Category Archives: PSST and Macro

Inflation Defies Fundamentals

Posted on by

Jan Groen writes,

Another issue is how to find the right variables to predict future inflation. Economists often use the Phillips curve relationship, with inflation depending inversely on unemployment—that is, lower unemployment puts upward pressure on wages and, eventually, on inflation. But while an unemployment rate variable is common, it isn’t clear that this is the best gauge. Stock and Watson (1999) and Wright (2009) consider a broader range of possible “economic slack” variables and then use different ways to condense the information in these variables to predict future inflation. Generally, these approaches do as well as or better than autoregressive models. Atkeson and Ohanian (2001), however, conduct a similar exercise but focus on the post-1985 period, which—up to the advent of the Great Recession—was a period typified by remarkably low and stable inflation in the United States. Using the Chicago Fed National Activity Index (CFNAI), which summarizes information across many activity variables, they show that the resulting inflation forecasts are worse than when one just relies on current inflation to forecast future inflation.

Links omitted. Pointer from Mark Thoma.

Standard macro theory includes a “tight” model of inflation. You might have a Phillips Curve in which inflation depends on unemployment. Or you might prefer a monetarist formulation, in which inflation depends on money growth.

However, if you rely on these relationships in the real world, your predictions for inflation will be awful. Think of your forecast as a weighted average of

a) inflation tomorrow will be what it was yesterday; and
b) inflation tomorrow will be determined by measures of economic slack and/or money growth

If you put any significant weight on (b), you will be hosed.

My own inclination, which I think is close to that of the late Fischer Black, is to treat both money and the average level of prices as consensual hallucinations. As far as money goes, we accept currency and abstractions representing currency today because we expect that other people will accept them tomorrow. The prices that we charge today are based on prices that we expect to face tomorrow. These habits and beliefs are extremely sticky, and they are usually self-validating.

Let me put it this way:

the average behavior of prices is an emergent phenomenon, not a phenomenon that is controlled top-down through the manipulations of the central bank.

Read that several times, until you appreciate the heresy.

I do subscribe to a fiscal theory of hyperinflation. If the government runs deficits and loses the ability to fund those deficits with long-term borrowing, then it has to go on a money-printing frenzy that will destroy the emergent properties of money and prices.

Along those lines, I am curious as to what will happen in Japan. The fiscal and monetary authorities there would like to engineer a moderate level of inflation. What they seem to be doing strikes me as sufficient to generate hyperinflation. So far, a consensual hallucination of zero inflation seems to be holding.

I think that economists should be very modest and humble in claiming to understand inflation.

Larry Summers on Upward-Sloping AD

Posted on by

He writes,

Notice that as Keynes, Tobin and subsequently Brad Delong and I have emphasized, wage and price flexibility may well exacerbate the problem. The more flexible wages and prices are, the more they will be expected to fall during an output slowdown leading to an increase in real interest rates. Indeed there is the possibility of destabilizing deflation with falling prices leading to higher real interest rates leading to greater output shortfalls leading to more rapidly falling prices and onwards in a vicious cycle.

Read the whole thing. There were many sentences I wanted to excerpt. Pointer from Mark Thoma.

In AS-AD, if your Y-axis is inflation rather than the price level, then AD slopes upward. That is, the more inflation you have, the lower the real interest rate, and the higher is AD. If AD gets steeper than AS, then have a nice day. Because if that happens, then a “favorable” supply shift leads to lower employment and output. One way to interpret secstag is as a claim that we have been experiencing an AD curve that is upward-sloping and steeper than AS.

Keep in mind that Summers and other mainstream macro economists talk about potential GDP as if it were some tangible quantity, rather than a made-up number. In mainstream macro, we all work in a GDP factory, and the factory has a capacity that we call potential GDP.

The PSST story rejects that. It says that we produce many different types of output, and we only have the potential to produce the output for which we have discovered patterns of sustainable specialization and trade. If we could discover other patterns of sustainable specialization and trade, we could produce a different mix of output, and perhaps this would raise the level of GDP. But raising GDP by discovering patterns of specialization and trade is akin to raising GDP by discovering a practical cold fusion technology. Complaining about the economy operating below potential is like complaining that we do not have cold fusion.

Related: Tyler Cowen on the difficulty of disentangling AD from AS. Plus Scott Sumner commenting on Tyler Cowen.

Robert Solow and Russ Roberts

Posted on by

On this podcast. As many commenters point out, Solow has a lot of acuity (and stamina!) for someone aged 90.

I think that in their discussion of what people believe in macro and why, I prefer Solow’s rendition. I think he is correct that it was not as simple as saying that Chicago taught monetarism and MIT taught fiscalism. (Small world note: Roberts’ undergraduate friend who went to MIT was my roommate during our first year of graduate school.) I think Solow is correct that the AS/AD paradigm was very different from what came before. However, I would note that until the 1970s, it was the AD paradigm alone, with the Phillips Curve (popularized by Samuelson and Solow) tacked on. In the 1970s, as economists started to think more about inflation and about what sort of theory might explain the Phillips Curve, there emerged the AS-AD paradigm. Most of the macro that came out of that 1970s theorizing ended up going in a direction that neither Solow nor I liked. As I point out in my macro memoir, that is how I was drawn to Solow as a thesis adviser.

PSST–it’s Bananagrams

Posted on by

The game Bananagrams may be a good metaphor for patterns of sustainable specialization and trade. In the game, each player is dealt a bunch of tiles with letters on them, like Scrabble tiles. You play as in Scrabble, but without a board and with each player playing his or her tiles alone. You form words free-form, but they must all connect, as in Scrabble. The object is to use all of your tiles. The rules are such that as the game proceeds, you are often forced to draw a new tile before you have used all of your tiles.

Sometimes, you have most of your tiles connected, with very few “unemployed” tiles, and when you get a new tile you can quickly “employ” it, meaning that you can add it to your existing set of words. That situation would represent a high-employment economy with patterns of sustainable specialization and trade.

However, sometimes you get a new tile, or a few of them, and you realize that you cannot make use of these tiles without breaking up several of your current words and starting over. You might be down to just one or two “unemployed” tiles, but breaking up some of your words gives you many “unemployed” tiles. That situation represents an economy where patterns of specialization and trade have become unsustainable. Getting back to a high-employment situation takes time and trial-and-error, just as in Bananagrams it takes time to recover when you find that your current word pattern won’t let you use all your tiles and you need to break up some of your words and create new words.

Paul Krugman on the State of Macro

Posted on by

He writes,

whenever somebody claims to have a deeper understanding of economics (or actually anything) that transcends the insights of simple models, my reaction is that this is self-delusion. Any time you make any kind of causal statement about economics, you are at least implicitly using a model of how the economy works. And when you refuse to be explicit about that model, you almost always end up – whether you know it or not – de facto using models that are much more simplistic than the crossing curves or whatever your intellectual opponents are using.

Pointer from Mark Thoma.

Shorter version: it takes a model to beat a model

I find that persuasive, although I wish we had a less pretentious term than “model,” which makes us sound more scientific than I believe we are. What I would like to do is compare the state of the IS-LM-Phillips Curve model with that of the PSST model. PSST, or patterns of sustainable specialization and trade, says that a sharp rise of unemployment is the result of patterns of trade made unsustainable by changes in technology and/or asset prices, and that only a gradual process of entrepreneurial trial and error can discover new patterns of sustainable specialization and trade. More about this model can be found here (see items 2,3, and 4).

Some comments:

1. I think that both PSST and IS-LM-Phillips are difficult to falsify. In some sense, the macroeconomic data are over-fit, so that the dataset cannot be used to decisively reject one model in favor of another.

2. I would counter one of Krugman’s narrow points, where he says that the liquidity trap explains why the huge reserves in the banking system have not had major effects. If this is true, then why did the Fed need to pay interest on reserves? I assume that the Fed thinks that if they had not paid interest on reserves, then its huge injection of reserves would have caused rapid money growth and high inflation. It would be interesting to poll economists on whether they would agree–my guess is that most would. It seems to me that Krugman would have to say “no” to the poll, but perhaps I am misunderstanding him. I am not sure how I would answer such a poll myself.

3. I do not align myself with those who see the Fed as the prime mover of inflation. My “model” of inflation is a pretty weak one. Basically, I just think that businesses get into habits about how much to raise workers’ wages each year. Maybe those habits are affected by the aggregate unemployment rate, as in the Phillips Curve, but I would caution that we do not have homogeneous “labor.” Some folks can be getting regular raises that are large, while others may fail to get raises at all. Each business looks at its own labor market, not at the economy-wide unemployment rate.

What about the 1970s? I would say that the 1970s were a period of “inflation consciousness.” Everyone became aware of it, and “cost-of-living” raises got built into the system, because so many employers incorporated recent inflation into current wage increases. I am tempted to suggest that the advent of wage-price controls starting in 1971 had the adverse consequence of raising inflation consciousness.

What about hyperinflation? I believe that really, big, long-term inflation is a fiscal phenomenon. That is, the government runs huge deficits, people stop lending to government, and then it meets its deficits with paper claims. We are not at that point in the U.S., but if we ever do reach the point where bond investors lose confidence–watch out.

4. So I have a PSST model for unemployment, and my “weak” model[s] for inflation. I think it is fair to criticize them as “just-so stories.” But I would say the same thing about the sorts of models preferred by Blanchard or Krugman. Just-so stories, dressed up in pretty math.

5. Elsewhere, Krugman writes,

Basically, the new IO [industrial organization, the field of recent Nobel Laureate Jean Tirole] made it OK to tell stories rather than proving theorems, and thereby made it possible to talk about and model issues that had been ruled out by the limits of perfect competition. It was, I can tell you from experience, profoundly liberating.

The State of Macroeconomic Analysis

Posted on by

Olivier Blanchard, who in August of 2008 describe it as “good,” has modified his views. Concerning the consensus methodology that he praised back then, Blanchard writes,

However, these techniques only made sense under a vision in which economic fluctuations were regular enough so that, by looking at the past, people and firms (and the econometricians who apply statistics to economics) could understand their nature and form expectations of the future; and simple enough so that small shocks had small effects, and a shock twice as big as another had twice the effect on economic activity. The reason for this assumption, called linearity, was technical. Models with nonlinearities – those in which a small shock, such as a decrease in housing prices, can sometimes have large effects, or in which the effect of a shock depends on the rest of the economic environment – were difficult, if not impossible, to solve under rational expectations.

Pointers from Mark Thoma and Greg Mankiw. Read the whole thing. I have to give Blanchard credit for writing this:

The reality of financial regulation is that new rules open new avenues for regulatory arbitrage, as institutions find loopholes in regulations. That in turn forces authorities to institute new regulations in an ongoing cat-and-mouse game (between a very adroit mouse and a less nimble cat). Staying away from dark corners will require continuous effort, not one-shot regulation.

That is the theme of The Chess Game of Financial Regulation.

However, my overall take on Blanchard’s essay will be harsh.*

1. He still wants to believe in equations and technical brilliance. He implies that if we were just better at manipulating nonlinear models, all would be well. Once again, an MIT economist is unable to grasp Hayek’s insight that there is knowledge embedded in the economic order that no individual can possess. No thanks to MIT, and long after I had left it, I wound up on the side of Hayek. In fact, when it comes to macro I would argue that I am more Hayekian than Hayek.

2. Rational expectations helped make the careers of Fischer, Blanchard, and other MIT contemporaries, and refusal to tool up in rational-expectations modeling is what un-made my own academic career. In hindsight, and with an assist from Frydman and Goldberg, I would say that rational expectations was the ultimate anti-Hayek proposition. In effect, Chicago said that everyone knows everything. Eventually, MIT countered with “behavioral economics,” which said that some people are often mistaken while assuming at least implicitly that the technocratic elites know everything.

3. My least favorite paragraph:

Now that we are more aware of nonlinearities and the dangers they pose, we should explore them further theoretically and empirically – and in all sorts of models. This is happening already, and to judge from the flow of working papers since the beginning of the crisis, it is happening on a large scale. Finance and macroeconomics in particular are becoming much better integrated, which is very good news.

I’ll be uncharitable (and sarcastic) and say that he is telling us once again that the state of macro is good, because the same modeling hubris still predominates. The way I see it, the drunks are still looking under the same lamppost.

As for integrating finance and macroeconoimcs, my prediction is that this will accomplish nothing. I believe that mainstream macroeconomists are over-stating the importance of the financial crisis. Instead, I am inclined to treat the financial crisis as a blip, one whose apparent macroeconomic impact was made somewhat worse by the very policies that mainstream economists claim were successful.

This blip took place in the context of key multi-decade trends:

–the transition away from goods-producing sectors and toward the New Commanding Heights of education and health care

–the transition of successful men away from marrying housekeepers and toward marrying successful women

–the integration of workers in other nations, most notably China and India, into the U.S. production system

–the increasing power of computer technology that ise more complementary to some workers than others

These trends are what explain the patterns of employment and relative wages that we observe. The financial crisis, and the government panic in response, pushed the impact of some of these developments forward in time. Overall, however, the focal points of mainstream macroeconomics, including fiscal and monetary measures, are not nearly as significant to the actual economy as they are on paper in the models.

I have always been harsh on Blanchard. You should discount for that.

The State of the Economy

Posted on by

1. There have been several posts pointing out that wage growth has been slow, even though the unemployment rate has fallen.

2. There have been several posts, including some of mine, on low long-term interest rates. More recently, the WSJ talked with James Bullard.

Right now, “the markets are making a mistake” and expect the Fed to maintain its ultra-easy policy stance longer than Fed officials themselves currently expect, Mr. Bullard said. When it comes to these expectations, “I would prefer that those be better aligned than they are.”

3. Scott Sumner writes,

The 4-week moving average of layoffs came out today at 287,750. Total civilian employment in September was 146,600,000. The ratio of the two, i.e. the chance of being laid during a given week if you had a job, was below 2 in 1000. That’s only happened once before in all of American history–April 2000.

However,

We are even seeing a lower employment/population ratio in the key 25-54 demographic, compared to seven years ago.

Read his whole post.

On (1), I would note that a few years ago wage growth was violating the Phillips Curve on the high side, and now it is violating the Phillips Curve on the low side. And yet mainstream macroeconomists stick to the Phillips Curve like white on rice. I would emphasize that the very concept of “the” wage rate is a snare and a delusion. Yes, the Bureau of Labor Statistics measures such a thing.

Instead, think of our economy as consisting of multiple labor market segments, not tightly connected to one another. There are many different types of workers and many different types of jobs, and the mix keeps shifting. I would bet that in recent years the official statistics on “the” wage rate have been affected more by mix shifts than by a systematic relationship between “the” wage rate and “the” unemployment rate.

On (2), I view this as evidence for my minority view that the Fed is not a big factor in the bond market. Instead, the Fed is mostly just following the bond markets. When it actually tries to affect the bond market, what you get are “anomalies,” i.e., the failure of the bond market to do as expected by the Fed.

On (3), I think that we are seeing a Charles Murray economy. In Murray’s Belmont, where the affluent, high-skilled workers live, I am hearing stories of young people quitting jobs for better jobs. On the basis of anecdotes, I would say that for young graduates of top-200 colleges, the recession is finally over. The machinery of finding sustainable patterns of specialization and trade is finally cranking again.

In Murray’s Fishtown, on the other hand, the recession is not over. I would suggest that we are seeing the cumulative effects of regulations, taxes, and means-tested benefits that reduce the incentive for firms to hire low-skilled workers as well as the incentive for those workers to take jobs. As Sumner points out, President Obama’s policies have moved in the direction of making these incentives worse.

Mark Thoma on the State of Macro

Posted on by

He writes,

The problem with macroeconomics is not that it has become overly mathematical – it is not the tools and techniques we use to answer questions. The problem is the sociology within the economics profession that prevents some questions from being asked.

But I see these as the same problem. The sociology of the profession essentially forced anyone who wanted to have an academic career to engage in mindless mathematical self-abuse. If the sociology of the profession had been better, very different sorts of articles would have been published in journals, very different sorts of economists would have earned tenure at major universities, and very different sorts of techniques would have prevailed. And don’t just blame Lucas. Fischer is every bit as much of a villain.

Asymmetric Employment Adjustment

Posted on by

From a paper by Ilut, Kehrig, and Schneider.

Suppose hiring after good news is slower than firing after bad news. A bad aggregate shock that lowers the mean of the distribution of private signals then has two effects. On the one hand, the mean signal is lower so hiring falls on average. On the other hand, the typical signal now brings bad news and thus generates a stronger employment response. As a result, dispersion also increases.

Asymmetric adjustment to provate signals leads to a number of predictions beyond the comovement of aggregate volatility and cross sectional dispersion. In particular, it should induce negative skewness of employment growth in both the cross section and the time series. Moreover, firms’ actions in anticipation of relevant fundamental shocks should depend on the sign of the shock realization: for example, bad (good) productivity realizations should be preceded, on average, by large drops (small increases) in hiring. We verify both sets of predicitions in Census data.

…One posssibility is that the logistics of the hiring process directly make hiring more costly than firing. This could be, for example, because hiring new workers is subject to costly search, whereas firing is free. A second candidate for asymmetric adjustment comes from information processing: if firm decision makers are averse to Knightian uncertainty (ambiguity) and are uncertain about the quality of signals, then it is also optimal to respond more to bad news.

From a PSST perspective, I think that lots of new information, which could include good news as well as bad, could lead to a recession. We have figured out that previous patterns of specialization and trade are no longer sustainable, but we are not sure where the new ones are.

Alex Tabarrok on Labor Market Flexibility

Posted on by

He writes,

more than half of current workers have jobs that are new since the end of the recession. A majority of workers have new jobs, some workers have wages that are increasing (and thus a fortiori not downwardly rigid) and quite a few workers have flexible wages due to piece rates, commissions, bonuses and so forth. Not all of these categories perfectly overlap. Thus, the scope for nominal wage rigidity as an explanation for current problems appears to be small.

In macroeconomics, the economy is one enormous GDP factory with one price and one wage. But macroeconomics is, in my view, misguided and misleading.