether they are crowded, or whether they are displaced, the paperclips with no circle of their own are sad. :(

            Now draw two new circles, maybe with a different color marker. This is New Housing. Now you have 8 circles. The new housing can only be occupied by rich guy paperclips. If you had a rich guy paperclip without his own circle, he can go in the New Housing, or you can move a rich guy paperclip from the crappy old circle he had been in to the new circle, and move a displaced regular guy paperclip into the newly empty circle. Now you’ll notice only 2 paperclips are displaced! Those two are still sad :( but the overall situation is better than when there were 4 paperclips that were sad.

            You can keep drawing circles until all paperclips have their own circle.

            • Phil says:

              Look, I agree, the rent is too damn high. It’s ridiculous.

              But I think if people in SF want the rent in SF to come down, they should be arguing to either build BELOW-market housing in SF, or to build market-rate housing in surrounding cities. I think building more market-rate housing in SF is counterproductive if the goal is to decrease rents there.

              • sonja trauss says:

                I don’t think you did the activity.

              • Dan says:

                The problem is that this is a strawman. YIMBYs want to reduce rents region-wide, not in a particular neighborhood.

                And the policy you’re recommending leads to a prisoner’s dilemma where everyone advocates for housing built ELSEWHERE. What do you think the net effect of that would be, given that zoning decisions are made at the local level?

  31. Cody L Custis says:

    “Of course there is plenty of development pressure, and new high-rise apartments are going in that have hundreds of apartments each, typically with a rent of $4000 – $8000 per month. If you let a developer build “market rate” apartments, that’s what they’ll build.”

    As long as the rent of apartments is greater than the cost of building (land + construction), developers will continue to build until the two equalize. The problem with San Fran is that zoning regulations prevent the two from meetings. Furthermore, current residents have great incentive for such regulations, as they drive up the price of the properties that they own.

    • Foster Boondoggle says:

      For most people, most of the time, owned housing is an amenity that they consume, not purely an investment. As a Berkeley resident, I’m pretty sure that most of the local NIMBY opposition has to do with quality of life, not an expectation of higher property values. (Which they’d never directly benefit from if they don’t move.) Personally I find that short-sighted. What’s great about dense cities like Manhattan is the amenities right outside your door. Berkeley could benefit by more walkable density in place like the gourmet ghetto and Solano ave. But I understand the desire to preserve the character of the community you moved into.

      • Phil says:

        I agree with Cody — and I think I said in my post — that if you build enough, you do eventually drive down rents even within the city. But in the case of San Francisco I think you’d be talking about totally changing the character of the city. As long as people find it an unusually attractive place to live — a combination of job availability, lifestyle, climate — it will be unusually expensive. “Let’s build until it’s not attractive anymore” is not an appealing message.

        I think Foster has it exactly right, that there are people who like high-density places like Manhattan, and people who like lower-density places, and people often want to preserve the character of the community they moved into. For a lot of people, the reason they live in their neighborhood is because they like it the way it is.

        Not that anyone asked but:

        I grew up in low-density suburbs (1-acre lots) so most amenities were pretty far away. It seemed normal to me, because that’s what I was used to, until I went to college and had friends, restaurants, movies, etc., within a few blocks. I decided I vastly prefer that.

        When I went to grad school, some of my classmates chose to live in nice new apartment buildings on the outskirts of town, in a complex with a pool, and laundry in the building, etc. They could have all of that, and a big apartment, for what I paid for a small apartment six blocks from campus. I thought they were crazy, and still do.

        Foster clearly knows Berkeley…for him and for other people who do, I’ll mention that when I moved to the area, I moved into Andrew’s old apartment on the border of Berkeley and Oakland (on Alcatraz just above Telegraph). Busy street and not a great apartment, but I loved the area: easy bike to work, to restaurants and shopping, and to Ultimate Frisbee games. A few years later I moved in with my girlfriend (now wife) in the Gourmet Ghetto area…not quite as lively as Rockridge but everything I need is within easy walking or biking distance. It’s awesome.

        There’s a lot of Berkeley I wouldn’t want to live in: too low a ‘walk score.’ There are lots of expensive homes in the hills that are thought to be nice, but any time you wanted to go anywhere you’d have to drive. Ugh.

        Anyway, I am a huge fan of medium-high density and I kinda feel like everybody should be. But if someone wants to live someplace much denser, that’s great, go for it.

        • sonja trauss says:

          “But in the case of San Francisco I think you’d be talking about totally changing the character of the city.”

          Yes, that is what we are talking about. We are talking about totally changing the character of the city. That is the political movement, that is the goal.

          • Phil says:

            Well, I’m not surprised that a lot of people in San Francisco object to it, then!

            Surely the goal is not to live in a city named “San Francisco.” If merely living in a city called San Francisco were the goal, Detroit could rename itself “San Francisco” and attract a whole passel of new residents. A pretty large fraction of people who live in San Francisco are there because they like many aspects of the city. “Totally changing the character of the city” is not going to be very appealing to them.

            • sonja trauss says:

              No. The goal is to live in one of the most job rich, opportunity rich, high productivity areas in the world. We don’t care if it’s called “San Francisco” or if it’s filled with 2 story bungalows built in the 1950s. In fact, we’d prefer it not be filled with 2 story bungalows, because if instead it was filled with 5 story apartment buildings, more of us could live here and have jobs here, and have opportunities here that we don’t have access to in the towns we were born in.

              If silicon valley moved to Detroit, we would happily move there. It’s not the name, it’s the jobs.

              Yes, thank you Phil, I am aware that there are many people in SF who do not share the political goal of my organization. We call them “our opponents” and my job, literally my job, I work full time on this, is to organize volunteers to frustrate their goals and to advance our goals.

              • Phil says:

                Sonja, San Francisco is not the only place with a hot job market. There’s Oklahoma City, Columbus, Milwaukee, Salt Lake City… if it’s really true that your members and supporters don’t care about the character of the place they live, there are other options.

  32. Jack says:

    He made a claim, you may disagree, but how many people above, bristling at the thought of being driven by spite and anger, only bothered to post because they were angry and subsequently took a spiteful tone?

    Many people are so upset by the tone taken in the article. It occurs to me that regardless of right or wrong, if we take a certain tone, e.g. “Why do people act against their own self-interest” that the argument becomes about the sonics/phonics, not about the facts. This is informative when thinking about the far left v left v the right atm, where when we identify policies that blatantly work against the facts as we know them, we are quick to try to rub the other group’s face in it. But being on the other side of it now… seems like that is unlikely to win hearts and minds, and instead people dig in further. And you can say “yes but we have the facts” but they certainly feel that way as well :)

    Additionally, there are a number of good responses and I don’t necessarily agree with the author’s key points, but some of the counterpoints explicitly ignore the author’s examples e.g. people excoriating him about his lack of understanding basic economics, but ignoring his example about NYC not getting substantial relief from incremental housing as those same economic principles may recommend, or people cherry-picking examples that don’t necessarily support the claim, e.g. Tokyo as though they are slam dunks.

    We need to, in the face of even a perceived lack of civility, find ways to offer people safe outs while we pick apart the policies, otherwise our polemic political situation will just continue to intensify.

    • I think above you will find a lot of professional economists. They bristle at people doing economics without first consulting their profession to ask what might have already been done on this issue. In this, they have a good point, but that point would be heard better if they had in fact been actually going down to the damn planning meetings and explicitly explaining the issue, and written some op-eds in the Chronicle and posted graphs and charts on blogs, and gotten involved with people to promote policies that help society. Unfortunately Economics seems to have a “hands off” attitude towards this stuff, perhaps it’s just too political.

      They also have jargon which Phil isn’t using. “median rents increase” to Phil I think means “taking the median over all occupied rental units, the value goes up” to an Economist it means “taking the median over unoccupied units offered in the market, the value goes up more than it would have if we had done nothing”

      So rents offered on Craigslist could go from say $3000 to $3500 and an economist would say “see rents went down” because they have some seasonal adjusted trend line that says they should have been at $3700 by the time we measured $3500 etc etc. This is Jargon because they are always talking about counterfactuals and so the “relative to what would have happened” is just tacked on in their heads. There is some very important thinking that goes into this point of view. Yet, it is not necessarily the case that this nuanced Economic point of view is the same as the YIMBY advocacy POV. Note that the dollar cost of stuff just tends to go up in time all the time, so if that’s the baseline, as an Economist it rapidly becomes more important whether the increase was more or less than under some alternative, whereas YIMBY advocates probably don’t think that way, most people don’t they think “I want to spend fewer dollars”.

      Phil claims the stated goals of YIMBY are to “[have] policies changed so that rents will go down, but [they] are strongly in favor of building more market-rate housing.”

      Phil’s claim is that all else equal, if you build more market rate housing, the median rental value of an apartment will be higher. As Carlos says, this is “mechanical” if no turnover occurs in the existing housing stock, and it’s essentially mechanical as well in the presence of rent control even if there is turnover in rental stock because at the growth rates seen in the last few years even someone who rented 2 years ago is enjoying 10 to 20% lower than market rents for their place, and new vacancies go for market rates.

      Economists rightly claim that they have studied this and rents really do “go down” in the sense that going up from $3000 to $3500 is “going down” compared to going up from $3000 to $3700, but this isn’t the stated goal of YIMBY “keep the growth rate of rents lower than they would have been” it’s *reduce the price of renting an apartment* and that *simply isn’t* going to happen until a tech bust or a massive building spree.

      One thing that I’ve only seen Carlos Ungil mention, but perhaps others did as well, is that what matters is not the dollar cost of renting an apartment, but rather the RATIO of dollar cost of renting to After Tax Income of say a Barista or a Plumber or a Hairstylist. It’s this “money illusion” of thinking that a dollar is a dollar which further complicates any clear rational thinking. And I notice that it certainly wasn’t something the Economists above were clearly pointing out, so I hope they will agree with me here once I mention it.

      Of course, if we stop talking about at the margin of a few thousand apartments, it’s easy to make SF housing prices go down. Here are several options:

      1) Small Tactical Nuclear Warhead
      2) Enormous Earthquake, massive radiation release, zombie infestation…
      3) Tech crash
      4) Build really A LOT of housing, like double the quantity currently available in SF.

      I vote 3 and 4 are probably better than 1 and 2, and probably most people who have been living in SF more than 5 years would prefer 3 to 4 because housing booms tend to damage incumbent’s property values, and also they definitely alter the environment in a way that has other externalities for the incumbents (traffic, blocked views, parking issues, ugly buildings, whatever).

      I conclude that the main mechanism by which the pressure relief valve will actually come is the popping of the current tech bubble.

      However, please note, many of the Tech Bros are holding straight up CASH in their bank account waiting for that crash to come. (evidence of this is available in the M1 money supply graph: I don’t know about you but my checking account didn’t double in value over the last few years, but the total money in checking did, and if that’s confined to say 10% of the population, then these people have about 20x as much cash as before the crash, so if they had $20k before now they have $400k in checking) An estate lawyer from the Bay Area tells me that the “wealthy” clients are the ones *STARTING* at $6 Million in cash in their checking account and there are plenty of them.

      So, while long term the tech crash may eventually reorient the SF Bay, I’d say overall it’s kind of too late to salvage, even after the pop, unless banks collapse and the FDIC makes that $6M in padding go up in a puff of smoke… we’re stuck with people who participated in the tech boondoggle being the top dogs in the area.

      If only we hadn’t started the whole mess with poorly thought out rent control laws back in the early 80’s and there had been gradual building and gradual rises in rent just like there were gradual rises in the price of cars and food and toiletries and basically everything else. Rent control that specifies a mechanical increase of half the CPI per year (as Berkeley does for example) is a terrible idea.

      • Carlos Ungil says:

        > Unfortunately Economics seems to have a “hands off” attitude towards this stuff, perhaps it’s just too political.

        Economists who have ventured into the alleged real world often quote Princeton’s Alan Blinder, who has formulated what he calls ”Murphy’s Law of economic policy”: ”Economists have the least influence on policy where they know the most and are most agreed; they have the most influence on policy where they know the least and disagree most vehemently.” It’s flip and cynical, but it’s true. (….) The analysis of rent control is among the best-understood issues in all of economics, and — among economists, anyway — one of the least controversial.

        • Yes, pretty much everyone who has ever looked at theory or empirical results on rent control agrees including heavily socialist economists who are all about helping the poor and labor and all that, rent control is a bad idea. Of course, 90+% of voters who are to vote on it benefit from it directly, and those who are hurt are typical people who are too young to vote (they will be hurt later when they want a place to live) or people who live outside the area who would like to move to the area currently or in the near or far future. The remainder are people who invested in real estate, and they comprise a few percent of the population. Also, voters are not among the people who have any knowledge of the theory or empirical results, and besides we’re talking about a major long-term continuous transfer of wealth into the pocket of renters if they vote yes.

      • Daniel,

        “I think above you will find a lot of professional economists. They bristle at people doing economics without first consulting their profession to ask what might have already been done on this issue. In this, they have a good point…”


        “…but that point would be heard better if they had in fact been actually going down to the damn planning meetings and explicitly explaining the issue, and written some op-eds in the Chronicle and posted graphs and charts on blogs, and gotten involved with people to promote policies that help society. Unfortunately Economics seems to have a “hands off” attitude towards this stuff, perhaps it’s just too political.”

        Sorry but no. There is LOTS of good stuff out there, good blogging, good papers, good op eds, good public policy work. But, sadly, lots of awful stuff too.

        This is a statistics blog. Lots of very intelligent stuff written by smart people who know their stats, yourself included. But consider the mass of terrible empirical applied stats being done out there. Indeed, it’s one of Andrew’s worthy crusades, and we support this and maybe even try to help.

        Are you going to suggest that it’s the fault of statisticians that there’s all this awful empirical work being done? And that there are no good blogs out there, or good op-eds, or whatever? Well, I guess there is an angle there about teaching stats etc. But mostly we blame the people who engage in these bad methods. And you know where the good stuff can be found, because you just know.

        Same with economics and economists. We’re not blameless, but most of the terrible economic policy out there can’t be blamed directly on the economics profession, just as most of the terrible applied stats work out there can’t be blamed on statisticians. And we know where the good stuff can be found, but that doesn’t make it easy for the non-economist to find it.

        There’s a related parallel that takes us back to Phil’s post. It’s easy for someone who doesn’t know any economics to think they can just walk and do some sensible analysis of a complex question. It’s also easy for someone who has no stats training to think they can just walk in and do some sensible analysis using data on whether A causes B. Many other fields are not like that – cosmology, immunology, whatever. But it’s a problem for stats and for economics alike.

        • I absolutely blame the stats profession for all the terrible stats being done. It’s right there in the textbooks how to do terrible stats!!!!!!

          In this particular case. I strongly believe that the two sides are talking past each other. For example

 and its replies


 and its replies

          and probably several other places. The thread is a bit of a mess due to many nested reply chains.

          My guess at what Phil means is basically here:

          And I’d be very interested to hear what a professional economist thinks about it, because I think it’s obviously correct given that it exactly tracks my personal knowledge of what goes in in rent controlled places (I personally know several instances such as uncles, friends, etc who live in heavily rent controlled areas or who have moved into SF temporarily and then moved out etc)

          The mistake that professional economists seem to be making in this thread is to pretend that this is a market economy instead of dealing with the reality that it’s a heavily rent controlled situation and heavily construction controlled. This is exactly the Economists version of “Consider a spherical racehorse in free space”. High spot prices DO NOT cause an increase supply, in fact they cause decreases in liquidity, no one moves out of a rent controlled apartment when they pay $900 and the spot price is $3800. Also, the existing prices do not represent equilibrium clearing prices precisely because there’s no liquidity. People advertise their apartment for $2800 and 36 people apply in the first 8 hours each bidding up the last, and the clearing price is $3700 which is never observed because all you see is the craigslist ad for $2800.

          Treating the SF market like an equilibrium perfect-information clearing-house instead of like an illiquid penny stock is a major mistake. Attacking Phils physics background instead of asking him questions about what he means and trying to get him to clarify and then teaching him how to think like an economist using the jargon and the appropriate concepts and then asking him to rephrase his questions in common terms… it comes off as smug. In an intellectual argument with a non-specialist, one should bend over backwards to see if the opposing side has some useful point that you’re missing or misunderstanding because they don’t use words the same way you do, and you should try to teach them how you use words.

          I think the economics profession shows itself in a bad light in this thread, in the same way that if some biologists had accidentally discovered Bayes rule and then some stats professors jumped down their throat about not understanding that all probabilities are frequencies… that’d be pretty bad behavior, especially if the stats profs know that Bayesian non-frequency distributions are a real thing, but they just don’t want to engage that and would rather stomp it in the bud.

          SF is an illiquid information poor regulated market that has experienced constantly increasing unobserved demand in the double-digit percentages for the last 10 years, with a major explosion of people in the surrounding areas just waiting to swoop in on anything that looks like a deal. I suspect you could add 5000 apartments in SF at $10,000/mo and they’d all be taken within the first week, and every person who moved out of an SF apartment to move into the new ones would leave behind an apartment that rents higher than it used to. Yes possibly not as high as if the 5000 new apartments hadn’t been made, but *that’s not what Phil was talking about*

          So, Phil may come off as not very economically sophisticated, but the economics profession comes off as smug, detached from reality, and unwilling to engage anyone outside their profession. It doesn’t look good.

          • Keith O'Rourke says:

            > one should bend over backwards to see if the opposing side has some useful point that you’re missing or misunderstanding
            Agree – at least if can make the time to do this.

            • “at least if can make the time to do this”

              And also if you think the person is willimg to learn andinterested in what you have to say.

              I’m sorry, but my reaction, and the reaction of the other members of my tribe, is (a) there is likely very little there for us to learn, (b) Phil just dived in without bothering to read any of the relevant and serious econ literature out there. Or maybe he read a bit and didn’t think it worth much. So you want us to engage patiently and carefully with someone who starts out by indicating he’s, well, just not very interested in what the economics discipline would say on this.

              To take your Bayes’ Rule and biologists example – say some biologist had posted “here’s a cool rule of thumb for combining prior information and new data” … cited a way of doing it that didn’t make use of Bayes correctly … and hadn’t shown any interest in trying to find out whether statisticians had already done anythiing like this. I think you’d be grumpy. I exaggerate a bit here, and maybe I’m being a bit harsh, but really, that’s how it comes across.

              I’m a lecturer by profession and I take teaching very seriously. But engaging in a time-consuming debate (after reading up on it first – I’m not an urban economics specialist, nor do I know the institutional details of the SF property market) with someone who is apparently not interested in what people in my area of expertise have to say, or at least not interested enough to, you know, like, read a bit first … meh. Life’s too short.

              • Ok, fine I think that’s a perfectly valid viewpoint, but evidently lots of economists read this blog and are willing to engage enough to denigrate Phil, but not enough to actually understand much about what his question really is. He’s got a perfectly reasonable and fairly mathematically well defined issue regarding the time evolution of the price distribution of occupied housing, it took him a while to express it but it’s right here in this comment:


                and yet although I think this is at the heart of his original question, I have yet to see an Economist actually even realize that this is what we’re discussing here, but there are a lot of comments like:

                “I usually love this blog. This was literally the worst article ever posted on the blog.”

                and the like, as well as tons of effectively “you don’t understand supply and demand” and “even if housing prices in SF might go up, they will go down in surrounding neighborhoods” (a fact Phil EXPLICITLY AGREES WITH in his original post) etc.

                I have yet to see an economist who even acknowledges the actual subject here. Phil was explicit enough in the above linked comment: What happens to the distribution of prices of occupied buildings when you add more market rate housing? Is there *any* conceivable mechanism by which adding market rate housing extends the left tail of the distribution lower? (ie. creates more living opportunities for people like plumbers and kinkos copy center employees and baristas and daycare teachers?)

              • also by “more” i mean on the order of 1-5% more housing stock not tripling the number of apartments.

              • If you just re-read Phils original post, and everywhere he says “median rent” replace it by “median rental rate of occupied units” then is there anything wrong with his post? I don’t see anything. Of course, if Economists interpret “median rent” as “median spot price of available units” and “go down” is allowed to mean “go down relative to the counterfactual where no additional units were built” they get their panties in a bunch… but those are the natural interpretations for an Economist and are TOTALLY UNNATURAL meanings for a Physicist.

                A physicist is thinking like “there are a bunch of binding sites for atoms, and a bunch of atoms are binding to the sites and we add some more binding sites at the high end of the energy distribution… and we shake it a little and allow it to equilibriate, and thanks to the large amounts of external high energy atoms, this just makes the net energy in bound atoms higher, even when it knocks some atoms free of their sites, the external forces are going to cause the emptied sites to bind at high energy”

                and Phil is *absolutely right* when you take that point of view. It’s just a totally unnatural view for economists to even think about.

              • Phil says:

                Mark, I encourage you to move on to the Sorry / Not Sorry post, I think nobody is reading this one anymore. I almost didn’t see your comment myself.

                I agree that I deserved some of the harsh comments. There are others I don’t think I deserved. I’m not complaining about the root-mean-squared treatment that I got here, and I learned some important lessons.

            • Alternate version: if Phil had started by with a bleg asking for good stuff to read on the the analysis urban housing markets in general, and the SF market in particular, and *then* followed up with his analysis, besides a better-informed analysis I think we’d also see that the reaction of the economists who read this blog would be rather different.

              • Yep, that’s also a good point. When it comes to politically sensitive ideas, it’s probably best to be very sensitive and careful in your communications strategies. But, I do think that the fact that I have some knowledge of both physics and economics has led me to see why Phil describes things in language natural for a physics guy, and economists interpret them entirely differently as if they were language natural for an econ guy, and then the two fight it out without ever even once realizing that they’re not talking about the same things!

              • “When it comes to politically sensitive ideas, it’s probably best to be very sensitive and careful in your communications strategies.”

                Sorry, I disagree again. Not with the statement. But being “sensitive and careful in your communications strategies” is not the issue.

                The issue is scholarship and what it means to be intellectually responsible (in the sense of “acting responsibly” – sorry, I am trying to be sensitive and careful :) in how I phrase this).

                I’m currently visiting an institute for advanced study, and I’m surrounded by biologists, anthropologists, mathematicians, political scientists, historians…. Very smart and very knowledgeable, one and all. How do you think they’d react if I tried to engage them in a topic in their area without making any serious attempt to first learn – from them or on my own – about what scholars like them and their predecessors had done on the topic? And I just waded in and made stuff up? Using my jargon and not theirs?

                Phil seems to be moving in the right direction, from what I read in the comments and the follow up post. But the way he started, in terms of being a good scholar, was just awful. And for that I think the harsh responses were well earned.

          • Also, economists have jumped directly to “the spot price is all that matters why would you even talk about the price distribution of occupied units?” and this isn’t necessarily the case. In fact Phil explicitly says that he wants to talk about the distribution of prices in occupied units in this comment:


            In particular any political movement whose purpose is to *try to increase the number of lower income people living in an area* (which is I think what Phil is assuming for YIMBY, possibly incorrectly) would care not just that the spot price didn’t go up, but specifically that *there exist apartments that could be rented to “low income” people* in other words, that plumbers or pre-school teachers or whatever could live in SF. If a pre-school teacher can only afford say $1300/mo and the the cheapest thing is $2200 but it would have been $2400 if the market rate apartments hadn’t been built…. then *this just goes to show that Phil was right in implicitly saying that making market rate apartments did NOT increase the supply of apartments affordable to pre-school teachers* in other words, the distribution of actually occupied homes shifted right, it didn’t even also spread into the left tail, the whole thing moved right.

            So, again, I think it’s all about miscommunication, and smug lack of engagement with the real questions. Economists in this thread immediately changed the question from “Why would YIMBY people think by building market rate apartments that we’re going to get more cheap apartments for pre-school teachers?” to “You obviously don’t understand supply and demand, why would you think the spot price of an unoccupied apartment would go up faster if we build market rate apartments?”

            I don’t think Phil has an opinion on the rate of change of the spot price of existing apartments. I think he has an opinion on the time evolution of the distribution of prices of occupied rental units, and I don’t think a single economist on this thread has actually engaged that.

  33. Ira says:

    In addition to the other problems with general sloppy thinking in this article (which others have pointed out) Manhattan is *not* “50% denser than San Francisco.” Manhattan is 400% denser than San Francisco. New York City (all 5 boroughs combined) is 50% denser than San Francisco, but it’s dragged down by Staten Island, which is both large and much lower density than the rest of the city. Brooklyn is twice as dense as SF, and Paris is three times as dense. There’s a lot of room to add more people without turning into Manhattan.

    • Phil says:

      Yikes! You are absolutely right. I did a Google search for “population density manhattan” and even though the giant text said 66940 people per square mile, what I saw was the lead sentence, which said “the density of the city is 26,430 per square mile” or something like that.

      I will go back and edit the article, I don’t want to spread fake facts.

  34. Sam says:

    >Of course, by the so-called “law of supply and demand”, building more housing does make housing cheaper. It’s easy to see why: those people with their billion dollars of disposable income are adding a lot economic activity in San Francisco, but they’re decreasing the economic activity in the cities they’re leaving, which no longer need so many waiters and barbers and shopkeepers. There is a cascade: some people move from Berkeley and Oakland to San Francisco, which allows replacements to move from Richmond and El Cerrito into Berkeley and Oakland, and so on. Ultimately, rents in San Francisco go up, and rents in some outlying communities go down. Yes, the increased supply of housing lead to decreased housing prices on average but they’ve gone up, not down, in San Francisco itself.

    This reminds me of a point that Nick Rowe made last year about a potential upward sloping demand curve for housing:

    If we imagine density is an amenity people are willing to pay for, then increasing supply could make that supply even more desirable. Of course, this is partial and ignores any changes in other prices or wages in the city. This is also purely theoretical and while casual empiricism suggests it holds in a very general sense (big cities have higher rents than small cities, even if the regulatory regime is not hugely different), the empirical evidence commentators (not the original poster, sadly) have marshaled suggests that in the short-run, increased supply is met with rents not rising as much.

    • Phil says:

      Thanks so much for posting that Nick Rowe link! I am going to point a few people towards it.

      • Sam says:

        Please do, but also note this from J.W. Mason (who agrees with Rowe) in the comments:
        >We should be clear that this is not an argument against allowing higher density. We absolutely should see zoning and other regulations changed to make it easier to allow housing (and mixed use) development at higher densities. But that’s because the benefits that people are paying for are real. There are good reasons that people from poorer areas all over the world choose to pay vastly higher rents in places like New York City. We should allow more people to make that choice. But we shouldn’t expect to get lower housing costs that way — that will require some form of subsidy, direct or indirect.

        That is, additional density is not a negative externality: rents are higher because living in a denser city is actually more desirable (so goes the theory, anyway).

        • Phil says:

          Oh sure, I’m not even sure where I stand on density myself. I love love love living at medium-high density (by which I mean about 10,000 people per square mile, a density at which every person in the US could live in Virginia with room to spare) and I think people who want to live in low-density suburbs like the one I grew up in are semi-nuts in addition to having too high an environmental impact. But I also wouldn’t want to live in Manhattan, although it really is a great place to visit.

          Different people like different things.

          One thing that I thought about mentioning in my post, but thought was too far off-topic, is that I think a lot of people who want to live in San Francisco really want to live in a city like San Francisco is now. In the sense that “you can never step twice in the same river”, a San Francisco that had 30% or 50% or 80% more people in it would not be the same San Francisco. Some people would prefer it — Andrew loves Manhattan, and presumably would prefer a San Francisco that is denser than the present one, compared to San Francisco as it exists now (I don’t know that this is true, I’m just speculating). Other people would not prefer it. On average, I would guess that people who already live in San Francisco would prefer that it not grow; after all, one of the reasons they live there is because they like it the way it is. And on average, I would guess that people who don’t live in San Francisco, but would like to, either would prefer higher density or at least are willing to tolerate it.

          • Steve Sailer says:

            Right. A lot of people like San Francisco the way it was.

            By the way, Toronto now has a giant high-rise suburb called Mississauga out by the airport. It has almost as many people as San Francisco. It seems okay, but I never hear anybody rhapsodizing about it the way they do about San Francisco.

            • Martha (Smith) says:

              San Francisco and Mississauga have very different histories: San Francisco was founded in 1776, became the largest West Coast city in the 1849 Gold Rush, was largely destroyed in the 1906 earthquake and fire (when it had a population of around 400,000), rebuilt in time for the Panama-Pacific Exposition in 1915 (and has since at least then been a major tourist attraction), and became a counterculture center in the fifties and sixties. So there’s lots about San Francisco to rhapsodize about, and lots of people who lived there, or visited, or heard exotic stories about it, to rhapsodize.

              Toronto Township (the precursor of Mississauga) was formed in 1805. Its population was around 100,000 in 1968, when the town of Mississauga was created. Mississauga’s population around 1970 was about 170,000. It was incorporated as a city in 1974, with some additional towns included. There are undoubtedly some people from the original communities encompassed by Mississauga who rhapsodize about the good old days, but there is no overall exotic history like San Fransisco’s, and there are fewer people involved than those who new about San Fransisco historically.

  35. Anon— says:

    Would have been nice to engage with any empirical evidence on this topic. Since there is a bunch and I’m not familiar with any that supports the model in this post.

    As others have commented, this seems really consistent with the stereotype of an arrogant physicist who proposes a simple model for a phenomenon while totally unaware of many of the characteristics of the data that this model is completely inconsistent with.

    • Phil says:

      Actually, the model I have proposed is more complicated than the model most of the YIMBYs and BARFs use, or at least what I have seen. Usually I see an appeal to the classic supply-demand curve: if supply goes up, price will down. My model is more complicated, in that I have at least a two-compartment model (San Francisco, and Outside San Francisco), as well as attempting to take into account the fact that the demand for workers in San Francisco depends on the income of the people who live there.

      So, arrogant physicist, fine. “Simple model”, well, compared to what? At least my model is more realistic than some that are often cited.

      • wow says:

        Phil, you really need to admit you’re out of your element here and move on. You don’t even have a real model. You have some sentences. That’s it. That might suffice for the quarterly journal of austrian economics, but you’re not fooling anyone else.

  36. Rob says:

    This is a misunderstanding of how supply/demand works for the reasons given by economists above.

    If you build more dwellings in SF, SF will become more affordable – and we’ll know, because more people will in fact be affording it (unless all the new houses somehow lie empty).

    • sonja trauss says:

      Yes, Nice point Rob. You might have to explain it again for everyone in the back, but beautifully put.

      “If you build more dwellings in SF, SF will become more affordable, _because more people will in fact be affording it_.”

      Here’s another sentence, that if you meditate on it, will make you smarter:

      “What’s the difference between being able to afford something that isn’t available, and not being able to afford something that is available?”

      • Phil says:

        “If you build more dwellings in SF, SF will become more affordable, _because more people will in fact be affording it_.”

        If you think that’s a good point, you’re going to love this one: SF is already more affordable than it has ever been, because more people are affording it!

        Problem solved!

        • NatashaRostova says:

          It’s more affordable than the counterfactual where less supply was built. I’d argue it makes sense to factor out exogenous growth (obviously there is a second order link between, well, everything, but I think we can somewhat safely assume it won’t overwhelm the first orders)

        • Sonja trauss says:

          No. problem not ‘solved.’ Rather, improvement is being made, more improvement is better.

          Do you really not know the difference between being on your way somewhere and being already there?

          Do you travel to Sacramento, but stop in El Cerrito because “I’m on my way! So I must already be there!”

  37. Sam says:

    Voters don’t have to be ‘rational’ since p(building | vote yes) = p(building | vote no) for any individual, he’ll vote with his heart not his head.

  38. jack says:

    oh jesus what an economically illiterate post, Phil. you could have spared yourself from embarrassment.

    • Phil says:

      The funny thing is, I definitely recognize and acknowledge the phenomenon of ‘physics arrogance.’ I try not to fall victim to it. Maybe not always successfully.

      But there is a parody-worthy aspect of classical statistics, too, in which perfectly-informed rational actors trade in frictionless markets, etc, etc.

      • Dan says:

        Your original post is basically a parody. You start with a laughably facile ‘explanation’ of the way things obviously work (with no reference to existing research), and condescendingly proceed to demean the character of anyone who doesn’t agree with your completely uninformed logic. This post is a travesty.

  39. Soho says:

    Andrew, this kind of ignorance is unworthy of your blog. Anyone who doesn’t understand that the east bay shares a housing market with San Francisco just isn’t even trying.

    • Raf says:

      I don’t think the author of the piece is ignorant, just intentionally obtuse. You can tell by reading his responses in the comments. He reveals that what he really cares about is neighborhood character.

      So this is my new theory: Phil knows that building more housing in the Bay Area will make the region more accessible, but this would be bad for him, because it might change the character of his single-family neighborhood. He doesn’t want to be seen as selfishly slamming the door behind him, so he invents weird models that describe a world where building more housing would increase demand so much that prices would go up.

      I hope this is the answer, because if it isn’t, I’m still at a loss.

      I agree that it’s pretty rude to strawman someone’s stated arguments, and then declare the arguments so weak that they must be acting out of spite instead of in good faith. On the other hand, he started it.

      • Dan says:

        Best comment of the thread. It wish I could upvote it to the top of the comments section. Thank you.

      • Ian Monroe says:

        “Phil” is projecting I think. He wants to keep Those People out, maintain his privileged lifestyle – single family home, cops who will push the homeless back downtown, etc – while being Liberal and a Good Person. He has to do a lot of motivated reasoning to end up at this unresearched slop of a post.

      • Phil says:

        I may be obtuse, but I’m not deliberately obtuse!

        Of course I think building more housing in the Bay Area will make the region more accessible. I say that in Paragraph 5! To save you from having to scroll back all the way up, I’ll repeat it here:
        Of course, by the so-called “law of supply and demand”, building more housing does make housing cheaper. It’s easy to see why: those people with their billion dollars of disposable income are adding a lot economic activity in San Francisco, but they’re decreasing the economic activity in the cities they’re leaving, which no longer need so many waiters and barbers and shopkeepers. There is a cascade: some people move from Berkeley and Oakland to San Francisco, which allows replacements to move from Richmond and El Cerrito into Berkeley and Oakland, and so on. Ultimately, rents in San Francisco go up, and rents in some outlying communities go down. Yes, the increased supply of housing lead to decreased housing prices on average but they’ve gone up, not down, in San Francisco itself.

        I might be wrong, but I’m not wrong on purpose!

  40. tcheasdfjkl says:

    One more thing: you say that adding housing in Manhattan hasn’t made rent come down. But you’re not comparing the status quo to the counterfactual where Manhattan did not build new housing. There’s no way to directly empirically test this, but I contend that if Manhattan had built significantly less housing, it would now be significantly more expensive. What matters isn’t the absolute number of units built, it’s the increase in demand *compared to* the increase in supply.

  41. Rob says:

    I usually love this blog. This was literally the worst article ever posted on the blog.

Leave a Reply