Fertility differences and agricultural techniques

There are times when you read a paper and you really wonder how the authors came up with the idea to check out a particular correlation in the data, because it seems to be so far-fetched. But thus a correlation can be beautiful if it also has a nice theory that comes with it.

The correlation that Alberto Alesina, Paola Giuliano and Nathan Nunn study is between current fertility and adoption of plough agriculture in history. OK, I did not think about that one. But now that they find a nice positive correlation, how could one explain it? They argue that this has to do that women and children are not particularly useful when ploughing, as strength is required. The traditional task of weeding, that fell on women and children, is not necessary with ploughing. Thus, there is a preference for fewer children that is ingrained in the culture of these regions to this day.

The unnecessary problems of the Euro

European leaders are currently struggling over a package to save the Euro, pouring large amounts of money into funds that should stabilize the fiscal situation in Greece, Portugal, Ireland and potentially other countries. It seems to me that this is a completely unnecessary problem, and all this grief could have easily been avoided with a simple change in policy.

Just look at what is happening in the United States. Several states are in serious financial difficulties and, as several times in the past, California is considering issuing IOUs, thereby essentially declaring it is insolvent. Is there any expectation that other states or the federal government will rush to California's aid because the dollar is threatened? Of course not, despite the fact that California is the largest state in the Union.

It should be the same for the Euro. None of the member countries can monetize its debt on its own, and the only reason that the Euro is threatened is that markets have an expectation that other countries will rush to help, thereby sending a message that monetary policy could be influenced by what is happening in those small countries. And why is this belief well anchored? Because European indeed rush to help (talk about a nice example of self-fulfilling expectations) and because of this silly concept that all national debt in Europe is fungible (talk about a nice example of the tragedy of the commons). Now of course it is a bit late to rectify those beliefs, but had it been clear no rescue package were in sight, those countries would probably not taken such a risky fiscal path in the first place (talk about a nice example of moral hazard). I guess that those silly policy decisions all boil down to European politics, once more (talk about a nice example where economists' advice has been ignored, and they will get blamed for it anyway).

What influences Fed presidents?

The European Central Bank is still a recent creation, so you can excuse its national governors for putting their country's interest first in the conduct of monetary policy. What about another federal central bank that is much older and whose governors territory does not necessarily coincide with political boundaries, the Federal Reserve System of the US?

Bernd Hayo and Matthias Neuenkirch have analyzed the speeches of Federal Reserve presidents over the span of twelve years and come to the conclusions that they equally represent the national and regional interests, except when it is not their turn to vote, when their region comes first. They find this by trying to fit a Taylor Rule to their positions, using a price index, a national and a regional unemployment rate. Unfortunately (and surprisingly), there are no regional price indexes in the US, which could have reinforced the regional focus of the presidents. Still, I am surprised how much they lobby for the general interest. After all, they are selected to represent their region.

You want to restrict bankers' pay

There has been and there still is much outrage about the large bonus payments bankers get. What the public does not understand is that bonus pay is a very large part of total pay, and it is so to encourage bankers to perform really well. And they certainly put in the hours. For example, bonus pay has been criticized because there is most often no "malus," but given that base pay is relatively low, this should capture it. The main criticism is aimed at the disparity of these bonus payments with respect to the average pay of a worker. This is, however, not something that should be regulated at the level of bonus pay, but through redistribution with income taxes. In this regard, whether it is regular pay or bonus pay makes no difference. So, should then bonus pay in banking be left unregulated?

John Thanassoulis does not think so. He argues that as bank compete for top bankers and try to shift the risk on them, they end up paying them too much and all in bonuses. This is optimal for the bank as it lowers its costs right when things get critical. But as a consequence, the bank gets too much into risky activities, as competition for bankers drives bonuses up higher than socially optimal, especially if there is a contagion risk of default for other banks. So you want a regulator to limit bonuses, but in a flexible way, or the benefit of having bonuses in the first place gets eroded. Indeed, it is the top brass that sets the bank level risk, whereas other employees all the way down to secretaries (who also get bonuses) are less influential, even collectively, on the aggregate risk. Thus the idea is not to cap bonuses individually, but at the bank level as a proportion of the balance sheet (which is what matters in terms of default). The pay structure would then presumably be readjusted by the bank, relying more on bonuses where it matters the most. Taxing bonuses has no risk impact, though, except for reducing bankers' pay.

Another possibility could be the dynamic incentive accounts I mentioned before.

Modelling without theory

In Economics, we have adopted the scientific method much like other sciences. As we teach our students, it consists of the following steps

1. Observe regularities in the data.
2. Formulate a theory.
3. Generate predictions from the theory (hypotheses).
4. Test your theory (is it consistent with data?)

In the context of Economics, the goal of the procedure is not only to explain why regularities in the data happened, but also to build a theory that is useful in predicting the consequences of particular policies or institutional designs.

David Hendry just published a paper about the scientific method in Economics that appears to fly in the face of what I just described. Here is an attempt to summarize his stand, and I apologize for quoting quite liberally:

1. Specify the object for modeling, usually based on a prior theoretical analysis in Economics. An example of such an object is y=f(z).
2. Defining the target for modeling by the choice of the variables to analyze, y and z, again usually based on prior theory. This is about deriving the data-generating process of the variables of interest, or fitting an equation with some statistical procedure.
3. Embed that target in a general unrestricted model (GUM), to attenuate the unrealistic assumptions that the initial theory is correct and complete. The idea is to add other variables, lags, dummies, shift variables and functional forms to improve the empirical accuracy of the initial model.
4. Search for the simplest acceptable representation of the information in that GUM. Or, now that the model has become huge (and may contain more variables than data points), let us get rid of some of them without loosing too much in accuracy.
5. Rigorously evaluate the final selection: (a) by going outside the initial GUM in step three, using standard mis-specification tests for the ‘goodness’ of its specification; (b) applying tests not used during the selection process; and (c) by testing the underlying theory in terms of which of its features remained significant after selection.

In other words, this amounts to take some linearized version of some theory, through data at it, whether theoretically relevant or not, then massage it until it fits the data.

A part from the fact that this is really the blueprint for an automated data mining exercise that is not driven in any way to answering a particular policy question, this procedure not only disregards the scientific method, but also Occam's Razor and the Lucas Critique. What use is it to learn that the CPI follows a polynomial of degree five with three lags on exports of cabbage, the number of sunny days, 25 other variables and three structural breaks (not an actual example used by Hendry, but it could)? If you want to make some very short term forecasts, that may be accurate, and this method is abundantly used in the City or Wall Street by neural networks "experts." But when it comes to advising policymakers, you need to have some Economics, and by that I mean economic theory, to explain why economic agents behave in such a way and what an intervention would lead to.

The scientific method starts with the observation of the data. Hendry dismisses this with a slight of hand, stating that stylized facts are "an oxymoron in the non-constant world of economic data." What if there are constants in economic data? In fact there are plenty, and this is what theories are trying to explain. Has Hendry never observed something in his surrounding that he then tried to explain? Or does he really spend his days feeding linear equations into his computer to see what it can come up with with his database?

Such papers, especially by people who enjoy respect like Hendry does in the UK, deeply upset me. To top it off, there are 33 self-citations.

The spaceship problem

Suppose you have to plan a very long term mission in space. It will last for many years, and you need to provide a group of people the means to live in a hermetic environment. You do not have access to Star Trek technologies like warp speed, replication and teleportation. Your population can reproduce, but life length and quality of life depends on resources and population density. How many people should be on such a mission? This is known as the spaceship problem. Of course, economists have something to say about this.

Pierre-André Jouvet and Grégory Ponthière are not going to solve the problem, there are too many biological and physical constraints, but they point out that the solution will yield solutions that contradict utilitarianism. They focus on the trade-off between the number of people and their life length. Indeed, longevity impacts population size and thus density. They assume that a social planner uses the sum of residents' utilities as a criterion and, unfortunately, that resources are unlimited, which makes the paper stray away from Economics.

What Jouvet and Ponthière really want to do it is compare different social welfare criteria in this environment. The Classical Utilitarian, for example, sums the utility of all individuals, the Average Utilitarian only the living ones. In a model without reproduction and a finite mission time, Classical Utilitarianism yields a small population living very long, while the second may want to have a large population that lives for a short time. Add reproduction to the mix and anything can happen depending on parameters values and initial population size. Make the mission life infinite, and the authors run into problems and need to define additional social welfare parameters. That is mainly due to the fact that there is no discounting, and infinitively lived economies and ill-defined.

What do I learn from this exercise? It is not very clear, except that social welfare criteria matter, adding utilities gives us a lot of trouble and that discounting is essential. But we knew that already, even when the spaceship is called Earth.

The impact of job search monitoring

Unemployment insurance is thought to be ripe with abuse, especially as job seekers do not appear to seek jobs that much. For example, time use surveys have established that they spend very little time on job search, the median being even zero minutes on any given day (see previous post on this). So it seems natural that you want to make sure the job seekers give sufficient effort to obtain benefits. But how effective is such monitoring?

Bart Cockx and Muriel Dejemeppe study this imposition of stronger monitoring on long term unemployed workers in 2004 in Belgium. For all practical purposes, there is no limit to the duration of unemployment insurance benefits there, so it seems baffling that only recently has there been some serious monitoring, and this only happens after eight months of unemployment. And then, it is only in the form of a stern letter threatening monitoring. Before this, monitoring was only targeted on those unemployed for more than 21 months...

This means that before 2004, the unemployed had essentially free rein. After that, there is a supposedly credible threat of monitoring after eight months. In Wallonia, the letter is followed up two months later with a counseling session. In Flanders, there is no systematic counseling That is probably what can be considered a clean natural experiment of a transition from no monitoring to some monitoring. What impact did it have? In Flanders, the transition to employment after eight months of unemployment increased by 28%. In Wallonia, this is 22%, the authors conjecture it is lower despite the more credible threat due to worse labor market conditions. But especially in Flanders, it appears the shorter duration implies that workers end up with worse jobs than before, both in terms of wage and duration of employment. And Walloon females are more likely to transition into sickness insurance, which has the same benefits as unemployment insurance.