July 2026 – J. W. Mason

Profits, Payouts and Equity Prices, Part 1

TLDR: The value of corporate equity relative to GDP is at a historical high. But this does not necessarily mean there’s a bubble: profits and shareholder payouts are also very high relative to historical values.

I first started thinking about economics thirty years ago, during the (first) tech boom.

This was the era of “irrational exuberance,” the phrase coined by the recently deceased Alan Greenspan and made famous by Robert Shiller. I wrote a review of Shiller’s book of that title for In These Times — one of my first published articles. My first paper in graduate school was a replication of a paper by Brad DeLong and Larry Summers¹, which argued that seemingly excessive stock valuations could be explained by rational investors extrapolating recent earnings growth into the future.

All of which is seeming at least a little bit relevant today.

Between the start of 1995 and the end of 1999, the price-earnings ratio for the S&P 500, as measured by Shiller, more than doubled, from 20 to nearly 44. Then over the next three years, it fell back nearly as far. Today, price-earnings ratios by the same metric are just shy of 40, not far from the peak of the first tech boom. Everything old is new again, it seems. (Except that, as Paul Krugman notes, people generally liked the products of the first internet companies.) So the obvious question is whether this is also a bubble — whether the second half of the late-1990s story will get a rerun as well.

There are many people out there with highly specialized expertise whose whole job is to think about stock valuations. I am not one of those people! If you are looking for investment advice, you have come to the wrong blog.

But I do think I have something to add.

Most of what makes financial-analyst jobs hard has to do with specific companies and specific markets. Things get easier when we are looking at the stock market as a whole. (And that is where my own background in heterodox macro is more likely to help.) When we are talking about corporate equity in the aggregate, rather than individual securities, some arithmetic comes into play that helpfully limits the space of possibilities.

One way to think of a share is that it gives a claim on future payments by the corporation that issued it. This is not all that a share is — as Arjun and I stress in Against Money, it’s important to keep sight of financial assets’ existence as concrete objects with their own specific properties, and not reduce them to simply a future cashflow. But certainly the cashflow it gives claim to is one very important property of a share.

Again, the value of share is not reducible to present value of expected (in either the statistical and/or psychological sense) future payments. But those should act as an anchor. Unless we have good reason to think there is a change on value market participants put on future payments, we should expect share prices to vary roughly in proportion to them. And even if we think valuation of shares can vary indefinitely with respect to payments they give claim to, it’s worth knowing how much of current share prices would have to be explained in these terms, and how much can be explained by variation in the payments.

In the rest of this post, I am going to look at the US nonfinancial corporate sector. This means excluding the 20-25% of corporate equity issued by financial businesses, and including closely-held corporations as well as those listed on public exchanges. This is mainly because that’s the universe for which it’s easiest to get consistent data. But I also think it’s a reasonable thing to be interested in substantively.

If we look at nonfinancial corporate equity over the 80 years since World War II, here is what we see:

The figure shows the total value of nonfinancial corporate equity,as a share of potential GDP. (I am skeptical of potential as a measure of actual economic potential, but here it is just functioning as a trend, to smooth out short-term changes in the denominator.) This, I would argue, is the most straightforward measure of the value of the stock market in broad social terms — how much claim wealth in this form gives on social labor and its products. This is also, of course, the metric used by Piketty.

As the figure shows, stock market value in that sense has had three long upswings. The first peaked in the late 1960s at 0.9, the second in 2000 at 1.6, and the third is ongoing, with the ratio currently at 2.3. (The data in this post is drawn from the Fed’s Financial Accounts, and goes through the first quarter of 2026.) Over the long run, there is a clear upward trend, especially over the past 15 years. (The dotted line shows the post-WWII average.) By this metric, the current stock market boom has now run well ahead of the late 1990s one.

How should we think about this?

Logically, the value of corporate equity relative to GDP must reflect a combination of four factors: value added in the corporate sector as a share of GDP; corporate profits as a share of their value added; payouts to shareholders as a percentage of profits; and the value placed by markets on each dollar of payouts.

In other words, if corporate stock is worth more relative to GDP, that can be either because more of GDP is now happening in the corporate sector; or because more of the income from that activity is going to profits; or because more of those profits are being paid out to shareholders (rather than retained in the firm); or because financial markets place a greater value on each dollar of payment. Or of course some combination of those.

We can write this as an accounting identity:

equity/GDP = value added/GDP * profits/value added * payouts/profits * equity/payouts

As with any accounting identity applied historically, it is useful insofar as it corresponds to (1) categories in the relevant data, and (2) distinct causal factors we believe to be at work.

For the first term, we are, again, using all nonfinancial corporate equity, which includes closely held as well as publicly-traded corporations, and the BEA’s estimate of potential GDP. (Both are in current dollars.) Value added is defined, as usual, as sales less the cost of non-labor inputs. Profits are after tax (and of course also after depreciation); conceptually, these are the funds potentially available for distribution to shareholders.²

Payouts I am defining as dividends less net new equity issues. That share repurchases are conceptually equivalent to dividends is not, I think, too controversial at this point (though it creates a lot of headaches). We are also adding shares retired through cash acquisitions, and subtracting shares issued whether in IPOs or otherwise. This might seem odd at the level of an individual firm, but at the aggregate level these other flows seem clearly equivalent to buybacks and dividends. If firm A buys up all the shares in firm B for cash, that is a payment from the corporate sector to shareholders, just as if firm A were buying back its own shares. Similarly, new shares issued are a reduction in the aggregate payments from the corporate sector to shareholders just as a reduction in dividend payments would be.

It might sound strange to define IPOs (which generally are quite exciting for stock market participants) as equivalent to dividend reductions (which generally are not.) But this is where the aggregate perspective matters. Shareholders as a whole already own all the equity of the corporate sector as a whole. An IPO is a payment from shareholders to the corporate sector, exactly like buyback is a payment from the corporate sector to shareholders, without in either case any change in aggregate ownership rights. Or looking at it from another point of view, new corporations are in general competing with existing firms; the profits flowing out to shareholders of the new firm are to a first approximation deductions from the profits flowing to claimants on existing firms. Nice for the shareholders in the new firm, if it succeeds; but no use to shareholders as a class.

Finally, the valuation term asks, in effect, what is the market price of a dollar of income from the corporate sector. It’s analogous to the price-earnings or price-dividend ratios one sees at the level of individual corporations or indexes, though not identical given the nonstandard (but, I would argue, appropriate) way I have defined payouts. Here it also functions as the residual term, reflecting any change in the value of equity not explained by the other factors.

The figures below show the values of each of these terms over the past 80 years. What do we see?

We will start with the first term, corporate value added as a share of GDP. This shows how much of economic activity takes place in the corporate sector, and is potentially available for shareholders.

As it turns out, the corporate value added term does not do anything interesting. Yes, it is modestly lower (around 50 percent) after 2000 than its average in the earlier decades (53 percent), suggesting that all else equal, we might expect the value of corporate equity to be slightly lower relative to GDP in the 21st than in the 20th century. But this change is very small compared with the movements in the other terms. This factor might be important if we were comparing the US to other countries, but it is not part of the story here.

Next, profits:

Profits as a share of value added shows much more variation, falling by half in the 1980s, then rising in this century, in two big jumps — one after 2000 and the second over the past five or so years. While the corporate share of GDP doesn’t vary by even 10 percent over the whole period, profits as a share of value added are fully three times greater today than they were for much of the 1980s.

The third term is payouts.

Payouts (as I’ve defined them) also show large variation, rising from a bit under 40 percent of profits in the early decades to over 80 percent in more recent ones. The timing here is a bit different — though there is plenty of short-term variation, the long-run shift happens in a single big jump in the early 1980s. (This was the topic of an essay in my dissertation, which I never managed to publish as an academic article but did turn into a report for the Roosevelt Institute.) This term gets relatively little attention in discussion of stock prices, but it seems to me that it is as fundamental as profits to any discussion of long-term trends in the value of corporate equity.

Finally, the valuation term shows a lot of short- and medium-term variation but, perhaps surprisingly, no long run trend. Today’s ratio of around 40 is close to what we see in the 1950s and 1960s.

Again, what we are measuring with this last term is the ratio of equity value to shareholder payouts, including net share repurchases. The big spikes in the early 1970s and in 2000 are because those years saw exceptionally high new equity issues, which means very low payouts by my metric, and therefore very high ratios of equity value to payouts.

It is more common to talk about equity in relation to earnings, on the implicit assumption that profits are of equal value to shareholders whether they are paid out or not. I’ve shown this latter ratio below. But personally, I do not think that that is a good assumption. Shareholders evidently care a great deal about payouts — why else would companies pay dividends and make share repurchases? I think it is important to distinguish between corporations and the shareholders who exercise claims on them — the former are not simply the personal property of the latter. From this point of view, it is more natural to talk about valuation in terms of the price shareholders place on the income they actually get from corporations, as opposed to the underlying profits.

All of these series (except the last one) are combined in the next figure, which is really the whole point of this post. If you take one thing from one I’ve written here, this picture is it.

Equity value relative to GDP and its components, 1947-2026:

For this figure, I’ve converted the values to logs. This has the big advantage of converting the multiplicative relationship to an additive one, so that we can visually see the contribution made by each of them. But it can make interpreting the figure a bit tricky. Here, zero is the average value over the full period; positive one is a value about 2.7 times greater than the average, while negative one is a value about one-third of the average. The black line similarly describes the deviation of the equity-GDP ratio from its full-period average; the heights of the bars correspond to the contribution each term makes to that deviation. The data is quarterly; for all the terms except equity, I use rolling one-year averages.

As we can see, the log of the equity-GDP ratio is currently about 1.1 above its long-run average, corresponding to a value nearly three times greater. (2.2 today, versus a long run average of 0.85.) Just over half of this (0.53) is explained by higher profits relative to value added, 0.19 is explained by higher payouts relative to profits, and 0.36 is explained by the valuation term.

So already we can see that a simple explanation of today’s high equity values is going to be incomplete. Relative to the long-run average, we have three distinct factors each of which explains a significant share of today’s higher values.

Another thing that jumps out from the figure is that the previous historical peaks in equity values reflect quite different mixes of these components.

In the 1960s, profits as a share of value added were, for a while, well above average, though not as high as today; but the fraction of those profits flowing out to shareholders was much lower. Thus the much lower ratio of equity to GDP, despite comparable valuation ratios.

In the late 1990s, profits as a share of value added were much lower — less than 5 percent at the height of the tech bubble, compared with 10 percent in the 1960s and 15 percent today. But the fraction of profits paid out to shareholders was historically high, averaging over 100 percent for the 1998-2000 period. It’s worth noting in this context, also, that the collapse of equity value in the 1970s reflected a fall in shareholder payouts much more than in profitability; this is perhaps important context for the shareholder revolt that followed.

The big takeaway from this decomposition is that we should be cautious about assuming the stock market is overvalued — that we’re in a bubble, that this is another bout of irrational exuberance — simply because equity prices are high relative to the historical norm. Shareholders have it better than the historical norm, too. Corporations are more profitable. And more of those profits are flowing out to them. A bit of exuberance might be rational, under the circumstances.

On the other hand: If we focus on just the past 20 years, as in the figure below, the picture looks a bit different.

Yes, both profits and payouts are high relative to their long-run averages; but those shifts mostly came earlier, while the big rise in equity prices is more recent. Apart from the relatively brief collapse in profits during the Great Recession, almost all the variation in equity prices over the past two decades comes from the valuation term, rather than changes in the underlying payments to shareholders.

This is even more true over the past two years — equity values have increased sharply while profits have been stable and aggregate payments to shareholders have fallen, as dividend growth has stalled and net equity issue has turned positive. As a share of GDP, the net payments flowing from corporations to shareholders today are very close to where they were a decade ago; but corporate equity is worth 60 percent more. It’s hard to avoid the conclusion that either equity was undervalued in the mid-2010s, or it is overvalued now.

So which side do we focus on? Over the long run, most — tho not all — of the increase in the value of wealth in the form of corporate equity, is explained by what we might call fundamentals — the flow of payments to owners of that wealth. Over the short to medium run, on the other hand, almost all of the increase in the value of equity comes from valuation, and whatever financial-market dynamics drive that. Or as the old saying goes, in the short run the market is a voting machine, but in the long run it’s a weighing machine.

I want to say a bit more about the profits and payouts parts of the picture.

That high stock prices reflect to some extent a high level of corporate profits seems to be reasonably well understood, at least based on recent coverage in the Financial Times. This of course does not mean that high stock prices are justified, or sustainable; it just shifts the question to how justified or sustainable the high profits are.

This goes double to the extent that high valuations are based on an expectation of further increases in profits, as this recent FT piece suggests:

Wall Street’s expectations for company profit growth are rising at the fastest pace since the post-pandemic rebound, fuelling concern that an “earnings bubble” could be forming in the estimates that have underpinned the US stock market’s rally.

Analysts are now forecasting a 25 per cent increase in S&P 500 company earnings for the coming year, according to Bloomberg data, boosted by a resilient US economy and the AI boom.

However, just ahead of the second-quarter earnings season, some investors are growing concerned about the speed at which analysts’ estimates are rising…

This Alphaville piece goes further, saying that “supernormal profits are unsustainable, because they always are.” I don’t know about that. I don’t know if there’s any reason to think the profit share is stationary, to use the statistics jargon — apart from a dip in 2008-2009, profits as a share of value added have been greater than their long-run average in every year of this century, and seem to be getting farther from it. Capital really has won some lasting victories in the class war.

That is one natural way to look at profits — as a distributional variable. But there’s another way of looking at them, from the demand side.

We know, as readers of Keynes, that an increase in investment automatically creates an equal quantity of additional saving. If, furthermore, there’s little or no incremental saving out of wage income (a reasonable assumption, in my opinion) and if the fiscal balance and trade balance don’t change significantly (perhaps less reasonable, but we’ll go with it) then this additional saving must take the form of an increase in profits. This relationship is often known as the Kalecki-Levy profits identity, and is one bit of heterodox economics that has established a foothold in finance and the business press. The same identity says that an increase in the fiscal deficit or trade surplus should similarly lead to an equal increase in aggregate profits.

Exploring the math of this and the extent to which it is a reasonable first approximation of real-world dynamics would be an interesting exercise for another post. But it raises another point which I think is very relevant for thinking about the current situation: Even if the AI companies themselves are not particularly (or at all) profitable, AI-related investment spending is probably an important factor in raising aggregate profits. Just like the California gold rush generated plenty of profits for somebody, even if the vast majority of prospectors themselves went broke.

Or as this recent FT piece puts it:

The AI boom is lifting the fortunes of hundreds of formerly drab industrial, utility and mining companies as investors turn to the “picks and shovels” needed to build and power vast data centres. …

The companies benefiting include Caterpillar, best known for construction equipment but now supplying generators for data centres, 150-year-old German engineering company Hochtief, which will enter the Dax later this month, and Nucor, a steel supplier that has credited “white hot” AI demand for a “tsunami of earnings power”. …

The vast amounts of electricity needed for AI training are also fuelling demand for specialised power management, high-voltage electronics and cooling technologies. This has led to big interest in traditional suppliers of electrical equipment…

You could think of it like this: As long as there is strong investment demand and easy financing for it, the profits will be there …. but not necessarily for the companies carrying out the investment and getting the financing.

And this, perhaps, is the point where the macro perspective needs to give way to the micro one. Because it may be that, yes, in the aggregate, an ease in financing brings forth additional investment, which generates enough profits to justify the initial financing. But debt must be paid back not in the aggregate, but by the specific companies that incurred it. If the investment is one place and the profits are somewhere else, then at some point somebody’s survival constraint is going to be violated.

And I think I will end this post here.

I very much want to discuss the payouts piece of the equation, which in my mind is as important as profits, and much less discussed. But this post is already too long, and has taken much too long to write. So the payouts piece should be along, well, if not this month, then next month, or soon.

The Case Against Hard Rules for City Reserve Funds

The following is a somewhat expanded version of testimony I presented on June 23 before the New York City Commission on Government Efficiency.

My name is Josh Mason. I am an associate professor and chair of the economics department at John Jay College, CUNY, and a senior fellow at the Groundwork Collaborative. It’s a pleasure to address the Committee on Government Efficiency, several of whose members I worked with back in my days as the Policy Director of the New York Working Families Party.

I am here today to speak in opposition to any measure to create stricter rules for the use of City reserve funds, or to enshrine limits on reserve-fund withdrawals in the New York City Charter. I believe that the City needs greater short-term flexibility in budgeting, not less.

Economics suggests two broad principles for thinking about the City budget position.

First, over the long run, growth city expenditure needs to match growth in revenue. Unlike the federal government, the city cannot run deficits indefinitely, nor can it use long-run debt to fund current expenditure.

Second, over the short run, adjustments in response to unexpected shocks to revenue or program costs should fall on those areas of spending with the greatest intertemporal substitutability. In other words, when faced with a mismatch between current revenues and current expenditure, the adjustment required in order to bring them into balance should as much as possible fall on those budget items for which a dollar of spending next year is a close substitute for a dollar of spending this year.

The first of these principles, presumably, is accepted by everyone here. The second one is less familiar. But it is also implicitly accepted by everyone when it comes to periods of a year or less.

New York City, like many governments has very large short-term fluctuations in revenue. Between quarters, the average change in tax receipts is 13 percent; the average change in total receipts is 10 percent. It is not unusual to see total City revenues fall by 10 percent from one quarter to the next, or to see tax revenues fall by as much as 15 percent over a quarter, as they did between the first and second quarters of this fiscal year. ³ (See Figure 1.)

Figure 1. Source: New York City Comptroller, New York City Quarterly Cash Report; and author’s analysis

No one believes that short-term variation in the timing of city receipts should lead to city departments cutting (or increasing) spending by 10 or 15 percent, simply because relatively little tax revenue comes in the second quarter compared with the first. Everyone, I think, agrees that these short-term fluctuations should be entirely absorbed on the city’s balance sheet via short-term borrowing or changes in the city’s cash holdings.

That is not controversial. But what I would add is that, economically, there is no sharp line separating periods of less than a year from periods of more than a year. The same logic that tells us that variations in revenue or program costs over the course of the year should be entirely absorbed on the balance sheet, suggests that variation over a period of few years should also be primarily absorbed in this way.

It is true that the timing of tax revenue means there are greater fluctuations in revenue from quarter to quarter than from year to year. But the city still faces substantial variation in revenue from year to year, much of which is temporary. In recent years, we’ve seen city revenues increase by over 10 percent in some years, by as little as 1.5 percent in other years. It is far preferable to see spending rise steadily at the average rate of revenue growth, than to have big increases in spending in some years and cuts in real terms (which a 1.5 percent growth in spending would be) in other years, in an effort to achieve balance in each fiscal year. Variation in revenue from year to year is often temporary, and reverses the next year; and even if slower revenue growth turns out to be persistent, a gradual adjustment to the new situation is almost always preferable to an abrupt one.

Again, this principle is well understood at the level of practical budgeting. That is the reason that the city has reserve funds in the first place. And it is why, historically, the city has often used surpluses to prepay future years’ expenses rather to increase spending. ⁴

The items in the city budget that are most intertemporally elastic — most substitutable between one year and the next — are fund contributions. A dollar contributed to the fund next year is almost as good as a dollar contributed this year. ⁵ If we were to contribute nothing to a given fund this year, and double the contribution next year, the overall funding position would be almost the same. In general, if the total contributions over some period are unchanged, there is very little economic cost to shifting those contributions around in time.

This is much less true of other city expenditures. If we were to shutter the city’s libraries this year, and double library spending next year, the overall value of library services provided to the public would be far less than with a stable level of spending. Additional hours of libraries open next year are a very poor substitute for hours the libraries are closed this year. The same goes for fire and police services, education, and most other public services.

In principle, capital expenditures are more substitutable — a major road improvement, say, is almost as valuable if it is carried next year as this year. But in practice, the process by which projects are approved makes them hard to shift around in time — a project that has passed all the necessary hurdles to go forward in one year cannot necessarily be deferred to a later year or advanced to an earlier one. So in practice, the least costly way to address unexpected changes in City revenues or program costs is via contributions to or withdrawals from the city’s reserve funds — a category in which I would include the Retiree Health Benefit Trust and the Budget Stabilization Account as well as the Revenue Stabilization Fund and General Reserve.

The proposals to mandate contributions to the reserve funds and limit withdrawals from them would reduce this flexibility, and create greater instability in other categories of city spending. Perversely, they would force the burden of adjustment onto budget items that have less intertemporal substitutability. This is the opposite of what we should be trying to achieve. The budget needs more short-term flexibility, not less.

A number of these proposals involve formulas that are intended to allow flexibility when economic conditions warrant it, but not otherwise. For example, a recent proposal from the Comptroller’s office suggests that except in the event of natural disasters or similar catastrophic events, withdrawals from reserves should be permitted only once there have been two quarters of declining employment in the city. ⁶

Since the idea of tying withdrawals to economic conditions may seem appealing, I want to explain why it is not a workable solution in practice. There are four reasons, in my view, why hard rules based on economic data are not a practical solution.

First (as the Comptroller’s proposal acknowledges, but other similar proposals do not), reliable macroeconomic data is often unavailable in real time; most economic data is subject to substantial revisions which can dramatically change the initial numbers.

For local employment, the final data are not released until a full year after the period which it covers, and are often quite different from the initial data. For example, in 2025, the jobs data as initially released showed a respectable gain of 50,000 jobs over the year. But the numbers have been subsequently revised downward and the most recent numbers show no job growth over the year at all. (See figure 2.)

Figure 2. Source: Bureau of Labor Statistics, State and Metro Area Employment, Hours, and Earnings; and author’s analysis

This does not mean that we should not use the most current economic data, of course. But data whose final value is not available until a year after the fact, and where the initial release may be very different from the revised value, needs to be used cautiously and weighed alongside other evidence on the state of the economy. It is not a suitable basis for imposing hard rules on the city budget.

Second, while national economic data is available sooner than for local areas, these are also unsuitable for budget rules, since business cycle dynamics in New York City can be quite different from national dynamics. For example, the 1990 recession was quite mild at the national level — employment fell by only about 1 percent and had fully recovered within two years of the end of the recession. But in New York, it was much more severe, with fully 10 percent of jobs lost and employment not returning to pre-recession levels until a decade later. This was also the case for the 2000 recession. The 2007-2009 recession, on the other hand, was milder in New York City, with employment returning to pre-recession levels two years after the recession ended, compared with five years nationally. So a rule based on national economic data may be a poor fit for local conditions.

Figure 3. Source: Federal Reserve Bank of St. Louis

A rule based on recessions, which has also been suggested as a trigger for drawing down reserve funds, combines both of these problems. The National Bureau for Economic Research often does not announce recession turning points until a year or more after the fact, and the timing of downturns may be significantly different at the local and national levels.

Third, even if we had reliable data, economic indicators do not move in sync, and it is not always obvious which is the appropriate one to use.

For New York City, as for most local governments, the single most important source of revenue is the property tax, which in recent years accounts for between 40 and 50 percent of all City tax revenue. Property tax receipts depend on property values, and these can move quite differently from employment or output. For example, while the 2007-2009 recession was, as noted, fairly mild in New York in terms of employment, home prices saw a steep and lasting fall — average New York home prices were lower in 2017 than they had been a decade earlier in 2007. (See Figure 3.) Given the city’s reliance on property taxes, this is arguably more important than employment conditions. A rule based on employment would not necessarily give a good sense of the economic conditions that are most relevant for the city budget position.

Finally, in practice, data-based rules create arbitrary cutoffs and thresholds. The nature of rules is to impose hard binaries — either withdrawals from the reserve funds are permitted or they are not. But in practice, economic conditions may be quite similar in periods when the threshold is not quite reached as in periods when it is, and whatever indicator is used as the basis of a rule will, in reality, only be one of many pieces of information relevant to economic and budget conditions. Policymakers in the moment can weigh various considerations to decide whether it is appropriate to draw down or to add to reserves; a predefined rule does not allow this flexibility.

More generally, advocates of rules for city reserve funds need to grapple with the full implications of such rules. The city will, inevitably, face unforeseen changes in its revenues and in the cost of the services it provides. The impact of these changes must be absorbed somewhere in the budget. Given the city’s limited ability to control its revenue, especially in the short run, shocks that are not absorbed in the balance sheet will in general, be absorbed by changes to the level of city services provided. Ensuring a steady rate of contributions to the employee retiree health benefit fund sounds like a good thing, in isolation. But, obviously, stable contributions to the fund do nothing to reduce instability in city revenues or program costs. So a rule imposing a more stable path of contributions to the fund necessarily imposes more instability elsewhere in the city’s budget. And cutbacks to funding for the school system, or for public safety, will have persistent costs that cannot be made good in future years in the way that a shortfall in fund contributions can be.

A myopic focus on stabilizing contributions to city trust funds (which is of course desirable in isolation) can blind us to the very large costs of instability in the provision of public services. It is certainly true that the City, unlike the federal government and even more than the State, is constrained in its ability to issue debt, and cannot fund ongoing deficits through new borrowing. Nor, of course, can the city’s financial assets be spent down indefinitely. In this sense, it is absolutely correct that public expenditures must be managed so as to keep them in line with revenue growth over time. It is unfortunate, however, that the idea of responsibility has been narrowed to mean only a focus only on financial outcomes, and not on the no less critical responsibility for consistent provision of the public services that New York’s residents and businesses depend on.

Even short-term reductions in the provision of education, public safety, transportation, health and other services can have lasting effects. Among other things, public services are directly relevant to decisions by both families and businesses about whether to move to, or remain in, the City, and thus have important consequences for the City’s future tax base. When faced with a tradeoff between consistent contribution to city reserve funds and consistent provision of public services, the former has no better a priori claim to be considered the “responsible” course than the latter.

A related mistake, in my view, is the idea that policymakers will systematically err on the side of overspending unless restrained by hard budgetary rules. Both common sense and history suggest that while this sort of error certainly occurs, there is no reason to think it is any more common than the opposite error, of excessive resort to spending cuts to close budget gaps and insufficient use of balance-sheet flexibility.

There is no reason to assume policymakers will systematically err on the side of irresponsibly drawing down reserves; it is just as plausible that they will underutilize them. This is clearly the case at the state level, where the State made no drawdowns from the Tax Stabilization Fund or Rainy Day Reserve Fund in either the 2000 or 2007-2009 recessions despite substantial falls in tax revenue, instead resorting to other, more costly measures to close the state budget gap.⁷ In general, there is no reason to think that today’s policymakers, who would impose this rule, are any more likely to strike the right balance between the balance-sheet position and public service provision than the future policymakers who would be bound by it. The one thing we know for sure is that future policymakers will be better informed about future economic and budget conditions than we are today.

To be clear, I think the existence of city reserve funds is a very good thing. Given the constraints on city borrowing, adequate reserve funds are essential to maintaining stable provision of city services in the face of unexpected shocks. It is appropriate for the City to contribute more to these funds in years when revenues are usually high, while drawing them down in years when revenue growth is weaker. And it may well be that the ideal funding of city reserves is greater than it has been historically.

There is nothing wrong with thinking about guidelines or targets for reserve funds. What I urge you to reject, however, is enshrining a hard limit on the use of reserves in the City Charter. The goal of maintaining reserves should be to provide future administrations with greater flexibility, not less, in responding to future challenges.

No use was made of “AI” in preparing this post.

At Vox, a Conversation on Rent Control

(I had a long conversation yesterday with Eric Levitz of Vox about the New York City rent freeze and the economics of rent regulation. I have posted the interview below just as it appeared there, for my archives and in case people want to read it without dealing with the paywall.)

An economist makes the case for Zohran Mamdani’s rent freeze

A new look at an issue that frequently divides voter and experts.

by Eric Levitz July 7, 2026 at 6:00 AM EDT

gettyimages-2274493525.jpg.webp

As America’s housing crisis deepens, policymakers are increasingly turning to an old idea for improving affordability: making large rent increases illegal.

In recent years. Oregon, Washington, and California have enacted statewide rent controls. In 2024, the Biden administration floated a nationwide cap on rent increases for large buildings. And last month, New York’s Rent Guidelines Board approved a two-year rent freeze on the city’s roughly 1 million stabilized units, fulfilling one of Mayor Zohran Mamdani’s signature campaign promises.

While rent control has long had some appeal to voters, it has historically provoked consternation among economists. In one 2012 survey, just 2 percent of economists agreed with the statement that local rent regulations “had a positive impact over the past three decades on the amount and quality of broadly affordable rental housing.”

The reasoning behind such skepticism is simple: When you make it less profitable to provide rental housing, people produce less of it. As a result, rent control reduces the supply of housing — and thus tends to make cities less affordable in the long run.

But this orthodoxy may soon be overturned – or so argues J.W. Mason, chair of the economics department at New York’s John Jay College of Criminal Justice and a senior fellow at Groundwork Collaborative, a progressive think tank.

In Mason’s view, the evidence that rent regulations discourage construction has been widely overstated: When designed well — and paired with zoning reforms — rent controls can protect tenants from displacement without reducing the long-term supply of housing.

We spoke this week about the case for (and against) rent control in general and New York City’s policies in particular. Our conversation has been edited for clarity and concision.

Among mainstream economists, conventional wisdom holds that rent control measures are misguided, partly on the grounds that they reduce the long-term supply of housing. In your view, what does that analysis get wrong?

When we talk about rent regulation, we’re typically talking about markets where there are already very substantial constraints on housing supply. Nobody is trying to pass rent regulation in exurban Texas or the Atlanta suburbs, where you have a lot of new housing construction.

Where it’s relatively easy to build housing, rents are going to be closely tied to the cost of building and operating new housing because, if you charge a lot more than that, then you create an opportunity for competitors.

The markets where you have rent regulation are markets like New York City, San Francisco, and a lot of European cities — places where there’s already really hard constraints on the capacity to build new housing. And in cases like that, where supply is already constrained by land use rules or just by an absolute scarcity of buildable land or by other factors, you’re not going to get any additional limitation on supply from rent regulation.

In that context, owners of existing housing collect rents in the broad, economic sense — income that doesn’t derive from any contribution they’ve made to production, but merely from others’ inability to build. Under those conditions, the only thing rent regulation does is redistribute some of that economic rent from property owners to tenants.

Most critics of rent control oppose restrictive zoning too. So, I think they might say that we should focus on ending the conditions that allow landlords to extract economic rents in the first place, rather than on redistributing them.

There’s an argument that, if we could achieve deep supply-side improvements in housing, we wouldn’t need rent regulation to the extent that we currently do. And I think that’s a perfectly reasonable argument. But it does not do any good for tenants who are facing displacement today. The fact that you have a different long-term goal does not remove the need for dealing with the short-term problem.

And there are good reasons to think that rent regulation is desirable, even if we think the real problem is on the supply side. For one, I think the politics of dealing with supply issues are much easier if you also have rent regulation. A lot of opposition to addressing supply-side problems is a perception that if you get new development, then that’s going to lead to displacement of people in the areas where development is taking place.

We can debate how true that is. But it’s a very deeply held perception. So, to the extent that you can offer real security to existing tenants, you remove one of the big sources of public opposition to supply-side measures: You don’t need to oppose removing restrictions on new housing because you are locked in. You are safe. Your landlord cannot kick you out to get somebody higher-paying in.

And honestly, I think that politics is very clear here in New York. I think that you would not have gotten the City of Yes land-use reforms, or the zoning reforms that passed on the ballot initiatives this past year, or a progressive like Zohran Mamdani coming out in favor of supply-side measures to increase housing production, if we had not strengthened the rent laws back in 2019.

Putting the politics to one side, do you think there is any tension between restricting rents and increasing construction? Say a city rolls back some of its restrictions on homebuilding, and new construction stops being effectively capped by zoning rules. If that city adopts rent controls, will that reduce the supply of housing at the margin? Or is that supposed tradeoff entirely illusory, in your view?

There’s no deterrent effect to many rent regulations, including those in New York City. Obviously, you can hypothetically imagine a much more rigorous form of rent control that could discourage new construction. I’m not going to say that it is impossible for that to happen. I think that we’re just very far from that point.

This is largely because new construction is typically exempt from rent control. In New York City, you are only required to comply with rent regulations if your building is more than 50 years old. And developers aren’t deciding whether to build based on how much rent a project will yield 50 years in the future.

The longevity of housing just makes it different from other goods. People often say, “If you impose a hard cap on milk prices, people will find it less worthwhile to produce milk. And we’re going to have shortages of milk in the stores.” We can debate whether that’s always true. But it’s a reasonable argument, since milk is consumed shortly after it’s produced. So your decision to produce more milk really is based on the price that you can get for that milk today.

But housing is at the opposite extreme. The median building in New York is 80 years old. When that housing was first produced, the price it’s going for today was not a factor.

Now, I should add that in New York City, the buildings with the highest rates of rent regulation are actually newer buildings. But that’s because developers voluntarily opt into the rent regulation system, as a condition of getting tax subsidies. At that point, clearly you’re not having a negative effect on supply when this is a voluntary decision.

Is that necessarily true? In theory, the tax subsidies are supposed to encourage housing investment. And developers weigh the benefit of those subsidies against their costs: If you accept them, you need to provide some units at below-market rates. So, if New York City makes providing rent-stabilized units less profitable — by freezing rents — then don’t the subsidies become less valuable? And wouldn’t that theoretically make investors slightly less inclined to fund new housing, all else equal?

Well, we’re seeing more new housing constructed in New York City right now than we’ve seen in many decades. So clearly something is working. And maybe what’s working is just that incomes are rising, that demand for housing in the city is rising.

But in my opinion, the tax abatements are badly structured. I really would not support housing development that way. But a huge fraction of new housing that gets built in the city uses these tax credits. So I think clearly they’re attractive to developers. Clearly, it’s a worthwhile trade-off from their point of view.

For critics of rent control, one study looms especially large: In 2018, a team of Stanford economists examined the impact of San Francisco’s rent control expansion in the 1990s. And they found that the policy led to a 15 percent reduction in the rental housing supply, which pushed up rents in the city by 5.1 percent. But in my understanding, you think the implications of that research are widely misinterpreted.

Yeah. I think that’s really a study about poor regulatory design. What it shows is: If you impose strict rent regulations but you don’t restrict people’s ability to convert rental properties to other uses, that may encourage landlords to convert rental housing into condos.

In the study, rental housing supply did not fall because of a decline in new construction. It fell because of condo conversions. And that distinction is important. If you have rental housing that’s converted to condos, that’s not reducing the overall supply of housing, but only that of rental housing. And it’s not necessarily increasing the cost of housing: It may be increasing market rents in the unregulated sector, but decreasing the cost of condos for condo buyers.

In any case, a well-designed rent regulation, like New York City’s 2019 reforms, can simply disallow people from moving housing out of the rental market in that way.

Many have argued that rent stabilization in general and New York City’s 2019 reforms in particular have negatively impacted the quality of the housing stock. Specifically, the argument is that landlords respond to rent restrictions by cutting back on maintenance. Is that a serious risk?

Of all of the concerns that you’ve raised, that is the most legitimate. I don’t think that we’re really seeing that yet. If anything, we’re seeing a reduction in a number of units that seem to have severe maintenance problems in New York. But you know, some people think we should not just have a rent freeze here, but a rent rollback. So, you roll back rents by five, 10, 15, 20, 25 percent, you’ll eventually reach a point at which you have real problems with building owners not doing basic maintenance and buildings falling into disrepair.

I’m not sure what the number is. Clearly there is a number where that happens. But I think we’re a very long way away from that. The vast majority of buildings are renting for much more than their operating and maintenance costs. The Rent Guidelines Board does studies. They suggest that the median margin is on the order of 40 or 50 percent.

As you’ve written, there is a minority of stabilized buildings in which maintenance and operating costs already exceed their total rents. But you attribute that primarily to the poverty of such buildings’ tenants, rather than to excessive restrictions on market rents?

I think that’s generally the case. There’s a sector of nonprofit-owned buildings, which tend to be the ones with the lowest rents and the lowest-income tenants. And in many cases, those buildings do face real problems with maintenance and upkeep. But they’re often not increasing rents even by the regulated amount, since tenants in these places simply can’t afford to pay more. In those cases, I think at some point you need either targeted subsidies or a change in ownership. But this is a very small fringe of buildings.

To name one last criticism of rent control: Some economists argue that it promotes an inefficient allocation of housing. The argument being: If you let people pay a below-market rent — on the condition that they don’t move — then you’re encouraging them to stay in place. And this leads to things like, for example, empty-nesters continuing to occupy three-bedroom apartments, which would have more utility for younger families.

I think we should recognize that there is a legitimate social interest in saying: Somebody who’s lived in an apartment for 15 years has a right to remain there, even if their landlord decides they could get a higher income by renting to somebody else. I think that’s a perfectly reasonable social goal.

And I think doing that actually makes the housing market more flexible and efficient. Why? Because it means that there’s less pressure to become a homeowner in order to get that security. Right now, in most markets, if you want security of tenure, the only way to get it is through ownership.

And ownership really locks you in. The transaction costs from buying and selling a house are very large. And obviously, in many cases, you get a financial risk, since a house is your main form of savings. If you sell at the wrong time, you lose a lot of money. So we get people who are locked into houses. They don’t have the same degree of geographic mobility. They can’t move to where the job opportunities are better. They stay in a big house even after their children are grown, which would really be better used by a younger family. If we give more security of tenure to renters, more people will choose to rent, and we’ll have actually, I think, a more flexible and efficient housing market.

At John Jay, We Study Economics to Change the World

Last week, the Rent Guidelines Board voted for a freeze on the rent for New York City’s one million rent-regulated apartments, fulfilling one of Mayor Mamdani’s defining campaign promises.

There has been plenty of discussion of the decision, both supportive and critical. But there’s one aspect of it, of particular interest to me, that has not been mentioned: Two out of the mayor’s six appointees to the board are recent graduates of the John Jay MA program in economics, where I teach.

I’m very proud of Sina Sinai and Lauren Melodia, who I know carefully studied the evidence and considered the full range of options before voting for the freeze. Lauren is also doing important work as the Director of Economic and Fiscal Policy at the Center for New York City Affairs, where she is producing a great deal of valuable research, most recently on working conditions in childcare. She’s recently been joined by David Lee, another John Jay graduate, who formerly worked as Legislative Director for New York Assemblymember Ron Kim and is now writing about fiscal policy at the Center.

Meanwhile on the rent regulation front, Anisha Steephen, a current student at John Jay, just released a major report from the Roosevelt Institute on rent regulation as financial regulation, which I hope to be writing more about soon.

This is what students from the John Jay economics program do. For a small program that’s existed for less than ten years, we have an impressive number of students out in the world contributing to progressive political projects.

Also in the housing space, consider Paul Williams. After finishing his MA with us a few years ago, he established the Center for Public Enterprise, where he now has a dozen staff, and has done as much as anyone to make the case that local government can be a major investor in housing, as well as in energy and other areas. This is a critical part of the both-and approach — boost supply and protect tenants — that defines the Mamdani agenda on housing.

Other current and former John Jay MA students include policy staff for socialist elected officials like State Senator Julia Salazar and former Representative Jamaal Bowman; the legislative director for the UAW; the chief of staff for former New York City Councilmember Carlina Rivera and State Senator Kristen Gonzalez; and analysts and researchers at various government agencies, including several at the Bureau of labor Statistics. Journalists like Aída Chavez (of The Intercept and The Nation) and Kate Aronoff (of The New Republic, and author of A Planet to Win: Why We Need a Green New Deal) were also students here. Jack Gross, founder of the outstanding web journal Phenomenal World, and Nathan Tankus, of the essential newsletter Notes on the Crises, were also briefly students here. (Neither got degrees, but the work and the community matter more than the credential.)

Why am I sharing this? Is it just to brag? Well, partially. I am very proud of what we’ve done with this program over the past decade, and of the students who have passed through it. And to update Hillel, if you don’t talk about your own work, who will talk about it?

But there’s also a more specific and timely reason: For the next two weeks, we are still accepting applications for Fall 2026. And I suspect that readers of this blog must know a few young (or not so young) people interested in studying heterodox economics at a public university in New York City.

If you do know someone who might fit that description, here is the pitch.

Unlike most economics programs, John Jay is unapologetically committed to a progressive, policy-oriented approach, and to the heterodox traditions of Marxian, Keynesian and feminist economics. Our students and faculty see the study of economics both as an end in itself and as a way of contributing to the most pressing struggles in our society.

While many of ours students take up roles in politics, advocacy, journalism and policy research (like on the Rent Guidelines Board) many others continue on to PhD programs. In one recent year, we had an entering class of 15 and eight students who went on to PhD programs, a proportion I suspect very few other MA programs in the country could match, even at much more prestigious institutions.

John Jay College is located at 59th St. and 10th Ave., near Columbus Circle in the heart of Manhattan. All classes in the MA program meet in person one day a week in the evening. Most students take three classes per semester and finish the program in two years, but there is no penalty for going at a different pace.

For anyone who has lived in New York State for at least one year as of September, full-time tuition is $5,545 per semester. This is pro-rated for those taking fewer classes, so the total cost for the program is approximately $22,000 regardless of the time over which it is completed. (This is less than a quarter the tuition at many comparable programs.) For non-resident full time students tuition is somewhat higher, but still cheap compared with most graduate programs.

There’s an online application here. Only the statement of purpose and transcript is required by July 15; recommendation letters can come in later.

There is no requirement to have previously studied economics; our students come from a wide range of backgrounds and many have undergraduate degrees in the humanities, physical sciences or other fields. We are less interested in what classes people have taken than in their intellectual curiosity, a willingness to work hard, and a commitment to using economics training to help change the world.

Does coming to John Jay guarantee that you’ll play a leading role in building municipal socialism? Obviously not. But based on our track record, it does seem to improve the odds.