What Do Liberal and Conservative Economists Agree About?

Dean (Biden) Baker and Kevin (Dow 36000) Hassett have a very nice piece in today’s NY Times focused on new evidence documenting the long run costs of short run job loss due to recession.  Unemployment raises suicide risk and divorce risk.  The new empirical research pays careful attention to trying to establish whether the correlation between unemployment and bad later outcomes is a causal effect.   It is possible that there is some  omitted 3rd factor that causes both events.

The recent collapse of Dewey and LeBoeuf offers an example of how modern applied economists seek to establish causal effects. In the year 2015, a researcher will be able to study suicide and divorce outcomes for those who have recently been fired by the law firm Dewey and LeBoeuf  .  These workers didn’t expect to be fired and this was a top law firm so nobody can claim that the  lawyers and staff were “losers”.   The control group might be similar workers (same age, gender and ethnicity) who remain employed at other law firms.  Since social scientists can’t randomly assign the treatment of “job loss” , this is the type of “natural experiment” that labor economists use to test the important hypothesis that “unemployment isn’t a vacation”.

This new research counters  the old half joke that  I was taught in graduate student at a Midwest university that  “unemployment is a vacation”.  This joke is true  if you have savings in the bank, and labor markets are competitive.  Under these assumptions you can easily transition to your next job without much pain.  This new empirical research on the long run cost of recessions is important and provides an intellectual underpinning for Dr. Krugman’s demand side push.   An alternative policy strategy is to  lower migration costs so that unemployed households can move to local labor markets where there is job growth.

While many RBC readers appear to believe that neo-classical economics is a smoke screen for pushing a core ideology covered with a lot of math and ivy league degrees, I believe that this Baker/Hassett piece (written by a well known Liberal and Conservative) highlights the consensus and the scientific approach at the heart of modern economics.  We are making slow progress but the target is moving and the questions are hard and game theory (i.e the strategic game between Bernanke, Congress, Obama, Europe, China and Wall Street) is hard!

Author: Matthew E. Kahn

Professor of Economics at UCLA.

20 thoughts on “What Do Liberal and Conservative Economists Agree About?”

  1. I’m almost completely ignorant of economics, neo-classical or otherwise. So perhaps I’ve been unclear.

    My concerns are two-fold. First, I have enough understanding of how systems a lot less complicated than our economy can get stuck in highly sub-optimal stable equilibria — or to put this another way, circumstances in which hill-climbing alone will fail as an optimization strategy — to be unenthusiastic about Panglossian tautologies to the effect that things are for the best in this best of all possible market economies.

    Second, I have enough experience with academic fields and their sociology to have seen this very circumstance, in which we get “stuck”, be enacted. So I’m a lot less enamored of long chains of logic than I used to be.

  2. “An alternative policy strategy is to  lower migration costs so that unemployed households can move to local labor markets where there is job growth.”

    One of the things I’ve been noticing is how much poorer we are as a result of the postwar embrace of worker mobility, especially since the real economy peaked with domestic oil production in the early 70s. Essentially we were sold the idea that people are and should be seen as just meat robots that can be shipped about to distant and sundry work stations in the global factory system. Those reluctant to get on the truck and follow the jobz become the poors, getting their due punishments for impeding the swirl of capital by trying to cling to the side of the bowl instead of letting themselves be flushed into the economic mainstream. Nearly every social pathology in America is worsened by the destruction (sorry, Creative Destruction!) of extended families and communities that results when workers are expected to relocate at the drop of a factory into a sunnier, more Right to Work state. Out here in the west we are enjoying the return of the work camp like the miner’s camps of old, madly fracking away on the treadmill, hoping to strike gold before the inevitable backlash crests, when already water starved states find that their buddies in the extractive industries haven’t changed a bit since the first Gilded Age.

    And up front on the pulpits, chanting the mantras of absolution for destroying the only habitable planet known to man, is a cadre of priests who call themselves economists and whose sole divine mission is to argue that, on the whole, it’s better to have a high GDP than intact communities, better to plot for “growth” than to accept that the planet is finite, and better to see the oceans exhausted and the entire web of life destroyed than to notice that the cult of economics is driving humanity to destroy itself and many other species with it.

    1. Is it only labor mobility?

      What about capitalist mobility – the ability of the owners of capital to live far removed from where their capital is invested. Even if you are willing to pay taxes to improve the community, wouldn’t you rather pay them where you live than a thousand miles away, where your factory is? So is labor stuck racing to the places that offer the biggest profits to the distant capitalist?

      And what are these “migration costs” that we want to lower anyway? One such, for example, might be accepting a poorer school system for one’s children, and there are other examples that have nothing to do with the cost of renting a van. Anyone serious about migration costs needs to explain how to handle those.

      And some simply cannot be lowered. An obvious and important cost is moving away from friends and family. That doesn’t exactly fit into economists’ schemes.

  3. Natural experiments may be the best available in the circs, but they are not ideal. The double-blind trial in medicine was developed to avoid not so much deliberate bias – if you are a bent scientist, you can rig any experiment – as unconscious bias and Hawthorne effects. There are for instance many ways to select the control group for the fired employees of Dewey and LeBoeuf, and no other firm will be an exact match. The best bet here might be for separate, non-communicating experimenters to select different control groups.

    You have to collect data on the subjects. Suicide and divorce are pretty objective indicators, but to get a full picture you probably want to ask about suicide attempts, depression, libido, behaviour problems of children, and so on. That involves talking to people, and unconscious bias comes in here as well. Using hired survey workers who don’t know or care about the research issue protects against the bias, but reduces the chance of truthful replies. To find out about libido requires trust and empathy: a human connection inseparable from bias on the research question.

    One of the claimed payoffs from electronic medical records is the possibility of data mining research on this sort of thing. You could plot consultations for STDs and depression against mentions of recent job loss. The privacy implications are difficult.

    1. Of course lots of other fields rely heavily (or even exclusively) on observational data. Evolutionary biology and ecology, for instance, or cosmology.

    2. The privacy issues can be managed by selecting a large sample and anonymizing the records prior to analysis. Privacy issues are most problematic when individuals can be identified. If studies are designed in a way that mitigates the identification issues, the problem is managed.

    3. Good points, but absolutely nothing new. Of course one would do a set of triple-blinded randomized controlled trials, but one frequently doesn’t have that opportunity, and that’s also true in medicine.

  4. What if you believe that neo-liberal economics is a smoke screen for core ideology with poor math and 2nd rate degrees (despite the universities that granted them)?

  5. So where in the US are these fabulous job markets that can absorb the millions of people who are currently unemployed if only they could move there?

    That solution only works if unemployment is structural, whereas in reality almost all sectors and geographical ares are hurting due to lack of demand. Conservatives want our unemployment to be structural because that fits their narrative, not because it is true. Their’s is only a solution for a problem we don’t have.

  6. I find the half joke ironic especially when followed by the acronym RBC which, to me, stll stands for real business cycle.

    I am an economist with low respect for economists, but calling Hasset one of us is going too far.

    Finally the natural experiment scintific approach is at most one of the multiple hearts of the too flexible but shar tongued hag fish of modern economics (don’t google the metaphor on a full stomach).

  7. many RBC readers appear to believe that neo-classical economics is a smoke screen for pushing a core ideology covered with a lot of math and ivy league degrees

    Actually, it often doesn’t require much math any more. A lot of it is just fatuous ideological dogma without the cover of a model, like this proposed alternative to Krugman:

    An alternative policy strategy is to lower migration costs so that unemployed households can move to local labor markets where there is job growth.

    This only works as a statement of dogma. If you try to examine it in any way – including with the use of math – you’d just expose it for what it is.

  8. Baker and Hassett are not doing economics. They’re doing sociology, There’s nothing wrong with that, but it would have been useful for them to talk a sociologist before going into print.

  9. An alternative policy strategy is to lower migration costs so that unemployed households can move to local labor markets where there is job growth.

    So I assume lowering those migration costs would include finding cost-effective ways for older workers to physically relocate their entire social network, right? Fascinating idea. If only workers were not actually people. Hmm, maybe that’s the problem we need to solve?

  10. @ Bloix Economics *is* sociology, focused on the social organization of production and distribution.

  11. Kahn: “This new research counters the old half joke that I was taught in graduate student at a Midwest university that ”unemployment is a vacation”.

    Is there any evidence that the Chicago guys didn’t believe this? Isn’t the Chicago/RBC attitude still that the Great Depression was caused by an increase in preference for leisure?

    ” This joke is true if you have savings in the bank, and labor markets are competitive. ”

    No, it’s true if you have money in the bank. If you don’t, life gets rough.

    ” Under these assumptions you can easily transition to your next job without much pain. ”

    IOW, if it’s the late 1990’s.

    ” This new empirical research on the long run cost of recessions is important and provides an intellectual underpinning for Dr. Krugman’s demand side push. An alternative policy strategy is to lower migration costs so that unemployed households can move to local labor markets where there is job growth.”

    Frankly, anybody who thought that unemployment was not rough should be kicked out of their (econ professor) office, and forced to look for a real job.

  12. This reminds me of the study that tried to find out if PMS really exists.

    But, if you’ve got the money to spend, by all means do this rather absurd study. I guess it will create some jobs?

  13. I have to admit to being deeply puzzled by the OP. If the intent is to point out that high-quality econometric research, peer-reviewed and with good design-of-experiment to factor out the researcher’s political biases, is valuable, I agree. If the intent is to claim that the U of C / Richard Posner “Law & Economics” theories are valid, good, and useful then color me unconvinced. Similarly if we are to say that combining some statistical research on individuals’ buying preferences with standard, deeply flawed microeconomics makes a “modern economics” that we must all get behind; no thanks.

    Cranky

  14. Seriously? What well-meaning adult of any experience at all (and not blinded by a certain warped view of classical economics) ever thought sustained mass unemployment was “voluntary” or a good thing? And while Dean Baker is a serious person, Dow 36000 and the obvious methodological flaws in it are enough to demonstrate that Kevin Hassett is not.

    Beyond that, just on methodological grounds I find it unlikely that a study just of Dewey LeBoeuf (or any single firm) in comparison to going law firms would be very convincing, because it’s entirely possible that a sick corporate culture was in part responsible for the firms downfall AND also had a role in either causing angst to its lawyers or had a selective role in the population of lawyers that ended up there (i.e. the ones who could go elsewhere did). I say this knowing nothing of Dewey LeBoeuf but just imagining the kinds of arguments that could be raised against this research design.

    As for the effects of unemployment, I thought there were already fairly strong cohort effects demonstrated on life prospects of people entering the labor force in booms versus busts. In the fields of economic and sociological research that I am familiar with, the more exotic quantitative research strategies almost never pay off because they offer too many opportunities to conform the results to investigator prejudice without being articulated enough to capture all the causal channels of real life. In other words they have enough degrees of freedom to be dangerous without enough power to replicate reality. For example, when randomized controlled trial results of social interventions are compared to regression studies, the simplest regression specifications come closest to the RCT results. So if the meaning of the original post is that a broad perspective, common sense, and judiciously applied data can illuminate social problems and policy responsibility, then fine, but if it is intended that narrower and more sophisticated data analyses applied to technical issues is the way to go, then I’m not particularly inclined to agree,

Comments are closed.