Industrial Policy: Further Thoughts

(Cross-posted from my Substack. If you like this blog, why not subscribe to that too?)

I just returned from Bangalore, where Arjun and I spent an intense 10 days working on our book, and on another project which I’ll be posting about in due time. I’d never been to India before, and it was … a lot. It took me a while to put my finger on the overarching impression: not chaos, or disorder, but incongruity — buildings and activities right on top of each other that, in an American context, you’d expect to be widely separated in space or time. That, and the constant buzz of activity, and crowds of people everywhere. In vibes, if not in specifics, it felt like a city of back-to-back Times Squares. I imagine that someone who grew up there would find an American city, even New York, rather dull.

It’s a city that’s gone from one million people barely a generation ago to 8 million today, and is still growing. There’s a modern subway, clean, reliable and packed, with the open-gangway cars New York is supposed to switch to eventually. It opened 15 years ago and now has over 60 stations — I wish we could build like that here. But the traffic is awesome and terrifying. Every imaginable vehicle — handpainted trucks, overloaded and dangling with tassels and streamers; modern cars; vans carrying sheep and goats; the ubiquitous three-wheeled, open-sided taxis; the even more ubiquitous motorbikes, sometimes carrying whole families; and of course the wandering cows — with no stoplights or other traffic control to speak of, and outside the old central city, no sidewalks either. Crossing the street is an adventure.

I realize that I am very far from the first person to have this reaction to an Indian city. Some years ago Jim Crotty was here for some kind of event, and the institution he was visiting provided him with a driver. Afterwards, he said that despite all the dodging and weaving through the packed roads he never felt anything but safe and comfortable. But, he added, “I would never get into a car with that guy in the United States. He’d be so bored, he’d probably fall asleep.”

Varieties of industrial policy. The panel I moderated on industrial policy is up on YouTube, though due to some video glitch it is missing my introductory comments. Jain Family Institute also produced a transcript of the event, which is here.

It was a very productive and conversation; I thought people really engaged with each other, and everyone had something interesting to contribute. But it left me a bit puzzled: How could people who share broad political principles, and don’t seem to disagree factually about the IRA, nonetheless arrive at such different judgements of it?

I wrote a rather long blog post trying to answer this question.

The conclusion I came to was that the reason Daniela Gabor (and other critics, though I was mostly thinking of Daniela when I wrote it) takes such a negative view of the IRA is that she focuses on the form of interface between the state and production it embodies: subsidies and incentives to private businesses. This approach accepts, indeed reinforces, the premise that the main vehicle for decarbonization is private investment. Which means that making this investment attractive to private business owners, for which profitability is a necessary but not sufficient condition. If you don’t think the question “how do we solve this urgent social problem” should be immediately translated into “how do we ensure that business can make money solving the problem,” then the IRA deserves criticism not just on the details but for its fundamental approach.

I am quite sympathetic to this argument. I don’t think anyone on the panel would disagree with it, either normatively as a matter of principle or descriptively as applied to the IRA. And yet the rest of us, to varying degrees, nonetheless take a more positive view of the IRA than Daniela does.

The argument of the post was that this is because we focus more on two other dimensions. First, the IRA’s subsidies are directed to capital expenditure itself, rather than financing; this already distinguishes it from what I had thought of as derisking. And second the IRA’s subsidies are directed toward narrowly specified activities (e.g. battery production) rather than to some generic category of green or sustainable investment, as a carbon tax would be. I called this last dimension “broad versus fine-grained targeting,” which is not the most elegant phrasing. Perhaps I would have done better to call it indicative versus imperative targeting, tho I suppose people might have objected to applying the latter term to a subsidy. In any case, if you think the central problem is the lack of coordination among private investment decisions, rather than private ownership s such, this dimension will look more important.

Extending the matrix. The post got a nice response; it seems like other people have been thinking along similar lines. Adam Tooze restated the argument more gracefully than I did:

Mason’s taxonomy focuses attention on two axes: how far is industrial policy driven by direct state engagement v. how far does it operate at arms-length through incentives? On the other hand, how far is green industrial policy broad-brush offering general financial incentives for green investment, as opposed to more fine-grained focus on key sectors and technologies?

Skeptics like Daniel Gabor, Mason suggests, can be seen as placing the focus on the form of policy action, prioritizing the question of direct versus indirect state action. Insofar as the IRA operates by way of tax incentives it remains within the existing, hands-off paradigm. A big green state would be far more directly involved. Those who see more promise in the IRA would not disagree with this judgment as to form but would insist that what makes the IRA different is that it engages in relatively fine-grained targeting of investment in key sectors.

My only quibble with this is that I don’t think it’s just two dimensions — to me, broad versus narrow and capital expenditure versus financing are two independent aspects of targeting.

I should stress that I wrote the post and the table to clarify the lines of disagreement on the panel, and in some similar discussions that I’ve been part of. They aren’t intended as a general classification of industrial policy, which — if it can be done at all — would require much more detailed knowledge of the range of IP experiences than I possess.

Tooze offers his own additional dimensions:

  • The relationship of economic policy to the underlying balance of class forces.
  • The mediation of those forces through the electoral system …
  • The agenda, expertise & de facto autonomy of state institutions…

These are certainly interesting and important questions. But it seems to me that they are perhaps questions for a historian rather than for a participant. They will offer a very useful framework for explaining, after the fact, why the debate over industrial policy turned out the way that it did. But if one is engaged in politics, one can’t treat the outcome one is aiming at as a fact to be explained. Advocacy in a political context presumes some degree of freedom at whatever decision point it is trying to influence. One wouldn’t want to take this too far: It’s silly to talk about what policies “should” be if there is no one capable of adopting them. But it seems to me that by participating in a political debate within a given community, you are accepting the premise, on some level, that the outcome depends on reason and not the balance of forces.

That said, Tooze’s third point, about state institutions, I think does work in an advocacy context, and adds something important to my schema. Though it’s not entirely obvious which way it cuts. Certainly a lack of state capacity — both administrative and fiscal — was an important motivation for the original derisking approach, and for neoliberalism more broadly. But as Beth Popp Berman reminds us, simple prohibitions and mandates are often easier to administer than incentives. And if the idea is to build up state capacity, rather than taking it as a fact, then that seems like an argument for public ownership.

I’ve thought for years that this was a badly neglected question in progressive economics. We have plenty of arguments for public goods — why the government should ensure that things are provided in different amounts or on different terms than a hypothetical market would. We don’t have so many arguments for why, and which, things should be provided by the public. The same goes for public ownership versus public provisions, with the latter entailing non-market criteria and intrinsic motivation, with the civil service protections that foster it.

The case for public provisioning. One group of people who are thinking about these questions seriously are Paul Williams and his team at  the Center for Public Enterprise. (Full disclosure: I sit on CPE’s board.) Paul wrote a blog post a couple weeks ago in response to some underinformed criticisms of public housing, on why public ownership is an important part of the housing picture. Looking at the problem from the point of view of the local government that are actually responsible for housing in the US, the problem looks a bit different than the perspective of national governments that I implicitly adopted in my post.

The first argument he makes for public ownership is that it economizes on what is often in practice the binding constraint on affordable housing, the fixed pot of federal subsidies. A public developer doesn’t need the substantial profit margin a private developer would expect; recovering its costs is enough. Public ownership also allows for, in my terms, more fine-grained targeting. A general program of subsidies or inclusionary zoning (like New York’s 421a tax credits) will be too lax in some cases, leaving affordable units on the table, and too stringent in others, deterring construction. A public developer can assess on a case by case basis the proportion and depth of affordable units that a given project can support. A third argument, not emphasized here but which Paul has made elsewhere, is that developing and operating public housing builds up the expertise within the public sector that is needed for any kind of transformative housing policy.

It’s telling but not surprising to see the but-this-one-goes-to-11 response to Paul’s post that all we need for more housing is land-use deregulation. Personally, I am quite sympathetic to the YIMBY position, and I know Paul is too. But it doesn’t help to oversell it. The problems of “not enough housing” and “not enough affordable housing” do overlap, but they are two distinct problems.

A somewhat different perspective on these questions comes from this report by Josh Wallack at Roosevelt, on universal childcare as industrial policy. Childcare doesn’t have some of the specific problems that industrial policy is often presented as the solution to – it doesn’t require specialized long-lived capital goods, or coordination across multiple industries. But, Wallack argues, it shares the essential element: We don’t think that demand on its own will call forth sufficient capacity, even with subsidies, so government has to intervene directly on the supply side, building up the new capacity itself. I’ve always thought that NYC’s universal pre-K was a great success story (both my kids benefited from it) that should be looked to as a model of how to expand the scope of the public sector. So I’m very glad to see this piece, which draws general lessons from the NYC experience. Wallack himself oversaw implementation of the program, so the report has a lot more detail on the specifics of implementation than you normally get. Very worth reading, if you’re at all interested in this topic.

One area where Wallack thinks the program could have done better is democratic participation in the planning process. This could be another dimension for thinking about industrial policy. A more political practice-oriented version of Tooze’s bullets would be to ask to what extent a particular program broadens or narrows the space for popular movements to shape policy. Of course the extent to which this is feasible, or even desirable, depends on the kind of production we’re talking about. In Catalyst, Matt Huber and Fred Stafford argue, persuasively in my view, that there is a tension between the need for larger-scale electricity transmission implied by the transition away from carbon, and the preference of some environmentalists for a more decentralized, locally-controlled energy system. I am less persuaded by their argument that the need for increased transmission and energy storage rule out a wholesale shift toward renewables; here as elsewhere, it seems to me, which obstacles you regard as insurmountable depend on where you want to end up.

The general point I would make is that politics is not about a final destination, but about a direction of travel. Whether or not we could have 100 percent renewable electricity — or 100 percent public ownership of housing, or whatever — is not so important. What matters is whether we could have substantially more than we have now.

On other topics.

Showing the inconsistencies between conservative free-market economics and actual conservative politics is, in my experience, much harder in practice than it seems like it ought to be, at least if you want to persuade people who actually hold one or both. So it’s fun to see Brian Callaci’s (excellent) arguments against non-compete agreements in ProMarket, the journal of the ur-Chicago Stigler Center.

Garbriel Zucman observes that the past few years have seen very large increases in the share of income at the very top, which now seems to have passed its gilded age peak.  Does this mean that I and others have been wrong to stress the gains for low-wage workers from tight post-pandemic labor markets? I don’t think so — both seem to be true. According to Realtime Inequality, the biggest income gains of the past two years have indeed gone to the top 1 percent and especially its top fractiles. But the next biggest gains have gone to the bottom half, which has outpaced the top 10 percent and comfortably outpaced the middle 40 percent. Their income numbers don’t further break out the bottom half, but given that the biggest wage gains have come a the very bottom, I suspect this picture would get even stronger if we looked further down the distribution.

This may well be a general pattern. The incomes that rise fastest in an economic boom are those that come from profits, on the one hand, and flexible wages that are strongly dependent on labor-market conditions on the other. People whose income comes from less commodified labor, with more socially embedded wage-setting, will be relatively insulated from swings in demand, downward but also upward. This may have something to do with the negative feeling about the economy among upper-middle class households that Emily Stewart writes about in Vox.

I’m still hoping to write something more at length about the debates around “greedflation” and price controls. But in the meantime, this from Servaas Storm is very good.

What I’ve been reading. On the plane to Bangalore, I finished Enzo Traverso’s Fire and Blood. I suppose it’s pretty common now to talk about the period from 1914 to 1945 as a unit, a second Thirty Years War. Traverso does this, but with the variation of approaching it as a European civil war — a war within a society along lines of class and ideology, rather than a war between states. A corollary of this, and arguably the animating spirit of the book, is the rehabilitation of anti-fascism as a positive political program. It’s a bit different from the kind of narrative history I usually read; the organization is thematic rather than chronological, and the focus is on culture — there are no tables and hardly any numbers, but plenty of reproductions of paintings. It reads more like a series of linked essays than a coherent whole, but what it lacks in overarching structure in makes up with endless fascinating particulars. I liked it very much.

 

“Earnings Shocks and Stabilization During COVID-19”

The other day, I put up a post arguing, on the basis of my analysis of the income data in the Current Population Survey, that the economic disruptions from the pandemic had not led to any reduction in real income for the lowest-income families. This is the opposite of the Great Recession, and presumably earlier recessions, where the biggest income losses were at the bottom. The difference, I suggested, was the much stronger fiscal response this time compared with previous downturns. 

My numbers were rough — tho I think informative — estimates based on a data set that is mainly intended for other purposes. Today I want to call attention to an important paper that reaches similar conclusions on the basis of far better data.

The paper is “Earnings Shocks and Stabilization During COVID-19” by Jeff Larrimore, Jacob Mortenson and David Splinter.1 If you’re following these debates, it’s a must-read.

The question they ask is slightly different from the one I did. Rather than look at the average change in income at each point in the distribution, they ask what fraction of workers experienced large declines in their incomes. Specifically they ask, for each point at the distribution of earnings in a given year, what fraction of workers had earnings at least 10 percent lower a year later? They include people whose earnings were zero in the second year (which means the results are not distorted by compositional effects), and do the exercise both with and without unemployment insurance and — for the most recent period — stimulus payments. They use individual tax records from the IRS, which means their sample is much larger and their data much more accurate than the usual survey-based sources.

What they find, first of all, is that earnings are quite volatile — more than 25 percent of workers experience a fall in earnings of 10 percent or more in a typical year, with a similar share experiencing a 10 percent or more increase. Looking at earnings alone, the fraction of workers experiencing large falls in income rose to about 30 percent in both 2009 and 2020; the fraction experiencing large increases fell somewhat in 2009, but not in 2019. See their Figure 1 below.

Turning to distribution, if we look at earnings alone, large falls were more concentrated at the bottom in 2020 than in 2009. This is shown in their Figure 2.  (Note that while the percentiles are based on earnings plus UI benefits, the  vertical axis shows the share with large falls in earnings alone.)  This pattern is consistent with the concentration of pandemic-related job losses in low-wage sectors. 

But when you add unemployment insurance in, the picture reverses. Now, across almost the whole lower half of the distribution, large falls in earnings were actually less common in 2020 than in 2019. And when you add in stimulus payments, it’s even more dramatic. Households in the bottom 20 percent of the distribution were barely half as likely to experience a larger fall in income in the crisis year of 2020 as in they were in the normal year of 2019.

The key results are summarized in their Table 1, below. It’s true that the proportion of low-wage households that experienced large falls in earnings during 2020 was greater than the proportion of high-wage households. But that’s true in every year — low incomes are just much more volatile than high ones. What’s different is how much the gap closed. Even counting the stimulus payments, households in the top fifth of earnings were somewhat more likely to experience a large fall in earnings in 2020 than in 2019. But in the bottom fifth, the share experiencing large falls in income fell from 43 percent to 27 percent. Nothing like this happened in 2009 — then, the frequency of large falls in income rose by the same amount (about 6 points) across the distribution. 

One thing this exercise confirms is that the more favorable experience low-income households in the pandemic downturn was entirely due to much stronger income-support programs. Earnings themselves fell even more disproportionately at the bottom than in the last recession. In the absence of the CARES Act, income inequality would have widened sharply rather than narrowed.

The one significant limitation of this study is that tax data is only released well after the end of the year it covers. So at this point, it can only tell us what happened in 2020, not in 2021. It’s hard to guess if this pattern will continue in 2021. (It might make a difference whether the child tax credit payments are counted.) But whether or not it does, doesn’t affect the results for 2020.

While the US experienced the most rapid fall in economic activity in history, low-wage workers experienced much less instability in their incomes than in a “good” year. This seems like a very important fact to me, one that should be getting much more attention than it is.

It didn’t have to turnout that way. In most economic crises, it very much doesn’t. People who are saying that the economy is over stimulated are implicitly saying that protecting low-wage workers from the crisis was a mistake. When the restaurant workers should have been left to fend for themselves. That way, they wouldn’t have any savings now  and wouldn’t be buying so much stuff. When production is severely curtailed, it’s impossible to maintain people’s incomes without creating excess demand somewhere else. But that’s a topic for another post. 

The point I want to make — and this is me speaking here, not the authors of the paper — is that the protection that working people enjoyed from big falls in income in 2020 should be the new benchmark for social insurance. Because the other thing that comes out clearly from these numbers is the utter inadequacy of the pre-pandemic safety net.  In 2019, only 9 percent of workers with large falls in earnings received UI benefits, and among those who did, the typical benefit was less than a third of their previous earnings. You can see the result of this in the table — for 2009 and 2019, the fraction of each group experiencing large  falls in earnings hardly changes when UI is included. Before 2020, there was essentially no insurance against large falls in earnings.

To be sure, the tax data doesn’t tell us how many of those with big falls in earnings lost their jobs and how many voluntarily quit. But the fact that someone leaves their job voluntarily doesn’t mean they shouldn’t be protected from the loss of income. Social Security is,  in a sense, a form of (much more robust) unemployment insurance for a major category of voluntary quits. The paid family and medical leave that, it seems, will not be in this year’s reconciliation bill but that Democrats still hope to pass, is another.

Back in the spring, people like Jason Furman were arguing that if we had a strong recovery in the labor market then we would no longer need the $400/week pandemic unemployment assistance. But this implicitly assumes that we didn’t need something like PUA already in 2019.

I’d like to hear Jason, or anyone, make a positive argument that before the pandemic, US workers enjoyed the right level of protection against job loss. In a good year in the US economy, 40 percent of low-wage workers experience a fall in earnings of 10 percent or more. Is that the right number? Is that getting us the socially optimal number of evictions and kids going to bed hungry? Is that what policy should be trying to get us back to? I’d like to hear why. 

A C-Shaped Recovery?

The coronavirus crisis has been different from normal recessions in many ways, but one of the most important is the scale of the macroeconomic response to it. 

Thanks to the stimulus payments, the pandemic unemployment insurance, the child tax credit, and a raft of other income support measures, this is the first recession in history in which household income actually rose rather than fell, and households ended up in a stronger financial position than before — with bankruptcies, for instance, running at half their pre-pandemic rates. It’s this that’s allowed spending to come back so quickly as the pandemic recedes. It wasn’t written in stone that the economic problem at the end of 2021 would be labor “shortages” and inflation, rather than double-digit unemployment and mass immiseration. The rising wave of hunger, homelessness and bankruptcies that people feared at the start of the pandemic hasn’t shown up. But that doesn’t mean that it couldn’t have. Without the stimulus measures of the past year and a half, it most likely would have. 

This extraordinary success story is the missing context for today’s macroeconomic debates. It’s somehow becoming conventional wisdom that the economy is “overstimulated,” as if the economic disruptions of the pandemic could have been managed some other way. As Claudia Sahm observed last week, the choice facing policymakerswas either to repeat the mistakes of the Great Recession or to go big. Fortunately, they went big.

The aggregate dimension of this story is familiar, even it’s sometimes forgotten these days. But I’ve seen much less discussion of the distributional side. Disposable income has held up overall, but what about for people at different income ranges?

For detailed statistics on this, we will have to wait for the American Community Survey produced by the Census. The ACS comes out annually; the first data from 2020 will be released in a month or so, and 2021 numbers will take another year. For real-time data we depend on the Current Population Survey, from the Bureau of Labor Statistics. This is the source for all the headline numbers on unemployment, wages and so on. 

The CPS is mainly focused on labor-market outcomes, but it does have one question about income: “What was the total combined income of all members of your family over the past 12 months?”1 The answer is given as one of 15 ranges, topping out at $150,000 or higher.

Compared with what we get from the ACS (or other more specialized surveys like the Survey of Consumer Finances or the Survey of Income and Program Participation) that’s not very much information. But it’s enough to get the big picture, and it has the major advantage of being available in close to real time. 

I have not seen anyone use the CPS to look at how real (inflation-adjusted) income changed across the distribution during the pandemic, compared with in the previous recession. So I decided to look at it myself. The results are shown in the figure nearby.

What I’ve done here is construct a household income measure by distributing households evenly within their buckets. Then I adjusted that income for inflation using the CPI. Then I compared family income at each point in the distribution in September 2021 — the most recent available — with September 2019, and then did the same thing for September 2009 and September 2007. I used the CPI for the inflation adjustment because the PCE index isn’t available yet for September.2 Using two-year periods ending in September seemed like the best way to make an apples-to-apples comparison and avoid seasonal effects.3 The idea is to see what happened to income across the distribution during the pandemic as compared to a similar time period during the Great Recession.

What you see here, for instance, is that a household at the 10th percentile — that is, whose income was higher than 10 percent of households and lower than 90 percent — had an income 4 percent higher in September 2021 than in September 2019. Over the 2007-2009 period, by contrast, real income at the 10th percentile fell by 8 percent. Real income the 80th percentile, on the other hand, fell by about 3 percent in both periods.4

As the figure makes clear, the difference between this recession and the previous one is not not just that disposable income fell last time but has been stable this time. The two crises saw very different patterns across income levels. The overall stability of personal income over the past two years is the result of substantial gains at the bottom combined with modest falls in the upper two-thirds. Whereas the fall in aggregate income during the Great Recession — as in most recessions — combines a much larger fall at the bottom with relative stability at the top. 

This seems to me like a very important and very under-appreciated fact about the past two years. This is not just the first recession in which household income didn’t fall. It’s the first recession — in modern times, if not ever — that hit higher income families harder than low-income ones. So far, it looks less like a K-shaped recovery than a C-shaped one.

Let’s look at it another way. Between December 2007 and December 2009 — the period of the Great Recession — the share of households who reported a total income under $30,000 rose from 26.3 percent to 28.6 percent. Incomes rose over the next decade, so that by December 2019, a similar roughly one-quarter share of households reported total income of under $35,000. But over the next two years, this share fell by almost two points, from 25.7 to 23.9 percent. The fraction reporting incomes under $30,000 fell from 20.5 to 18.8 percent, while the fraction reporting incomes under $20,000 fell from 16.3 percent to 14.6 percent. This suggests a substantial decline in the number of families facing serious material hardship. 

You might say: But real income did fall across most of the distribution. That is true.5 But think about it: We have just lived through a pandemic that, among other things, caused the most rapid fall in economic activity in US history. 20 million jobs disappeared overnight, and millions of them still have not come back. Of course income fell! What’s surprising is that it didn’t fall by more — that the short-term disruption was followed by a rapid bounce back rather than the long jobless recovery we’ve had after previous crises. What’s also a departure from previous downturns is whose incomes fell and whose didn’t.

Because the CPS income data is top-coded at $150,000 — about 15% of US households are above this — and the bucket below that is quite wide, the CPS isn’t informative about income at the top end. That’s why the figures cut off at the 80th percentile. I don’t see any obvious reason why high-income families should have had very different experiences in the two recessions, but we will have to wait for other data to be released to find out for sure.

There are certainly problems with measuring income with a single question. It’s not always clear what households are counting as income, especially at the low end where transfers make up a higher portion of the total. But it’s the same question in all four years. I find it hard to believe that the contrasting shifts in the numbers don’t reflect a genuine difference in the experience of low-income families over the two periods.

After all, this is consistent with what we know from other sources. Wage gains have been stronger at the bottom than at the top, by a growing margin. In the Household Pulse survey that the Census has been conducting regularly since the start of the pandemic, the dog that didn’t bark is the lack of any increase in most measures of material deprivation. In the most recent survey, for example, 9 percent of families reported that in the past week, they sometimes or often didn’t have enough to eat. That’s a shockingly high number — but it is a somewhat lower number than in April 2020. And of course, what’s all the talk about labor shortages but complaints — sometimes in so many words — that people no longer feel they have to accept underpaid drudge work out of sheer desperation?

Maintaining or improving access to necessities for the most vulnerable through an economic catastrophe is a major accomplishment. Yet what’s striking about the current moment is how little anyone is taking credit for it. 

Of course there are reasons why the focus is where it is. It’s easier to talk about the problems we are actually facing than the much worse crisis we didn’t have. (There ought to be a name for the fallacy where a timely response to head off some danger is retroactively treated as a sign there was no danger in the first place.) Conservatives obviously don’t want to acknowledge the success of a massive public spending program, especially when Democrats are in office (and don’t necessarily approve of making poor people less poor in the first place.) Progressives are more comfortable criticizing bailouts than celebrating economic success stories. (And of course there is plenty to criticize.) And with the Build Back Better agenda on the line, one might worry that talking about how the measures of the past year and a half have raised up the bottom will feed a dangerous complacency, a sense that we’ve done enough already.

As it happens, I’m not sure that last worry is justified. Back when I did political work, one of things that came though most clearly talking to organizers, and to people at doors myself, is that for most people the biggest obstacle to political engagement isn’t satisfaction with the way things are, but doubt that collective action can change them. Most people,I think, are quite aware that, as we used to say, “Shit is fucked up and bullshit.” What they lack is a sense of the connection of politics and policy with the concrete problems they face. Even among political professionals, I suspect, doubt that things can be very different is often a more powerful conservative force than a positive attachment to things as they are. Remembering how policymakers made the choice go big during the pandemic might, then, strengthen, rather than undermine, the case for going big today.

Be that as it may, if it is in fact the case that during a period when unemployment spiked to 15 percent, incomes at the bottom end actually rose, that seems like an important fact about the world that someone ought to be talking about.

 

UPDATE:

Some people have asked whether the apparent rise in incomes at the bottom might be due to changes in family size — maybe more people moved in together and pooled their income during the pandemic? To address that, here’s another version of the figure, this one showing the change in real income divided by household size.

As it turns out, average household size actually shrank slightly over 2019-2021. This was not the case in 2007-2009, so adjusting for household size makes the recent performance look a bit better relative to the previous one. But as you can see, the broad picture is essentially the same.

 

Inflation for Whom?

A point I’ve been emphasizing about inflation (see here and here) is that it is just an average of price changes; it doesn’t have any independent existence.

One implication of this is that there is not, even in principle, a true inflation rate. Pick any basket of goods and measure their prices over time; that is an inflation rate. The “all urban consumers” basket used by the BLS for the headline CPI inflation rate is a useful benchmark, but it’s just one basket among others. Any individual household or subgroup of households will have its own consumption basket and corresponding inflation rate.

Because a small number of items have gone up in price a lot recently, the average price increase in the CPI basket is greater than increase in wages over the past year. In this sense, real wages have gone down. I am not convinced this is a meaningful statistic. For one thing, car prices are almost certain to come back down over the next year, once the current semiconductor bottleneck is relieved and manufacturers ramp up output. Wage gains, on the other hand, have a lot of inertia. This year’s wage gains are likely to continue; certainly they will not be given back.

But there’s another reason the “falling real wage” claim is misleading. When price increases are concentrated in a few areas, the inflation rate facing people who are buying stuff in those areas will be very different from the rate facing those who are not. Most Americans do buy a car every few years, but relatively few need to buy a car right now.1 And even averaged over time, different groups of people spend more or less on cars relative to other things. The same goes for other categories of spending.

The BLS’s Consumer Expenditure Survey (CEX) tries to measure the distribution of consumption spending by different demographic groups. In principle, you could construct a separate CPI for each group, like CPI-E the BLS reports for elderly households. (For what it’s worth the CPI-E increased by 4.8 percent over the past year, a bit slower than the headline rate.) In practice the challenges in doing this are formidable — for the headline measure weights can be based on retail sales, but the weights for demographic group have to be based on household surveys, which are slower and much less reliable. (I have some discussion of these issues in Section 7 of this paper.) Still, the CEX can give us at least a rough sense of the difference in consumption baskets and inflation rates across different groups.

It’s particularly interesting to look at consumption baskets across income groups. One of the central arguments for running the economy hot is that it tends to compress wages. From this point of view, an increase in prices paid disproportionately by lower income households is more concerning than a similar aggregate increase in prices paid more by the better off.

For this post, I chose to focus on the consumption basket of households with pre-tax income below $30,000 a year — about one quarter of the population.

In the table below, I show 20 items, accounting for almost 95% of the CPI basket. The first column shows its share of the CPI-U basket, taken from the most recent CPI Table 2. The second column shows the difference between the weight of the item in consumption by households earning less than $30,000 and its weight in total consumption.2 So a positive value means something that makes up a larger share of consumption for households with incomes under $30,000 than of consumption for the population as a whole. This comes from the most recent Consumer Expenditure Survey, covering July 2019 through June 2020. The third column shows the price change of that item from July 2020 to July 2021, again from CPI Table 2. The items are ordered from the ones that make up the largest relative share of the consumption basket for low-income households to the ones that make up the smallest relative share. So it gives at least a rough sense of the different inflations experienced by lower versus higher income families.

Expenditure Category Overall share (CPI) Relative share, income <$30k (CEX) Inflation, July 2020-July 2021 (CPI)
Rent of primary residence 7.6 8.3 1.9
Food at home 7.6 2.4 2.6
Electricity 2.5 1.5 4
Medical care services 7.1 1 0.8
Medical care commodities 1.5 0.35 -2.1
Recreation commodities 2.0 0.35 3.2
Water and sewer and trash collection 1.1 0.3 3.7
Education and communication services 6.1 0.2 1.2
Motor fuel 3.8 0.2 41.6
Utility (piped) gas service 0.7 0.2 19
Apparel 2.7 0.2 4.2
Motor vehicle parts and equipment 0.4 0.05 4.3
Fuel oil and other fuels 0.2 0.05 30.9
New vehicles 3.7 -0.15 6.4
Transportation services 5.3 -0.15 6.4
Lodging away from home 1.0 -0.3 21.5
Used cars and trucks 3.5 -0.3 41.7
Alcoholic beverages 1.0 -0.3 2.4
Food away from home 6.2 -0.35 4.6
Recreation services 3.7 -0.6 3.7
Household furnishings and supplies 3.7 -0.7 3
Owners’ equivalent rent 22.4 n/a3 2.4

As you can see, the items that are increasing at less than 2 percent a year — highlighted in blue — are all things disproportionately consumed by lower-income households. Rent, in particular, makes up a much higher share of spending for low-income households. Rent growth slowed sharply during the pandemic and, unlike many other prices, it has not so far accelerated again. Rent growth over the past year is about half the average rate in the three years before the pandemic.

Medical goods and services also make up a larger share of spending for lower-income households; prices there have grown slowly or income cases actually fallen over the past year. Prescription drug prices, for example, fell by 2 percent over the past year. Finally, education services, including childcare, have pulled inflation down over the past year, rising by about 1 percent (college tuition was flat.) Education inflation has been slowing for a long time — a trend I don’t recall seeing discussed much — but it slowed even more during the pandemic. Education and childcare make up a slightly higher fraction of spending for low-income households than for others.

On the other side, almost all the sectors where inflation is notably high — highlighted in red — make up a larger share of spending for higher-income households. Lodging away from home, for example, where prices are up over 20 percent, makes up less than 1 percent of the consumption basket for households with incomes under $30,000, but 2.5 percent of the basket for households with incomes over $200,000. Transportation services, food away from home, and new and used cars, which account for  the majority of non-energy inflation, are also disproportionately consumed by higher income households.

In general, it seems clear that lower-income households are facing less inflation than higher income ones. The biggest price increases are in areas that are disproportionately consumed by higher-income families, while several of the most important consumption categories for lower-income families are seeing prices rise more slowly than before the pandemic. Any discussion of “falling real incomes” that ignores this fact is at best incomplete.

There is, of course, one big exception: energy. Gasoline especially, but also electricity and heating gas, are seeing big price increases and make up a larger share of consumption for lower-income families. And unlike auto purchases, energy consumption can’t be postponed. If you want to tell a story about higher prices eating up wage gains, it seems to me that energy is your best bet.

Except, of course, that these are prices that we want to see rise, if we are serious about climate change. Many of the same people fretting about inflation eroding real wages, are strong supporters of carbon taxes or permits. If you think a goal of policy is to raise the relative price of fossil fuels, why object when it happens via the market?

At the end of the day, perhaps the current debate about inflation and real wages doesn’t belong in the macroeconomics box at all, but in the climate box. The difficult problem here is not how to keep demand strong enough to raise wages without also raising prices. The price spikes we’re seeing right now are mainly about short-term supply constraints. I am confident that prices for autos and many other goods will  come back down or at least stabilize over the next year, even if demand remains strong. The really difficult problem is how we make the transition away from fossil fuels without unacceptably burdening the people who are currently dependent on them.

UPDATE: I am getting some very confused readers, who note that historically rent, education and health care have historically risen in price faster than most goods, while in this post I’m saying they are rising more slowly. The original post, should have, but did not, make clear that the pattern of price changes over the past year or so is quite different from what we are used to. That said, this is not all about the pandemic. As I did note, inflation in education has been slowing for a long time; health care inflation has fallen dramatically during the pandemic but was also slowing before that, arguably thanks to the ACA. But the key point is that I am not saying that poor people face lower inflation in general; I’m saying this is a distinct feature of the inflation we’re experiencing now.

The Class Struggle on Wall Street: A Footnote

Remember back at the beginning of February when the stock markets were all crashing? Feels like ages ago now, I know. Anyway, Seth Ackerman and I had an interesting conversation about it over at Jacobin.

My rather boring view is that short-term movements in stock markets can’t be explained by any kind of objective factors, because in the short run prices are dominated by conventional expectations — investors’ beliefs about investors’ beliefs… [1] But over longer periods, the value of shares is going to depend on the fraction of output claimed as profits and that, in general, is going to move inversely with the share claimed as wages. So if working people are getting raises — and they are, at least more than they were in 2010-2014 — then shareholders are right to worry about their own claim on the product.

One thing I say in the interview that a couple people have been surprised at, is that

there has been an upturn in business investment. In the corporate sector, at least, business investment, after being very weak for a number of years, is now near the high end of its historical range as a fraction of output.

Really, near the high end? Isn’t investment supposed to be weak?

As with a lot of things, whether investment is weak or strong depends on exactly what you measure. The figure below shows investment as a share of total output for the economy as a whole and for the nonfinancial corporate sector since 1960. The dotted lines show the 10th and 90th percentiles.

Gross capital formation as a percent of output

 

As you can see, while invesment for the economy as a whole is near the low end of its historic range, nonfinancial corporate investment is indeed near the high end.

What explains the difference? First, investment by households collapsed during the recession and has not significantly recovered since.  This includes purchases of new houses but also improvements of owner-occupied houses, and brokers’ fees and other transactions costs of home sales (that last item accounts for as much as a quarter of residential investment historically; many people don’t realize it’s counted at all). Second, the investment rate of noncorporate businesses is about half what it was in the 1970s and 80s. This second factor is exacerbated by the increased weight of noncorporate businesses relative to corproate businesses over the past 20 years. I’m not sure what concrete developments are being described by these last two changes, but mechanically, they explain a big part of the divergence in the figure above. Finally, the secular increase in the share of output produced by the public sector obviously implies a decline in the share of private investment in GDP.

I think that for the issues Seth and I were talking about, the corporate sector is the most relevant. It’s only there that we can more or less directly observe quantities corresponding to our concepts of “the economy.” In the public (and nonprofit) sector we can’t observe output, in the noncorproate sector we can’t observe profits and wages (they’re mixed up in proprietors income), and in the household sector we can’t observe either. And financial sector has its own issues.

Anyway, you should read the interview, it’s much more interesting than this digression. I just thought it was worth explaining that one line, which otherwise might provoke doubts.

 

[1] While this is a truism, it’s worth thinking through under what conditions this kind of random walk behavior applies. The asset needs to be and liquid and long-lived relative to the relevant investment horizon, and price changes over the investment horizon have to be much larger than income or holding costs. An asset that is normally held to maturity is never going to have these sort of price dynamics.

Saving and Borrowing: A Response to Klein

Matthew Klein has a characteristically thoughtful post disagreeing with my new paper on income distribution and debt. I think his post has some valid arguments, but also, from my point of view, some misunderstandings. In any case, this is the conversation we should be having.

I want to respond on the specific points Klein raises. But first, in this post, I want to clarify some background conceptual issues. In particular, I want to explain why I think it’s unhelpful to think about the issues of debt and demand in terms of saving.

Klein talks a great deal about saving in his post. Like most people writing on these issues, he treats the concepts of rising debt-income ratios, higher borrowing and lower saving as if they were interchangeable. In common parlance, the question “why have households borrowed more?” is equivalent to “why have households saved less?” And either way, the spending that raises debt and reduces saving, is also understood to contribute to aggregate demand.

This conception is laid out in Figure 1 below. These are accounting rather than causal relationships. A minus sign in the link means the relationship is negative.

 

We start with households’ decision to consume more or less out of their income. Implicitly, all household outlays are for consumption, or at least, this is the only flow of household spending that varies significantly. An additional dollar of household consumption spending means an additional dollar of demand for goods and services; it also means a dollar less of savings. A dollar less of savings equals a dollar more of borrowing. More borrowing obviously means higher debt, or — equivalently in this view — a higher debt-GDP ratio.

There’s nothing particularly orthodox or heterodox about this way of looking at things. You can hear the claim that a rise in the household debt-income ratio contributes more or less one for one to aggregate demand as easily from Paul Krugman as from Steve Keen. Similarly, the idea that a decline in savings rates is equivalent to an increase in borrowing is used by Marxists as well as by mainstream economists, not to mention eclectic business journalists like Klein. Of course no one actually says “we assume that household assets are fixed or nonexistent.” But implicitly that’s what you’re doing when you treat the question of what has happened to household borrowing as if it were the equivalent of what has happened to household saving.

There is nothing wrong, in principle, with thinking in terms of the logic of Figure 1, or constructing models on that basis. Social science is impossible without abstraction. It’s often useful, even necessary, to think through the implications of a small subset of the relationships between economic variables, while ignoring the rest. But when we turn to  the concrete historical changes in macroeconomic quantities like household debt and aggregate demand in the US, the ceteris paribus condition is no longer available. We can’t reason in terms of the hypothetical case where all else was equal. We have to take into account all the factors that actually did contribute to those changes.

This is one of the main points of the debt-inequality paper, and of my work with Arjun Jayadev on household debt. In reality, much of the historical variation in debt-income ratios and related variables cannot be explained in terms of the factors in Figure 1. You need something more like Figure 2.

Figure 2 shows a broader set of factors that we need to include in a historical account of household sector balances. I should emphasize, again, that this is not about cause and effect. The links shown in the diagram are accounting relationships. You cannot explain the outcomes at the bottom without the factors shown here. [1] I realize it looks like a lot of detail. But this is not complexity for complexity’s sake. All the links shown in Figure 2 are quantitatively important.

The dark black links are the same as in the previous diagram. It is still true that higher household consumption spending reduces saving and raises aggregate demand, and contributes to lower saving and higher borrowing, which in turn contributes to lower net wealth and an increase in the debt ratio. Note, though, that I’ve separated saving from balance sheet improvement. The economic saving used in the national accounts is quite different from the financial saving that results in changes in the household balance sheet.

In addition to the factors the debt-demand story of Figure 1 focuses on, we also have to consider: various actual and imputed payment flows that the national accounts attribute to the household sector, but which do not involve any money payments to or fro households (blue); the asset side of household balance sheets (gray); factors other than current spending that contribute to changes in debt-income ratios (red); and change in value of existing assets (cyan).

The blue factors are discussed in Section 5 of the debt-distribution paper. There is a much fuller discussion in a superb paper by Barry Cynamon and Steve Fazzari, which should be read by anyone who uses macroeconomic data on household income and consumption. Saving, remember, is defined as the difference between income and consumption. But as Cynamon and Fazzari point out, on the order of a quarter of both household income and consumption spending in the national accounts is accounted for by items that involve no actual money income or payments for households, and thus cannot affect household balance sheets.

These transactions include, first, payments by third parties for services used by households, mainly employer-paid premiums for health insurance and payments to healthcare providers by Medicaid and Medicare. These payments are counted as both income and consumption spending for households, exactly as if Medicare were a cash transfer program that recipients then chose to use to purchase healthcare. If we are interested in changes in household balance sheets, we must exclude these payments, since they do not involve any actual outlays by households; but they still do contribute to aggregate demand. Second, there are imputed purchases where no money really changes hands at all.  The most important of these are owners’ equivalent rent that homeowners are imputed to pay to themselves, and the imputed financial services that households are supposed to purchase (paid for with imputed interest income) when they hold bank deposits and similar assets paying less than the market interest rate. Like the third party payments, these imputed interest payments are counted as both income and expenditure for households. Owners’ equivalent rent is also added to household income, but net of mortgage interest, property taxes and maintenance costs. Finally, the national accounts treat the assets of pension and similar trust funds as if they were directly owned by households. This means that employer contributions and asset income for these funds are counted as household income (and therefore add to measured saving) while benefit payments are not.

These items make up a substantial part of household payments as recorded in the national accounts – Medicare, Medicaid and employer-paid health premiums together account for 14 percent of official household consumption; owners’ equivalent rent accounts for another 10 percent; and imputed financial services for 4 percent; while consolidating pension funds with households adds about 2 percent to household income (down from 5 percent in the 1980s). More importantly, the relative size of these components has changed substantially in the past generation, enough to substantially change the picture of household consumption and income.

Incidentally, Klein says I exclude all healthcare spending in my adjusted consumption series. This is a misunderstanding on his part. I exclude only third-party health care spending — healthcare spending by employers and the federal government. I’m not surprised he missed this point, given how counterintuitive it is that Medicare is counted as household consumption spending in the first place.

This is all shown in Figure 3 below (an improved version of the paper’s Figure 1):

The two dotted lines remove public and employer payments for healthcare, respectively, from household consumption. As you can see, the bulk of the reported increase in household consumption as a share of GDP is accounted for by healthcare spending by units other than households. The gray line then removes owners’ equivalent rent. The final, heavy black line removes imputed financial services, pension income net of benefits payments, and a few other, much smaller imputed items. What we are left with is monetary expenditure for consumption by households. The trend here is essentially flat since 1980; it is simply not the case that household consumption spending has increased as a share of GDP.

So Figure 3 is showing the contributions of the blue factors in Figure 2. Note that while these do not involve any monetary outlay by households and thus cannot affect household balance sheets or debt, they do all contribute to measured household saving.

The gray factors involve household assets. No one denies, in principle, that balance sheets have both an asset side and a liability side; but it’s striking how much this is ignored in practice, with net and gross measures used interchangeably. In the first place, we have to take into account residential investment. Purchase of new housing is considered investment, and does not reduce measured saving; but it does of course involve monetary outlay and affects household balance sheets just as consumption spending does. [2] We also have take into account net acquisition of financial assets. An increase in spending relative to income moves household balance sheets toward deficit; this may be accommodated by increased borrowing, but it can just as well be accommodated by lower net purchases of financial assets. In some cases, higher desired accumulation of financial asset can also be an autonomous factor requiring balance sheet adjustment. (This is probably more important for other sectors, especially state and local governments, than for households.) The fact that adjustment can take place on the asset as well as the liability side is another reason there is no necessary connection between saving and debt growth.

Net accumulation of financial assets affects household borrowing, but not saving or aggregate demand. Residential investment also does not reduce measured saving, but it does increase aggregate demand as well as borrowing. The red line in Figure 3 adds residential investment by households to adjusted consumption spending. Now we can see that household spending on goods and services did indeed increase during the housing bubble period – conventional wisdom is right on that point. But this was a  spike of limited duration, not the secular increase that the standard consumption figures suggest.

Again, this is not just an issue in principle; historical variation in net acquisition of assets by the household sector is comparable to variation in borrowing. The decline in observed savings rates in the 1980s, in particular, was much more reflected in slower acquisition of assets than faster growth of debt. And the sharp fall in saving immediately prior to the great recession in part reflects the decline in residential investment, which peaked in 2005 and fell rapidly thereafter.

The cyan item is capital gains, the other factor, along with net accumulation, in growth of assets and net wealth. For the debt-demand story this is not important. But in other contexts it is. As I pointed out in my Crooked Timber post on Piketty, the growth in capital relative to GDP in the US is entirely explained by capital gains on existing assets, not by the accumulation dynamics described by his formula “r > g”.

Finally, the red items in Figure 2 are factors other than current spending and income that affect the debt-income ratio. Arjun Jayadev and I call this set of factors “Fisher dynamics,” after Irving Fisher’s discussion of them in his famous paper on the Great Depression. Interest payments reduce measured saving and shift balance sheets toward deficit, just like consumption; but they don’t contribute to aggregate demand. Defaults or charge-offs reduce the outstanding stock of debt, without affecting demand or measured savings. Like capital gains, they are a change in a stock without any corresponding flow. [3] Finally, the debt-income ratio has a denominator as well as a numerator; it can be raised just as well by slower nominal income growth as by higher borrowing.

These factors are the subject of two papers you can find here and here. The bottom line is that a large part of historical changes in debt ratios — including the entire long-term increase since 1980 — are the result of the items shown in red here.

So what’s the point of all this?

First, borrowing is not the opposite of saving. Not even roughly. Matthew Klein, like most people, immediately translates rising debt into declining saving. The first half of his post is all about that. But saving and debt are very different things. True, increased consumption spending does reduce saving and increase debt, all else equal. But saving also depends on third party spending and imputed spending and income that has no effect on household balance sheets. While debt growth depends, in addition to saving, on residential investment, net acquisition of financial assets, and the rate of chargeoffs; if we are talking about the debt-income ratio, as we usually are, then it also depends on nominal income growth. And these differences matter, historically. If you are interested in debt and household expenditure, you have to look at debt and expenditure. Not saving.

Second, when we do look at expenditure by households, there is no long-term increase in consumption. Consumption spending is flat since 1980. Housing investment – which does involve outlays by households and may require debt financing – does increase in the late 1990s and early 2000s, before falling back. Yes, this investment was associated with a big rise in borrowing, and yes, this borrowing did come significantly lower in the income distribution that borrowing in most periods. (Though still almost all in the upper half.) There was a debt-financed housing bubble. But we need to be careful to distinguish this episode from the longer-term rise in household debt, which has different roots.

 

[1] Think of it this way: If I ask why the return on an investment was 20 percent, there is no end to causal factors you can bring in, from favorable macroeconomic conditions to a sound business plan to your investing savvy or inside knowledge. But in accounting terms, the return is always explained by the income and the capital gains over the period. If you know both those components, you know the return; if you don’t, you don’t. The relationships in the figure are the second kind of explanation.

[2] Improvement of existing housing is also counted as investment, as are brokers’ commissions and other ownership transfer costs. This kind of spending will absorb some part of the flow of mortgage financing to the household sector — including the cash-out refinancing of the bubble period — but I haven’t seen an estimate of how much.

[3] There’s a strand of heterodox macro called “stock-flow consistent modeling.” Insofar as this simply means macroeconomics that takes aggregate accounting relationships seriously, I’m very much in favor of it. Social accounting matrices (SAMs) are an important and underused tool. But it’s important not to take the name too literally — economic reality is not stock-flow consistent!

 

Two Papers in Progress

There are two new papers on the articles page on this site. Both are work in progress – they haven’t been submitted anywhere yet.

 

[I’ve taken the debt-distribution paper down. It’s being revised.]

The Evolution of State-Local Balance Sheets in the US, 1953-2013

Slides

The first paper, which I presented in January in Chicago, is a critical assessment of the idea of a close link between income distribution and household debt. The idea is that rising debt is the result of rising inequality as lower-income households borrowed to maintain rising consumption standards in the face of stagnant incomes; this debt-financed consumption was critical to supporting aggregate demand in the period before 2008. This story is often associated with Ragnuram Rajan and Mian and Sufi but is also widely embraced on the left; it’s become almost conventional wisdom among Post Keynesian and Marxist economists. In my paper, I suggest some reasons for skepticism. First, there is not necessarily a close link between rising aggregate debt ratios and higher borrowing, and even less with higher consumption. Debt ratios depend on nominal income growth and interest payments as well as new borrowing, and debt mainly finances asset ownership, not current consumption. Second, aggregate consumption spending has not, contrary to common perceptions, risen as a share of GDP; it’s essentially flat since 1980. The apparent rise in the consumption share is entirely due to the combination of higher imputed noncash expenditure, such as owners’ equivalent rent; and third party health care spending (mostly Medicare). Both of these expenditure flows are  treated as household consumption in the national accounts. But neither involves cash outlays by households, so they cannot affect household balance sheets. Third, household debt is concentrated near the top of the income distribution, not the bottom. Debt-income ratios peak between the 85th and 90th percentiles, with very low ratios in the lower half of the distribution. Most household debt is owed by the top 20 percent by income. Finally, most studies of consumption inequality find that it has risen hand-in-hand with income inequality; it appears that stagnant incomes for most households have simply meant stagnant living standards. To the extent demand has been sustained by “excess” consumption, it was more likely by the top 5 percent.

The paper as written is too polemical. I need to make the tone more neutral, tentative, exploratory. But I think the points here are important and have not been sufficiently grappled with by almost anyone claiming a strong link between debt and distribution.

The second paper is on state and local debt – I’ve blogged a bit about it here in the past few months. The paper uses budget and balance sheet data from the census of governments to make two main points. First, rising state and local government debt does not imply state and local government budget deficits. higher debt does not imply higher deficits: Debt ratios can also rise either because nominal income growth slows, or because governments are accumulating assets more rapidly. For the state and local sector as a whole, both these latter factors explain more of the rise in debt ratios than does the fiscal balance. (For variation in debt ratios across state governments, nominal income growth is not important, but asset accumulation is.) Second, despite balanced budget requirements, state and local governments do show substantial variation in fiscal balances, with the sector as a whole showing deficits and surpluses up to almost one percent of GDP. But unlike the federal government, the state and local governments accommodate fiscal imbalances entirely by varying the pace of asset accumulation. Credit-market borrowing does not seem to play any role — either in the aggregate or in individual states — in bridging gaps between current expenditure and revenue.

I will try to blog some more about both these papers in the coming days. Needless to say, comments are very welcome.

Varieties of the Phillips Curve

In this post, I first talk about a variety of ways that we can formalize the relationship between wages, inflation and productivity. Then I talk briefly about why these links matter, and finally how, in my view, we should think about the existence of a variety of different possible relationships between these variables.

*

My Jacobin piece on the Fed was, on a certain abstract level, about varieties of the Phillips curve. The Phillips curve is any of a family graphs with either unemployment or “real” GDP on the X axis, and either the level or the change of nominal wages or the level of prices or the level or change of inflation on the Y axis. In any of the the various permutations (some of which naturally are more common than others) this purports to show a regular relationship between aggregate demand and prices.

This apparatus is central to the standard textbook account of monetary policy transmission. In this account, a change in the amount of base money supplied by the central bank leads to a change in market interest rates. (Newer textbooks normally skip this part and assume the central bank sets “the” interest rate by some unspecified means.) The change in interest rates  leads to a change in business and/or housing investment, which results via a multiplier in a change in aggregate output. [1] The change in output then leads to a change in unemployment, as described by Okun’s law. [2] This in turn leads to a change in wages, which is passed on to prices. The Phillips curve describes the last one or two or three steps in this chain.

Here I want to focus on the wage-price link. What are the kinds of stories we can tell about the relationship between nominal wages and inflation?

*

The starting point is this identity:

(1) w = y + p + s

That is, the percentage change in nominal wages (w) is equal to the sum of the percentage changes in real output per worker (y; also called labor productivity), in the price level (p, or inflation) and in the labor share of output (s). [3] This is the essential context for any Phillips curve story. This should be, but isn’t, one of the basic identities in any intermediate macroeconomics textbook.

Now, let’s call the increase in “real” or inflation-adjusted wages r. [4] That gives us a second, more familiar, identity:

(2) r = w – p

The increase in real wages is equal to the increase in nominal wages less the inflation rate.

As always with these kinds of accounting identities, the question is “what adjusts”? What economic processes ensure that individual choices add up in a way consistent with the identity? [5]

Here we have five variables and two equations, so three more equations are needed for it to be determined. This means there are large number of possible closures. I can think of five that come up, explicitly or implicitly, in actual debates.

Closure 1:

First is the orthodox closure familiar from any undergraduate macroeconomics textbook.

(3a) w = pE + f(U); f’ < 0

(4a) y = y*

(5a) p = w – y

Equation 3a says that labor-market contracts between workers and employers result in nominal wage increases that reflect expected inflation (pE) plus an additional increase, or decrease, that reflects the relative bargaining power of the two sides. [6] The curve described by f is the Phillips curve, as originally formulated — a relationship between the unemployment rate and the rate of change of nominal wages. Equation 4a says that labor productivity growth is given exogenously, based on technological change. 5a says that since prices are set as a fixed markup over costs (and since there is only labor and capital in this framework) they increase at the same rate as unit labor costs — the difference between the growth of nominal wages and labor productivity.

It follows from the above that

(6a) w – p = y

and

(7a) s = 0

Equation 6a says that the growth rate of real wages is just equal to the growth of average labor productivity. This implies 7a — that the labor share remains constant. Again, these are not additional assumptions, they are logical implications from closing the model with 3a-5a.

This closure has a couple other implications. There is a unique level of unemployment U* such that w = y + p; only at this level of unemployment will actual inflation equal expected inflation. Assuming inflation expectations are based on inflation rates realized in the past, any departure from this level of unemployment will cause inflation to rise or fall without limit. This is the familiar non-accelerating inflation rate of unemployment, or NAIRU. [7] Also, an improvement in workers’ bargaining position, reflected in an upward shift of f(U), will do nothing to raise real wages, but will simply lead to higher inflation. Even more: If an inflation-targetting central bank is able to control the level of output, stronger bargaining power for workers will leave them worse off, since unemployment will simply rise enough to keep nominal wage growth in line with y*  and the central bank’s inflation target.

Finally, notice that while we have introduced three new equations, we have also introduced a new variable, pE, so the model is still underdetermined. This is intended. The orthodox view is that the same set of “real“ values is consistent with any constant rate of inflation, whatever that rate happens to be. It follows that a departure of the unemployment rate from U* will cause a permanent change in the inflation rate. It is sometimes suggested, not quite logically, that this is an argument in favor of making price stability the overriding goal of policy. [8]

If you pick up an undergraduate textbook by Carlin and Soskice, Krugman and Wells, or Blanchard, this is the basic structure you find. But there are other possibilities.

Closure 2: Bargaining over the wage share

A second possibility is what Anwar Shaikh calls the “classical” closure. Here we imagine the Phillips curve in terms of the change in the wage share, rather than the change in nominal wages.

(3b) s =  f(U); f’ < 0

(4b) y = y*

(5b) p = p*

Equation 3b says that the wage share rises when unemployment is low, and falls when unemployment is high. In this closure, inflation as well as labor productivity growth are fixed exogenously. So again, we imagine that low unemployment improves the bargaining position of workers relative to employers, and leads to more rapid wage growth. But now there is no assumption that prices will follow suit, so higher nominal wages instead translate into higher real wages and a higher wage share. It follows that:

(6b) w = f(U) + p + y

Or as Shaikh puts it, both productivity growth and inflation act as shift parameters for the nominal-wage Phillips curve. When we look at it this way, it’s no longer clear that there was any breakdown in the relationship during the 1970s.

If we like, we can add an additional equation making the change in unemployment a function of the wage share, writing the change in unemployment as u.

(7b) u = g(s); g’ > 0 or g’ < 0

If unemployment is a positive function of the wage share (because a lower profit share leads to lower investment and thus lower demand), then we have the classic Marxist account of the business cycle, formalized by Goodwin. But of course, we might imagine that demand is “wage-led” rather than “profit-led” and make U a negative function of the wage share — a higher wage share leads to higher consumption, higher demand, higher output and lower unemployment. Since lower unemployment will, according to 3b, lead to a still higher wage share, closing the model this way leads to explosive dynamics — or more reasonably, if we assume that g’ < 0 (or impose other constraints), to two equilibria, one with a high wage share and low unemployment, the other with high unemployment and a low wage share. This is what Marglin and Bhaduri call a “stagnationist” regime.

Let’s move on.

Closure 3: Real wage fixed.

I’ll call this the “Classical II” closure, since it seems to me that the assumption of a fixed “subsistence” wage is used by Ricardo and Malthus and, at times at least, by Marx.

(3c) w – p = 0

(4c) y = y*

(5c) p = p*

Equation 3c says that real wages are constant the change in nominal wages is just equal to the change in the price level. [9] Here again the change in prices and in labor productivity are given from outside. It follows that

(6c) s = -y

Since the real wage is fixed, increases in labor productivity reduce the wage share one for one. Similarly, falls in labor productivity will raise the wage share.

This latter, incidentally, is a feature of the simple Ricardian story about the declining rate of profit. As lower quality land if brought into use, the average productivity of labor falls, but the subsistence wage is unchanged. So the share of output going to labor, as well as to landlords’ rent, rises as the profit share goes to zero.

Closure 4:

(3d) w =  f(U); f’ < 0

(4d) y = y*

(5d) p = p*

This is the same as the second one except that now it is the nominal wage, rather than the wage share, that is set by the bargaining process. We could think of this as the naive model: nominal wages, inflation and productivity are all just whatever they are, without any regular relationships between them. (We could even go one step more naive and just set wages exogenously too.) Real wages then are determined as a residual by nominal wage growth and inflation, and the wage share is determined as a residual by real wage growth and productivity growth. Now, it’s clear that this can’t apply when we are talking about very large changes in prices — real wages can only be eroded by inflation so far.  But it’s equally clear that, for sufficiently small short-run changes, the naive closure may be the best we can do. The fact that real wages are not entirely a passive residual, does not mean they are entirely fixed; presumably there is some domain over which nominal wages are relatively fixed and their “real” purchasing power depends on what happens to the price level.

Closure 5:

One more.

(3e) w =  f(U) + a pE; f’ < 0; 0 < a < 1

(4e) y = b (w – p); 0 < b < 1

(5e) p =  c (w – y); 0 < c < 1

This is more generic. It allows for an increase in nominal wages to be distributed in some proportion between higher inflation, an increase in the wage share,  and faster productivity growth. The last possibility is some version of Verdoorn’s law. The idea that scarce labor, or equivalently rising wages, will lead to faster growth in labor productivity is perfectly admissible in an orthodox framework.  But somehow it doesn’t seem to make it into policy discussions.

In other word, lower unemployment (or a stronger bargaining position for workers more generally) will lead to an increase in the nominal wage. This will in turn increase the wage share, to the extent that it does not induce higher inflation and/or faster productivity growth:

(6e) s = (1  – b – c) w

This closure includes the first two as special cases: closure 1 if we set a = 0, b = 0, and c = 1, closure 2 if we set a = 1, b = 0, and c < 1. It’s worth framing the more general case to think clearly about the intermediate possibilities. In Shaikh’s version of the classical view, tighter labor markets are passed through entirely to a higher labor share. In the conventional view, they are passed through entirely to higher inflation. There is no reason in principle why it can’t be some to each, and some to higher productivity as well. But somehow this general case doesn’t seem to get discussed.

Here is a typical example  of the excluded middle in the conventional wisdom: “economic theory suggests that increases in labor costs in excess of productivity gains should put upward pressure on prices; hence, many models assume that prices are determined as a markup over unit labor costs.” Notice the leap from the claim that higher wages put some pressure on prices, to the claim that wage increases are fully passed through to higher prices. Or in terms of this last framework: theory suggests that b should be greater than zero, so let’s assume b is equal to one. One important consequence is to implicitly exclude the possibility of a change in the wage share.

*

So what do we get from this?

First, the identity itself. On one level it is obvious. But too many policy discussions — and even scholarship — talk about various forms of the Phillips curve without taking account of the logical relationship between wages, inflation, productivity and factor shares. This is not unique to this case, of course. It seems to me that scrupulous attention to accounting relationships, and to logical consistency in general, is one of the few unambiguous contributions economists make to the larger conversation with historians and other social scientists. [10]

For example: I had some back and forth with Phil Pilkington in comments and on twitter about the Jacobin piece. He made some valid points. But at one point he wrote: “Wages>inflation + productivity = trouble!” Now, wages > inflation + productivity growth just means, an increasing labor share. It’s two ways of saying the same thing. But I’m pretty sure that Phil did not intend to write that an increase in the labor share always means trouble. And if he did seriously mean that, I doubt one reader in a hundred would understand it from what he wrote.

More consequentially, austerity and liberalization are often justified by the need to prevent “real unit labor costs” from rising. What’s not obvious is that “real unit labor costs” is simply another word for the labor share. Since by definition the change real unit labor costs is just the change in nominal wages less sum of inflation and productivity growth. Felipe and Kumar make exactly this point in their critique of the use of unit labor costs as a measure of competitiveness in Europe: “unit labor costs calculated with aggregate data are no more than the economy’s labor share in total output multiplied by the price level.” As they note, one could just as well compute “unit capital costs,” whose movements would be just the opposite. But no one ever does, instead they pretend that a measure of distribution is a measure of technical efficiency.

Second, the various closures. To me the question of which behavioral relations we combine the identity with — that is, which closure we use — is not about which one is true, or best in any absolute sense. It’s about the various domains in which each applies. Probably there are periods, places, timeframes or policy contexts in which each of the five closures gives the best description of the relevant behavioral links. Economists, in my experience, spend more time working out the internal properties of formal systems than exploring rigorously where those systems apply. But a model is only useful insofar as you know where it applies, and where it doesn’t. Or as Keynes put it in a quote I’m fond of, the purpose of economics is “to provide ourselves with an organised and orderly method of thinking out particular problems” (my emphasis); it is “a way of thinking … in terms of models joined to the art of choosing models which are relevant to the contemporary world.” Or in the words of Trygve Haavelmo, as quoted by Leijonhufvud:

There is no reason why the form of a realistic model (the form of its equations) should be the same under all values of its variables. We must face the fact that the form of the model may have to be regarded as a function of the values of the variables involved. This will usually be the case if the values of some of the variables affect the basic conditions of choice under which the behavior equations in the model are derived.

I might even go a step further. It’s not just that to use a model we need to think carefully about the domain over which it applies. It may even be that the boundaries of its domain are the most interesting thing about it. As economists, we’re used to thinking of models “from the inside” — taking the formal relationships as given and then asking what the world looks like when those relationships hold. But we should also think about them “from the outside,” because the boundaries within which those relationships hold are also part of the reality we want to understand. [11] You might think about it like laying a flat map over some curved surface. Within a given region, the curvature won’t matter, the flat map will work fine. But at some point, the divergence between trajectories in our hypothetical plane and on the actual surface will get too large to ignore. So we will want to have a variety of maps available, each of which minimizes distortions in the particular area we are traveling through — that’s Keynes’ and Haavelmo’s point. But even more than that, the points at which the map becomes unusable, are precisely how we learn about the curvature of the underlying territory.

Some good examples of this way of thinking are found in the work of Lance Taylor, which often situates a variety of model closures in various particular historical contexts. I think this kind of thinking was also very common in an older generation of development economists. A central theme of Arthur Lewis’ work, for example, could be thought of in terms of poor-country labor markets that look  like what I’ve called Closure 3 and rich-country labor markets that look like Closure 5. And of course, what’s most interesting is not the behavior of these two systems in isolation, but the way the boundary between them gets established and maintained.

To put it another way: Dialectics, which is to say science, is a process of moving between the concrete and the abstract — from specific cases to general rules, and from general rules to specific cases. As economists, we are used to grounding concrete in the abstract — to treating things that happen at particular times and places as instances of a universal law. The statement of the law is the goal, the stopping point. But we can equally well ground the abstract in the concrete — treat a general rule as a phenomenon of a particular time and place.

 

 

 

[1] In graduate school you then learn to forget about the existence of businesses and investment, and instead explain the effect of interest rates on current spending by a change in the optimal intertemporal path of consumption by a representative household, as described by an Euler equation. This device keeps academic macroeconomics safely quarantined from contact with discussion of real economies.

[2] In the US, Okun’s law looks something like Delta-U = 0.5(2.5 – g), where Delta-U is the change in the unemployment rate and g is inflation-adjusted growth in GDP. These parameters vary across countries but seem to be quite stable over time. In my opinion this is one of the more interesting empirical regularities in macroeconomics. I’ve blogged about it a bit in the past  and perhaps will write more in the future.

[3] To see why this must be true, write L for total employment, Z for the level of nominal GDP, Y for per-capita GDP, W for the average wage, and P for the price level. The labor share S is by definition equal to total wages divided by GDP:

S = WL / Z

Real output per worker is given by

Y = (Z/P) / L

Now combine the equations and we get W = P Y S. This is in levels, not changes. But recall that small percentage changes can be approximated by log differences. And if we take the log of both sides, writing the log of each variable in lowercase, we get w = y + p + s. For the kinds of changes we observe in these variables, the approximation will be very close.

[4] I won’t keep putting “real” in quotes. But it’s important not to uncritically accept the dominant view that nominal quantities like wages are simply reflections of underlying non-monetary magnitudes. In fact the use of “real” in this way is deeply ideological.

[5] A discovery that seems to get made over and over again, is that since an identity is true by definition, nothing needs to adjust to maintain its equality. But it certainly does not follow, as people sometimes claim, that this means you cannot use accounting identities to reason about macroeconomic outcomes. The point is that we are always using the identities along with some other — implicit or explicit — claims about the choices made by economic units.

[6] Note that it’s not necessary to use a labor supply curve here, or to make any assumption about the relationship between wages and marginal product.

[7] Often confused with Milton Friedman’s natural rate of unemployment. But in fact the concepts are completely different. In Friedman’s version, causality runs the other way, from the inflation rate to the unemployment rate. When realized inflation is different from expected inflation, in Friedman’s story, workers are deceived about the real wage they are being offered and so supply the “wrong” amount of labor.

[8] Why a permanently rising price level is inconsequential but a permanently rising inflation rate is catastrophic, is never explained. Why are real outcomes invariant to the first derivative of the price level, but not to the second derivative? We’re never told — it’s an article of faith that money is neutral and super-neutral but not super-super-neutral. And even if one accepts this, it’s not clear why we should pick a target of 2%, or any specific number. It would seem more natural to think inflation should follow a random walk, with the central bank holding it at its current level, whatever that is.

[9] We could instead use w – p = r*, with an exogenously given rate of increase in real wages. The logic would be the same. But it seems simpler and more true to the classics to use the form in 3c. And there do seem to be domains over which constant real wages are a reasonable assumption.

[10] I was just starting grad school when I read Robert Brenner’s long article on the global economy, and one of the things that jumped out at me was that he discussed the markup and the wage share as if they were two independent variables, when of course they are just two ways of describing the same thing. Using s still as the wage share, and m as the average markup of prices over wages, s = 1 / (1 + m). This is true by definition (unless there are shares other than wages or profits, but none such figure in Brenner’s analysis). The markup may reflect the degree of monopoly power in product markets while the labor share may reflect bargaining power within the firm, but these are two different explanations of the same concrete phenomenon. I like to think that this is a mistake an economist wouldn’t make.

[11] The Shaikh piece mentioned above is very good. I should add, though, the last time I spoke to Anwar, he criticized me for “talking so much about the things that have changed, rather than the things that have not” — that is, for focusing so much on capitalism’s concrete history rather than its abstract logic. This is certainly a difference between Shaikh’s brand of Marxism and whatever it is I do. But I’d like to think that both approaches are called for.

 

EDIT: As several people pointed out, some of the equations were referred to by the wrong numbers. Also, Equation 5a and 5e had inflation-expectation terms in them that didn’t belong. Fixed.

EDIT 2: I referred to an older generation of development economics, but I think this awareness that the territory requires various different maps, is still more common in development than in most other fields. I haven’t read Dani Rodrik’s new book, but based on reviews it sounds like it puts forward a pretty similar view of economics methodology.

The End of the Supermanager?

Everyone is talking about this new paper, Firming Up Inequality. It uses individual-level data from the Social Security Administration, matched to employers by Employer Identification Number (EIN), to decompose changes in earnings inequality into a within-firm and a between-firm component. It’s a great exercise — marred only modestly by the fact that the proprietary data means that no one can replicate it — exactly the sort of careful descriptive work I wish more economists would do.

The big finding from the paper is that all the rise in earnings inequality between 1982 and 2012 is captured by the between-firm component. There is no increase in the earnings of a person in the top 1% of the earnings distribution within a given business, and the earnings of someone at the median for that same business. The whole increase in earnings inequality over this period consists of a widening gap between the firms that pay more across the board, and the firms that pay less.

I’m not sure we want to take the results of this study at face value. Yes, we should be especially interested in empirical work that challenges our prior beliefs, but at the same time, it’s hard to square the claims here with all the other evidence of a disproportionate increase in the top pay within a given firm. Lawrence Mishel gives some good reasons for skepticism here. The fact that the whole increase is accounted for by the between-firm component, yet none by the between-industry component, is very puzzling. More generally, I wonder how reliable is the assumption that there is a one to one match between EINs and what we normally think of as employers.

That said, these findings may be pointing to something important. As a check on the plausibility of the numbers in the paper, I took a look at labor income of the top 1 percent and 0.01 percent of US households, as reported in the World Top Incomes Database. And I found something I didn’t expect: Since 2000, there’s been a sharp fall in the share of top incomes that come from wages and salaries. In 2000, according to the tax data used by Piketty and his collaborators, households in the top 0.01 percent got 61 percent of their income from wages, salaries and pensions. By 2013, that had fallen to just 33 percent. (That’s excluding capital gains; including them, the labor share of top incomes fell from 31 percent to 21 percent.) For the top 1 percent, the labor share falls from 63 percent to 56 percent, the lowest it’s been since the 1970s.

Here is the average income of the top 0.01 percent over the past 40 years in inflation-adjusted dollars, broken into three components: labor income, all other non-capital gains income, and capital gains.

01percent_income
Average income of top 0.01% of US households, from World Top Incomes Database. 3-year moving averages.

As you can see, the 1990s look very different from today. Between 1991 and 2000, the average labor income of a top 0.01% household rose from $2.25 million to $10 million; this was about 90 percent of the total income increase for these households. During the 1990s, rising incomes at the top really were about highly paid superstars. Since 2000, though, while average incomes of the top 0.01% have increased another 20 percent, labor income for these households has fallen by almost half, down to $5.5 million. (Labor income has also fallen for the top 1 percent, though less dramatically.) So the “Firming” results, while very interesting, are perhaps less important for the larger story of income distribution than both the authors and critics assume. The rise in income inequality since 2000 is not about earnings; the top of the distribution is no longer the working rich. I don’t think that debates about inequality have caught up with this fact.

Fifteen years ago, the representative rich person in the US was plausibly a CEO, or even an elite professional. Today, they mostly just own stuff.

 

Causes and Effects of Wage Growth

Over here, a huge stack of exams, sitting ungraded since… no, I can’t say, it’s too embarrassing.  There, a grant proposal that extensive experimentation has shown will not, in fact, write itself. And I still owe a response to all the responses and criticism to my Disgorge the Cash paper for Roosevelt. So naturally, I thought this morning would be a good time to sit down and ask what we can learn from comparing the path of labor costs in the Employment Cost Index compared with the ECEC.

The BLS explains the difference between the two measures:

The Employment Cost Index, or ECI, measures changes in employers’ cost of compensating workers, controlling for changes in the industrial-occupational composition of jobs. … The ECI is intended to indicate how the average compensation paid by employers would have changed over time if the industrial-occupational composition of employment had not changed… [It] controls for employment shifts across 2-digit industries and major occupations. The Employer Costs for Employee Compensation, or ECEC… is designed to measure the average cost of employee compensation. Accordingly, the ECEC is calculated by multiplying each job quote by its sample weight.

In other words, the ECI measures the change in average hourly compensation, controlling for shifts in the mix of industries and occupations. The ECEC simply measures the overall change in hourly compensation, including the effects of both changes in compensation for particular jobs, and changes in the mix of jobs.

Here are the two series for the full period both are available (1987-2014), both raw and adjusted for inflation (“real”).

What do we learn from this?

First, the two series are closely correlated. This tells us that most of the variation in compensation is driven by changes within occupations and sectors, not by shifts in employment between occupations and sectors. This is clearly true at annual frequencies but it seems to be true over longer periods as well. For instance, let’s compare the behavior of compensation in the five years since the end of the recession to the last period of strong wage growth, 1997-2004. The difference between the two periods in the average annual increase in nominal wages is almost exactly the same according to the two indexes — 2.7 points by the ECI, 2.6 points by the ECEC. In other words, slower wage growth in the recent period is entirely due to slower wages growth within particular kinds of jobs. Shifts in the composition of jobs have played no role at all.

On the face of it, the fact that almost all variation in aggregate compensation is driven by changes within employment categories, seems to favor a labor/political story of slower wage growth as opposed to a China or robots story. The most obvious versions of the latter two stories involve a disproportionate loss of high-wage jobs, whereas stories about weaker bargaining position of labor predict slower compensation growth within job categories. I wouldn’t ask this one piece of evidence to carry a lot of weight in that debate. (I think it’s stronger evidence against a skills-based explanation of slower wage growth.)

While the two series in general move together, the ECEC is more strongly cyclical. In other words, during periods of high unemployment and falling wages in general, there is also a shift in the composition of employment towards lower-paid occupations. And during booms, when unemployment is low and wages are rising in general, there is a shift in the direction of higher-paid job categories. [1] Insofar as wages and labor productivity are correlated, this cyclical shift between higher-wage and lower-wage sectors could help explain why employment is more stable than output. I’ve had the idea for a while that the Okun’s law relationship — the less than one-for-one correlation between employment and output growth — reflects not only hiring/firing costs and overhead labor, but also shifts in the composition of employment in response to demand. In other words, in addition to employment adjustment costs at the level of individual enterprises, the Okun coefficient reflects cyclically varying degrees of “disguised unemployment” in Joan Robinson’s sense. [2] This is an argument I’d like to develop properly someday, since it seems fairly obvious, potentially important and empirically tractable, and I haven’t seen anyone else make it. [3] (I’m sure someone has.)

What’s going on in the most recent year? Evidently, there has been no acceleration of wage growth for a given job, but the mix of jobs created has shifted toward higher-wage categories. This suggests that to the extent wages are rising faster, it’s not a sign of labor-market pressures. (Some guy from Deutsche Bank interprets the same divergence as support for raising rates, which it’s hard not to feel is deliberately dishonest.) As for which particular higher-wage job categories are growing more rapidly — I don’t know. And, what’s going on in 1995? That year has by far the biggest divergence between the two series. It could well be an artifact of some kind, but if not, seems important. A large fall in the ECEC relative to the ECI could be a signature of deindustrialization. I’m not exploring the question further now (those exams…) but it would be interesting to ask analogous question with some series that extends earlier. It’s likely that if we were looking at the 1970s-1980s, we would find a much larger share of variation in wage growth explained by compositional shifts.

Should we adjust for inflation? I give the “real” series here, but I am in general skeptical that there is any sense in which an ex post adjustment of money flows for inflation is more real than, say, The Real World on MTV. I am even more doubtful than usual in this case, because we are normally told to think that changes in nominal wages are the main determinant of inflation. Obviously in that case we have to think of the underlying labor-market process as determining a change in nominal wage. Still, if we do compute a “real” index, things look a little different. Real ECI rises 14 percent over the full 1987-2014 period, while real ECEC rises only 5 percent. So now we can say that about two-thirds of the increase in real wages within particular job categories over the past three decades, was offset by a shift in the composition of employment toward lower-paid job categories. (This is all in the first decade, 1987-1996, however.) This way of looking at things makes sense if we think the underlying wage-setting process, whatever it is, operates in terms of a basket of consumption goods.

This invites another question: How true is it that nominal wages move with inflation?

Conventional economics wisdom suggests we can separate wages into nominal and “real” components. This is on two not quite consistent grounds. First, we might suppose that workers and employers are implicitly negotiating contracts in terms of a fixe quantity of labor time for, on the one hand, a basket of wage goods, and on the other, a basket of produced goods (which will be traded for consumption good for the employer). This contract only incidentally happens to be stated in terms of money. The ultimate terms on which consumption goods for the workers exchange with consumption goods for the employer should not be affected by the units the trade happens to be denominated in. (In this respect the labor contract is just like any other contract.) This is the idea behind Milton Friedman’s “natural rate of unemployment” hypothesis. In Friedman’s story, causality runs strictly from inflation to unemployment. High inflation is not immediately recognized by workers, leading them to overestimate the basket of goods their wages will buy. So they work more hours than they would have chosen if they had correctly understood the situation. From this point of view, there’s no cost to low unemployment in itself; the problem is just that unemployment will only be low if high inflation has tricked workers into supply too much labor. Needless to say, this is not the way anyone in the policy world thinks about the inflation-unemployment nexus today, even if they continue to use Friedman’s natural rate language.

The alternative view is that workers and employers negotiate a money-wage, and then output prices are set as a markup over that wage. In this story, causality runs from unemployment to inflation. While Friedman thought an appropriate money-supply growth rate was the necessary and sufficient condition for stable prices, with any affect on unemployment just  collateral damage from changes in inflation, in this story keeping unemployment at an appropriate level is a requirement for stabilizing prices. This is the policy orthodoxy today.  (So while people often say that NAIRU is just another name for the natural rate of unemployment, in fact they are different concepts.) I think there are serious conceptual difficulties with the orthodox view, but we’ll save those for another time. Suffice it to say that causality is supposed to run from low unemployment, to faster nominal wage growth, to higher inflation. So the question is: Is it really the case that faster nominal wage growth is associated with higher inflation?

Wage Growth and Inflation, 1947-2014

A simple scatterplot suggests a fairly tight relationship, especially at higher levels of wage growth and inflation. But if we split the postwar period at 1985, things look very different. In the first period, there’s a close relationship — regressing inflation on nominal wage growth gives an R-squared of 0.81. (Although even then the coefficient is significantly less than 1.)

Wage Growth and Inflation, 1947-1985

Since 1985, though, the relationship is much looser, with an R-squared of 0.12. And even is that driven almost entirely by period of falling wages and prices in 2009; remove that and the correlation is essentially zero.

Wage Growth and Inflation, 1986-2014

So while it was formerly true that changes in inflation were passed one for one into changes in nominal wages, and/or changes in nominal wage growth led to similar changes in inflation, neither of those things has been true for quite a while now. In recent decades, faster nominal wage growth does not translate into higher inflation.

Obviously, a few scatterplots aren’t dispositive, but they are suggestive. So supposing that there has been a  delinking of wage growth and inflation, what conclusions might we draw? I can think of a couple.

On the one hand, maybe we shouldn’t be so dismissive of  the naive view that inflation reduces the standard of living directly, by raising the costs of consumption goods while incomes are unchanged. There seems to be an emerging conventional wisdom in this vicinity. Here for instance is Gillian Tett in the FT, endorsing the BIS view that there’s nothing wrong with falling prices as long as asset prices stay high. (Priorities.) In the view of both Keynes (in the GT; he modified it later) and Schumpeter, inflation was associated with higher nominal but lower real wages, deflation with lower nominal but higher real wages. I think this may have been true in the 19th century. It’s not impossible it could be true in the future.

On the other hand. If the mission of central banks is price stability, and if there is no reliable association between changes in wage growth and changes in inflation, then it is hard to see the argument for tightening in response to falling unemployment. You really should wait for direct evidence of rising inflation. Yet central banks are as focused on unemployment as ever.

It’s perhaps significant in this regard that the authorities in Europe are shifting away from the NAIRU (Non-Accelerating Inflation Rate of Unemployment) and increasingly talking about the NAWRU (Non-Accelerating Wage Rate of Unemployment). If the goal all along has been lower wage growth, then this is what you should expect: When the link between wages and inflation weakens, the response is not to find other tools for controlling inflation, but other arguments for controlling wages. This may be the real content of the “competitiveness” discourse. Elevating competitiveness over price stability as overarching goal of policy lets you keep pushing down wages even when inflation is already low.

Worth noting here: While the ECB’s “surrender Dorothy” letter to the Spanish government ordered them to get rid of price indexing, their justification was not, as you might expect, that indexation contributes to inflationary spirals. Rather it was that it is “a structural obstacle to the adjustment of labour costs” and “contribute to hampering competitiveness.” [4]  This is interesting. In the old days we would have said, wage indexing is bad because it won’t affect real wages, it just leads to higher inflation. But apparently in the new dispensation, we say that wage indexing is bad precisely because it does affect real wages.

[1]  This might seem to contradict the previous point but it doesn’t, it’s just that the post-2009 recovery period includes both a negative composition shift in 2008-2009, when unemployment was high, and a positive compositional shift in 2014, which cancel each other out.

[2] From A Theory of Employment: “Except under peculiar conditions, a decline in effective demand which reduces the amount of employment offered in the general run of industries will not lead to ‘unemployment’ in the sense of complete idleness, but will rather drive workers into a number of occupations [such as] selling match-boxes in the Strand, cutting brushwood in the jungles, digging potatoes on allotments which are still open to them. A decline in one sort of employment leads to an increase in another sort, and at first sight it may appear that, in such a case, a decline in effective demand does not cause unemployment at all. But the matter must be more closely examined. In all those occupations which the dismissed workers take up, their productivity is less than in the occupations that they have left.”

[3] The only piece I know of that makes the connection between demand and productivity variation across sectors is this excellent article by John Eatwell (which unfortunately doesn’t seem to be available online), but it is focused on long run variation, not cyclical.

[4] The ECB’s English is not the most felicitous, is it? The Spanish version is “contribuyen a dificultar la competitividad y el crecimiento,” which also doesn’t strike me as a phrase that a native speaker would write. Maybe it sounds better in the original German.