Tuesday, December 06, 2016

More about the Econ 101 theory of labor markets

I've received many good responses to my post about the "Econ 101" (undifferentiated competitive partial equilibrium short run) theory of labor markets. But there are a couple responses that I got many times, which I think are worth discussing in detail.

To recap: I pointed out that the Econ 101 theory can't simultaneously reconcile the small sizes of the labor market responses to immigration shocks and minimum wage increases. I said this implies that the Econ 101 theory isn't a good way to think about labor markets, and that instead we should pick another go-to mental model - general equilibrium, or search and matching theory, or monopsony, etc.

Two common responses I got were:

1. "Immigration doesn't just shift labor supply; it shifts labor demand too."

2. "Labor is probably very heterogeneous, so one S-D graph shouldn't have to fit both these facts at once."

The first response is a very good one. I like it a lot. What it means is that we should always think about labor markets in terms of general equilibrium, not partial equilibrium. 

An essential part of partial equilibrium thinking (or as David Andolfatto likes to call it, "Marshall's Scissors") is that the supply and demand curves can shift independently of each other. This is not apparent just from drawing the graphs, but if you think about it, it's obvious. If any shock - a storm, a policy change, a change in the price of a substitute or complement, etc. - shifts both curves in the same direction, the effect of shocks on both price and quantity will be indeterminate. Econ 101 doesn't allow you to reason from a price change, but it should allow you to reason from a price change and a quantity change.

Immigration shifts labor demand because of general equilibrium effects. Realizing that more labor is now available in a certain city, more companies will move there or start up there, in order to take advantage of the demand for goods and services from the new immigrants. That will push the price of labor back up to where it was before the immigrants came. If that shift happens quickly, studies won't show any sizable dip in wages from the influx of immigrants. General equilibrium is harder to think about than partial equilibrium (quick mental exercise: what are the general equilibrium results of a minimum wage hike?), but in the case of labor, the evidence shows that the simpler theory just won't cut it.

OK, on to response #2. It might be the case that immigrants don't compete directly with native-born workers, but fill economic niches that previously were mostly left unfilled. It might be that if we sliced the data in just the right way, we could find a slim subgroup of people whose wages are clobbered by immigration. In fact, George Borjas has tried to do exactly this. But any properly trained empirical researcher will recognize this as data mining. 

Meanwhile, some people claim that minimum wage and immigration affect two very different segments of the labor market. That's certainly possible, but both types of studies are usually about low-wage, low-education workers. Many limit themselves to teenagers, and find the same thing. How much more granularity do you want?

The problem with invoking extreme heterogeneity to explain seemingly incommensurate labor market facts is that it turns predictive theory into useless just-so stories. Assuming unobservable characteristics - or mucking around in the data til you find some that seem to work - introduces free parameters into the model, meaning it can never really be compared with data. Heterogeneity assumptions are just the proverbial epicycles. Without some assumptions about which segment of the labor market the theory applies to, the Econ 101 theory becomes totally useless.

So while some degree of labor market heterogeneity is real, the more it gets invoked, the more it weakens the overall theory. Theories should always be penalized for sticking in more parameters.

Anyway, those were two thoughtful and interesting categories of responses. I also got a few that were...well...a little less deep.  ;-)


  1. Noah: so on a mostly unrelated topic, I'm wondering why manufacturing jobs would command higher wages than service jobs, especially since manufacturing jobs have to compete over a broad geographic area and many service jobs (retail, restaurants, perhaps tourism) don't. Just in case you were looking for topics to blog about...

    1. Perhaps higher wages in manufacturing jobs are simply a result of unionization, and unionization occurs more in manufacturing because big plants, with many people working together, are more easy to unionize?

    2. My hypothesis is that service jobs, unlike manufacturing, cannot be geographically relocated and yield the same results. Typical 101 example- you cannot export a haircut. This leads to low skill manufacturing jobs being outsourced to low wages countries like Mexico and China, while you cannot outsource low skill service jobs like waiter-ing. If you look at any high skilled service job in the medical or academic professions, many of these are high-paying. There are just more observable low-skilled service jobs as low-skilled manufacturing jobs have been exported abroad.

  2. Actually you can show the increase in demand due to immigration by introducing long run. In the short run, immigration increase the labor supply causing wages to decrease for a while. This prompts firms to expand and in the long firm can do this by increasing capital (plant size), causing the demand curve to shift to the right. The effect on wages will be zero if supply and demand curve shift by the same amount. There you go--Mariel boat lift didn't cause wages in Miami to go down!

    An empirical example of this is expansion of meat rendering plant in Iowa because of the availability of immigrant workers. Those plant would not have started without availability of immigrant workers because Iowa is getting depopulated by the emigration of the young people out of the state.

    1. "In the short run, immigration increase the labor supply causing wages to decrease for a while. This prompts firms to expand..."

      Is that really what prompts firms to expand? Or would it be more demand for their goods? Now if *I* were producing something, I would only expand in response to increased demand for my goods.

      Hopefully, I'd be able to foresee the increased demand in the future, and get ahead of the train.

  3. My mind is focusing more and more centrally these days on the idea of "monopsony" shock ("shock" is more about sales then science; but that seems necessary as with anyone else with the progressive econ crowd) -- as the unifying concept to explain all exploitative wages.

    Monopsony shock + immigration of low skilled workers = labor market death to half the workforce -- no exaggeration.

    !00,000 out of my guess 200,000 Chicago, gang age males are in drug dealing gangs (and responsible for 80% of gun homicides) because immigration WITHOUT COLLECTIVE BARGAINING (necessarily buttressed by centralized bargaining) prevents the labor market from clearing.

    Foreign raised workers will work for $400/wk ($40 below 1968 fed min -- double the per cap income later) while Am raised will not. Ditto for McDonald's which pays more like $300 in most of country.

    My old taxi job used to pay $800/wk -- now $400 (before Uber!). All of above because of immigration AND NO COLLECTIVE BARGAINING. Collective bargaining may not apply as directly to my old (28 year) job but a decent union density would mean a country in which workers whose pay relies on gov would be taken care of.

    Immigration ("insourcing") or not -- outsourcing or not -- 6% private union density is like 20/10 BP -- it starves every other healthy economic and political process. Can be reversed state by progressive state adding enforcement power to fed defined but unenforced illegality of union busting (no pipe dream).

    Re-constitute normal union density and unions will be your social cop on every corner -- goodbye "reduced progressive tax, dereged industries and services espec' financial and the eroded minimum wage."

  4. Methodology 101, economic filibuster, and the mother of all excuses
    Comment on Noah Smith on ‘More about the Econ 101 theory of labor markets’

    The ancient Greeks started science with the distinction between doxa (= opinion) and episteme (= knowledge). And scientific knowledge is well-defined by material and formal consistency.

    Obviously, economics is a failed science, that is, there is NO knowledge that satisfies scientific criteria. This means, economists are forced to backpedal methodologically or in Blaug’s words ‘to play tennis with the net down.’

    How this is done? Watch Noah Smith: “What falsification really means ― or should mean, anyway ― is that a theory is shown to not work as well as we’d like it to under a well-known set of conditions. So since people have different expectations for a theory ― some demand that theories work with high degrees of quantitative precision, while others only want them to be loose qualitative guides ― whether a theory has been falsified will often be a matter of opinion.”

    Here you have it, we are back at opinion and all scientific retards happily agree because opinion means nothing other than (i) anything goes, and (ii), the coexistence of false theories. And this is what we actually have: the four main approaches Walrasianism, Keynesianism, Marxianism, Austrianism are mutually contradictory and axiomatically false.

    A scientist is supposed to abandon a falsified theory. This does not happen in economics. Morgenstern reminded his fellow economists long ago: “In economics we should strive to proceed, wherever we can, exactly according to the standards of the other, more advanced, sciences, where it is not possible, once an issue has been decided, to continue to write about it as if nothing had happened.”

    Walrasianism, Keynesianism, Marxianism, Austrianism is provable false, i.e. materially and formally inconsistent, but they are still around ‘as if nothing had happened’.

    Because falsification in the original sense does not happen, the heap of scientific rubbish grows with every peer-reviewed issue of ranked quality journals. All grand debates end where they started, that is, in the morass of conceptual confusion, undecidability, impenetrable mishmash, category error, and inconclusiveness. As Clower put it: “... we know little more now about ‘how the economy works,’ ... than we knew in 1790, after Adam Smith completed the last revision of The Wealth of Nations.”

    Economists cannot explain how the economy works but they can explain why economics does not work. Here is the mother of all excuses: “Economics is a strange sort of discipline. The booby traps I mentioned often make it sound as it is all just a matter of opinion. That is not so. Economics is not a Science with a capital S. It lacks the experimental method as a way of testing hypotheses. . . . There are always differences of opinion at the cutting edge of a science, . . . . But they last longer in economics . . . and there are reasons for that. As already mentioned, rival theories cannot be put to an experimental test. All there is to observe is history, and history does not conduct experiments: too many things are always happening at once. The inferences that can be made from history are always uncertain, always disputable, . . . You can’t even count on a long and undisturbed run of history, because the ‘laws’ of behavior change and evolve. Excuses, excuses. But the point is not to provide excuses.” (Solow)

    See part 2

  5. Part 2

    Or: “We know all that. Nothing is perfect … The assumptions are reasonable. The assumptions don’t matter. The assumptions are conservative. You can’t prove the assumptions are wrong. The biases will cancel. We can model the biases. We’re only doing what everybody else does. Now we use more sophisticated techniques. If we don’t do it, someone else will. What would you do? The decision-maker has to be better off with us than without us … The models aren’t totally useless. You have to do the best you can with the data. You have to make assumptions in order to make progress. You have to give the models the benefit of the doubt. Where’s the harm?” (Freedman)

    Economics could be ignored as Circus Maximus were it not for the fact that the scientific incompetence of economists is the ultimate cause of mass unemployment, deflation, depression, stagnation. No state/society can afford the intellectual deadweight of its economists.

    Let us come to the methodological key point. Standard economics is built upon this set of foundational hard core propositions: “HC1 economic agents have preferences over outcomes; HC2 agents individually optimize subject to constraints; HC3 agent choice is manifest in interrelated markets; HC4 agents have full relevant knowledge; HC5 observable outcomes are coordinated, and must be discussed with reference to equilibrium states.” (Weintraub)

    The representative economist has not realized it, but methodologically these premises are forever unacceptable. It should be pretty obvious that the neo-Walrasian hard core contains THREE NONENTITIES: (i) constrained optimization (HC2), (ii) rational expectations (HC4), (iii) equilibrium (HC5). Every model that contains a nonentity is A PRIORI false. In practical terms: as soon as the word equilibrium/disequilibrium appears in an economic paper it can be thrown into the waste basket. The same holds for all other nonentities.

    Conclusion: The labor market theory is false in the incarnation as partial supply-demand-equilibrium and a fortiori in the incarnation as general equilibrium.* Methodology 101 tells us: every economist who accepts supply-demand-equilibrium as an explanation of how markets work disqualifies himself as a scientist.

    Egmont Kakarot-Handtke

    * See ‘Unemployment is high because economics is false: period, full stop, end of story’

  6. Anonymous7:43 AM

    Well, 2 things:

    1) you took the lines to be linear. If I remember correctly, this is not econ 101 assumption, just what you use to get easy math. You can have both effects true (up to a limit) with non-linear curves for demand and supply.
    2) Well, it might have been falsified in the short-term for linear lines, but not for long term (what is "long term"? what is "modest"?).

  7. Isn't the direction of causation of ' immigration shocks' likely to be the other way round?

    Aren'migrants more likely to be drawn to locations of high/rapidly growing labour demand?

  8. Agree that the simply model just won't do. The problem here is one of rates. The rate of immigration and the rate that the economy absorbs them by creating new jobs. If the rate that new jobs are created lags behind the immigration rate then this will depress wages as there will be surplus workers waiting to get employment and will likely take jobs at lower wages rather than wait longer for jobs with better wages.


  9. In all fields of learning, the preferred mode is to see what happens, record it, then draw theoretical conclusions. But not economics, which is largely "here's a theory that kinda makes sense in optimal, non-existent conditions: let's try to superimpose it on reality".

    I don't understand why economics doesn't simply move to a more data-driven, "x and y moved, what happened to z" model. We have the data, we have the processing power.

    It should be fairly easy to construct new and accurate economic theory based on what actually happens out there. After all, Renaissance Tech has been netting 20% a year doing this for decades.

    Let's get with it and make economics actually useful.