The Missing Mini-Recession

So I've been thinking about this idea of the "mini-recession" -- low-duration, low-intensity periods of economic contraction -- and how the U.S. does not seem to have them, which Scott Sumner first brought to my attention in this excellent post in December.

Sumner's explanatory story was that "inertial central banks that target nominal rates and observe the macroeconomy with a lag might occasionally produce short contractions, typically 9 to 12 months." He looked at the change in unemployment rates, and found that increases beyond a small amount appear to trigger far larger recessions.

We see the same effect in the NBER's recession dating -- see the graph above, which is a histogram of the length in months of economic recessions (negative) and expansions (positive). Short recessions are very rare. Instead, recession lengths appear tightly clustered around 12 months, with a pronounced negative skew, i.e. there are more recessions of length > 12 m. than < 12 m. Similarly, expansion lengths are more loosely clustered around 24 months, with a positive skew.

This is the signature of the absence of mini-recessions.

My method to arrive at the graphed data above was pretty basic. I downloaded the monthly NBER business cycle dating data from fred, calculated using Excel the length of each recession or expansion, and then used Excel's "FREQUENCY" function to create the histogram. (The graph is in Google Docs so it can engage you interactively.)

It's more or less impossible to explain this without a lagged policy response.

Six Signs of the DisConnect

"Washington is out of touch" -- that's a line one hears a lot from politicians who want to play the outsider. Is it true?

Tyler Cowen of the "Marginal Revolution" blog points us to a disturbing article, "Bubble on the Potomac," in Time magazine which discusses the extent to which that has become true, economically speaking, for the District of Columbia and its environs -- defined, for the purposes of this post, as the Washington-Arlington-Alexandria metropolitan statistical area.

Although I tend to regard the economics coverage of such newsmagazines as sloppy, I was very impressed by the quality of this article. I'd just like to further the point, demonstrating that there is substantial evidence suggesting growth of the federal government in Washington, D.C., and moreover, disproportionate prosperity in the D.C. area.

So here are six signs that D.C. has economically "DisConnected" from the rest of the nation:

#6: Home prices in D.C. have risen 43% over a nationwide average since 2000.

#5: Federal government employment has grown 4% in the District and D.C. area -- a net 40,000 D.C. government jobs -- since President Obama took office in January 2009.

#4: The rise of the contractor conceals the extent to which the federal government has grown in the D.C. area, supporting the argument of the Time article.

#3: The unemployment rate of the D.C. area is consistently two-thirds of the nation's unemployment rate.

#2: The average American's income is 63% of what the average D.C. area resident makes.

#1: D.C. employment contracted only one-fourth as much as the nation's did, and their payrolls are already above their pre-recession highs.

Jersey Workers

Paul Krugman has a brief post up on his blog pointing out that the recovery in nonfarm payrolls in New Jersey has been unusually sluggish. He writes that he's "actually not sure why NJ is doing so much worse than New York or Pennsylvania, and I doubt that Christie has much to do with it, but he’s the one trying to claim credit for … what?"

N.J.'s slow growth in nonfarm payrolls, in fact, can largely be explained by changes in government employment since Christie took office in January 2010.

Here's a graph of New Jersey's nonfarm payrolls, its private payrolls, and New York's nonfarm payrolls, indexed to 100 at January 2010, when Christie came into office and also at the trough of the New Jersey payroll numbers.Notice, in fact, that the gap really emerged in late 2010, when the major cuts in government employment occurred. Had N.J.'s government payrolls held constant, the N.Y.-N.J. growth gap would have been half its current size.

Here's a graph which gives you a sense of the magnitude of contraction in government employment in New Jersey as compared to New York. Both states cut back towards the end of the recession, but the former far more so than the latter.I suppose one can read two different conclusions into this post. (1) The Jersey Comeback is happening, but because of public sector cuts, you see it in the private numbers, not the nonfarm numbers, or (2) the Jersey Comeback is being restrained by imprudent public sector cuts. That is a political and normative decision which I'll refrain from making here.

Major Irrationality

In a matter of months, I will head off to college. One of the things I've been thinking about recently is how students pick a major -- I'm certain as to mine -- and in particular, I've been asking myself to what extent people think rationally when they are choosing a major.

I don't think they really do. In fact, I think it's almost impossible for such a choice to be made rationally, i.e. there are tremendously strong forces driving such irrationality.

(I find behavioral economics fascinating -- Daniel Kahneman's Thinking Fast and Slow is on my summer reading list, and it should be on yours too -- but for whatever reason I don't write about it too often. This post will begin to rectify that oversight.)

This matters, because whether you regard the major selections of college students as rational or irrational should determine whether you do or do not sympathize with them when they, immediately after graduation, complain about low pay or scarce opportunities for remunerative work, as was discussed in this recent New York Times feature. (The whole thing is excellent reading, in fact.)

If you regard the choice of major as a rational decision-making process, then it stands to reason that all fresh graduates from art programs were and are in full knowledge of what side of the tradeoff they are on when it comes to the probability distribution of lifetime incomes. Therefore, when they complain that they can find no work which values their degree, you shouldn't sympathize with them -- they understood in advance the likelihood of this happening, they should have been prepared, they had other options.

If you regard the choice of major as a irrational decision-making process, then it stands to reason that the newly-minted artists weren't in full knowledge of the tradeoffs -- they didn't recognize or correctly appraise the tradeoffs they faced. Therefore, when they complain that, hey, there's no good work out there for an artist, then there is room for sympathy. In fact, I would argue there could very well be room for a public policy intervention; the state should try to encourage ("nudge") more rational decisions. Or, if that proves impossible, one might argue that such graduates have a valid moral claim for income support from the state. It's not that one would want to encourage such choices, but given that the trade-offs are not understood, changing them doesn't change behavior -- and there are analogous situations in which we forgive or otherwise lessen the consequences of bad decisions given the irrational circumstances.

Consider the U.S. criminal justice system. In most states, murder is considered in "degrees" of reprehensibility. The penalty for first-degree murder -- those which are "willful and premeditated" -- is far more severe, possibly even death, than second- or third-degree murders -- those which are not willful or premeditated, and those which are committed in a "heat of passion...[that would] cause a reasonable person to become emotionally or mentally disturbed," respectively. The reason we lessen the punishment isn't because someone killed in cold blood is any less dead than someone killed in a "heat of passion." It's because juries make an informed judgment about the capacity of the murderer to make a rational decision under his or her circumstances.

I'm not going to argue, for the rest of this post, that artists, English majors, etc. have a valid claim to state aid. Rather, I want to think about if these choices are irrational, and if they are, why these individuals may be making irrational choices.

Regret is the telltale sign of a decision-maker understanding post hoc that his or her decision was irrational. And we see a lot of regret in the New York Times article:

“As an 18-year-old, it sounded like a good fit to me, and the school really sold it...I knew a private school would cost a lot of money. But when I graduate, I’m going to owe like $900 a month. No one told me that.”

There's even more in the associated video.

When you consider the forces acting on college students, it's hard to see how a rational decision of major could even be made, even when they sit down to think about it.

First, the fact is that they are considering utility gains well into the future as compared to present utility -- let's face it, math and science requires more work than English and art -- and this invites a common force of cognitive bias, present-biased preferences.

Second, this cognitive bias is magnified by a very big framing problem. When you go to college, your "frame" -- the cultural influences under which you make a decision -- is shaped by who you are taking classes with, the professors teaching you, the school environment in general...everything that, in a rational decision-making process, should be irrelevant.  What is relevant is how you calculate future utility according to a tradeoff about income, and your "frame" doesn't include this, or underweights it because of the college environment.

The framing problem is exacerbated by the fact that your frame of reference shifts dramatically between college and the "real world" -- what matters, and how you value it, changes. This makes your utility calculations incorrect. In fact, because this future life is so hard to imagine, that inability to imagine may even be the underlying force causing people to undervalue future utility. This could be an amazing study: read people a short news story about a tragedy abroad -- say, a famine in Africa -- and ask them to donate a certain amount from $10 you give them there. Then, with another group, show them a color photo which depicts exactly what is described in the text and ask them to donate; with another group, show them a video which provides the same information as the text or photo. (This is my take on an old theory of Adam Smith from the Theory of Moral Sentiments.)

My prediction is that where D = average donations, D_video > D_photo > D_text. That would be solid evidence that "imagination matters " when it comes to framing and cognitive bias.

The Principled Regulator?

Arnold Kling has a fresh perspective on how regulation should be articulated, calling for "principles-based regulation" rather than "bright-line regulation":

When we think of regulation, we think of specific rules that spell out the boundaries between what is approved and what is forbidden...What I want to propose is an alternative approach...legislation would lay out broad but well-defined principles that businesses are expected to follow.

Kling recognizes a need for some bright-line regulations in some instances, but he implies that regulation (in particular, financial reg.) would be better off if we made a broad move toward principles and away from bright lines.

I agree with him in some respects -- particularly that bright-line regulation is made toothless by creative circumvention efforts which violate the spirit, but not the text, of a regulation. Kling's essay is motivated by the failure of Dodd-Frank to effectively codify important principles of financial regulation -- why not, then, just put the principles themselves into law?

While I like the idea, I'm very concerned that principles-based regulation may perform even worse than bright-line regulation.

My problems with principles-based regulation are that could establish due process rights where it may be more effective to draw bright lines, that it could create considerable regulatory and political uncertainty, that it would make markets less efficient due to imperfect information, and that trying to establish principles goes far beyond minimalist "rules of the road."

The problem of due process

When firms violate regulations, the regulations can often be enforced without turning to courts. "Your kitchen has a rodent infestation," the regulator says. "Here's a $500 fine, and you have to close down the restaurant until the rodent infestation is removed and you are re-certified by the Department of Public Health." Only in the cases that the regulator's facts, or his application or reading of the relevant regulation, are challenged do such situations require due process -- that is, the involvement of the judicial system. "That wasn't a rodent, Mr. Regulator, and in any case, you didn't realize that rodents are allowed in some restaurant kitchens under the conditions specified under §110.A.4 of the Health Code."

Now consider the scope of expansion of due process when regulations move from bright lines to broad principles. "Your kitchen is unclean," the regulator says. "Here's a $500 fine, and you have do close down the restaurant until it is cleaned and re-certified as such." Given that such a decision has an additional interpretation step beyond the facts, this interpretation by regulators can be challenged in Court in the United States. "Is the Department of Public Health's definition of cleanliness reasonable? Did the Department make a reasonable judgement of the cleanliness in this particular instance?" The example of a rodent infestation bright line becoming a cleanliness principle may seem absurd, but as the principles become even less clear -- for example, Kling's "consumer protection," "reckless behavior," or "fiduciary irresponsibility" tests -- they will invite even more due process challenges.

The need to involve courts in regulatory enforcement strikes me as the defeat of the idea of principles itself. In furthering due process, our court system will establish precedent cases to define the details of the principle in question. And at that point, we've merely outsourced the author of lines to the courts, rather than the legislature. It doesn't resolve the question that, at some point, criteria must be identified for regulations or their applications to meet a reasonability test. Judicial involvement, furthermore, seems to me suboptimal, given the fact that legislatures and regulatory bodies are more open to the democratic process and can actively write new law without waiting for a case.

Regulatory and political uncertainty

The most salient advantages of bright lines is that they are bright, and that they are lines -- that is, you know when you're in violation of them. Firms don't necessarily know when they are in violation of nebulous principles. That may mean they are more cautious to expand or adapt in circumstances where they are uncertain of how regulation will be enforced. For example, let's turn to Kling's Capital One case -- could Capital One recommend its own payday lending service, or another one, to its low-income customer with a small credit limit? Capital One might say that the customer had a chance of benefiting from the product -- and you and I might agree, but would a regulator who thought that payday lending was usurious and exploitative? As a result, firms will have to consider whether their behavior can be construed as in violation of regulations, rather than if their behavior runs directly contrary to a bright line.

Using the same Capital One payday lending example, we can see that the views of the regulators matter under principles-based regulation. What about when the political party of administrations changed in the federal, state, and local governments? Would the enforcement patterns of the principle change? For example, would a Republican commissioner of the relevant enforcement agency be different from a Democratic commissioner in how he judges particular actions of firms as consistent or inconsistent with the principles? This is plausible, perhaps even certain to occur. Also, the uncertainty of principles-based regulation opens the door not just for inconsistent and unclear applications of the principle to specific cases, but outright biased cases. The influences acting upon the regulatory agency become a more important consideration when one expands the discretion of the regulators by making them enforce principles instead of bright lines. (Consider the potential for corruption and for bigotry, just off the top of my head.)

Imperfect information and competition

What follows from this uncertainty and due process requirement is a massive headache: imperfect information. When firms do not know how a regulator will judge, they participate imperfectly in markets -- they will put the principles into practice differently, with some, say, having more clean kitchens than others according to their own liberal or conservative definitions of cleanliness, or some being more aggressive in what financial products they try to sell.

With imperfect information, the firms closest to the regulators (read: the big firms which always hire their compliance officials from top positions in government regulatory offices) will have the best information, giving them a distinct competitive advantage over smaller firms with less information.

Just as bright line regulation creates a perverse incentive for firms to try to circumvent and game the text of the regulation, so does principles-based regulation. The problem is inherent to regulation, not to lines, as Kling suggests. I would contend, in fact, that the incentive problems are not fixed -- perhaps even worsened -- due to the considerable scope of interpretation in principles. The game merely shifts, as firm knowledge of how principles are applied becomes valuable.

Regulation and the role of the state

Finally, Kling and I are both sympathetic to the idea that "government is the problem," or is often so in economic regulation. He puts forth the idea of principles-based regulation in hope that it will reduce the regulatory burden incumbent upon firms. The regulatory codes themselves may be shorter -- something I think is important and a strong argument for a limited application of principles-based regulation -- but it does not follow that the regulatory burden has been lessened, or that the involvement of the state in commerce has contracted.

The tests themselves have changed, but the expectations of regulators have not. Kling doesn't say that the goal is to reduce the involvement of the regulatory state, just to change how it operates. Unfortunately, having the state enforce principles is a far more aggressive posture for the state than bright lines. The regulator is no longer setting "rules of the road" -- they are not the state trooper on the side of the road with the speed gun -- they are for all intents and purposes a judge. (See my discussion of due process.) That's considerably more power granted to the state.

Eurocrisis = NGDP Crisis

Paul Krugman wrote a post a while back in which he looked at the data for debt-to-GDP ratios and government spending as a percentage GDP and effectively no explanatory relationship there for what's happening in the Eurozone, contrary to popular narrative. Instead, Krugman argued, Europe had a balance of payments problem.

I don't disagree with him, but I see the balance of payments problem as fundamentally driven by nominal GDP gaps, which differ strikingly between the different Eurozone nations, as monetary policy at the ECB cannot obviously stabilize on a path trend -- or keep growing even at roughly the trend rate -- nominal GDP in the 23 Eurozone nations.

The result is massive, unfathomably large nominal recessions in countries like Spain, Portugal, and Greece -- depressions, really -- amid nominal booms above trend in Germany, Belgium, and Austria. (Remember where the Eurozone is in its business cycle, and you should recognize that to be running even a small positive NGDP gap at this time amounts to a huge positive gap in normal times.)

What's going on in the Eurozone is ultimately not a debt issue, nor an issue of structural government spending. Those have undoubtedly slowed growth in the Eurozone over the long run -- but this is not a structural problem, it is a cyclical problem, a nominal problem. (See this post of mine for more on this.) Instead, these structural problems surfaced because of the nominal problem.

Note: An earlier version of this post failed to include Portugal; I redid my analyses for several of the countries to confirm that my data had not been mislabeled, but everything appears alright now. Please let me know in the comments if anything else seems faulty. Thanks. -- ES