Job market seminar: Duhaime-Ross

Leave a reply

The most recent job market seminar that I attended was from Alix Duhaime-Ross (who happens to have the perfect initials for a PhD student: A DR). Duhaime-Ross works on empirical microeconomics, an area of economics that has been popularised by books such as Freakonomics. This approach to economics is heavily data driven, and Duhaime-Ross focuses on labour economics and the economics of education.

Duhaime-Ross’ presentation focused on the effects of educating school-aged children in the dominant language (instead of a minority language). The context here is Bill 101, a law that was passed in Quebec in 1977 that required all children of immigrants to be educated in French language schools (previously a majority of immigrant children were educated in English). An exception was allowed for children of immigrants who had at least one parent who had been educated in English in Canada. This law is still in place today, so that if, for example, I moved to Quebec and had children they would be required to attend school in French. Any children who had already started school in English were allowed to finish their schooling in English, so the law only affected new students.

What effect did this law have on the children who would be educated in English without the law, but are not being educated in French? It might improve the student’s future employment prospects because they will have better French skills in a French speaking province, or it might harm them because they might do worse in a French school than they would have in an English school (perhaps leading to them not going to university, for example).

Normally a question like this would be very hard to answer because parent’s will usually try and make the best choices for their kids, creating a selection bias. This means that Duhaime-Ross can’t just compare English educated children in 1976 with French educated children in 1977, for example, because the characteristics of these children will differ (and other changes might confound things as well). What we really need to do is exploit the mandatory nature of the law, and create a natural control group to compare outcomes with.

Duhaime-Ross does exactly this. She begins by looking at a control group – those children with only one foreign born parent. These children usually have one parent who was educated in English in Canada, so are exempt from the law. The change in outcomes over the period of time when the law was enacted amongst this group of children forms a control group. The change in outcomes for children affected by the law can be compared to the change of outcomes in the control group to isolate the effect of the law.

But, there still might be some other changes that are not accounted for with this process. What if the changes in the law caused French school class sizes to become larger and therefore reduce the quality of French language education? This would bias the estimates from the previous paragraph. Duhaime-Ross therefore used a second control group consisting of native French speaking who would always be educated in French regardless of the law.

There are, therefore, three groups. Children who were educated in English both before and after the law change, children educated in French both before and after the law change, and a group of children who would have been educated in English before the law change but were forced to be educated in French after the law change. By comparing the outcomes of these three groups Duhaime-Ross can isolate the effects of the law change on the outcomes of the children affected by the law.

The outcomes that Duhaime-Ross used were taken from the 2006 Canadian census. What this means is that the outcome data is more than 20 years after these children finished school, so that we are looking at the long term effects of schooling language.

Duhaime-Ross found that the law was unambiguously good for the children of immigrants. Children whose education was in French (instead of English purely because of the law change) earned higher incomes, were more likely to be employed, and were more likely to have gone to university. The law was not, however, unambiguously good for the province of Quebec. The law caused some immigrant families to leave Quebec and move to other parts of Canada so that their children could be educated in English. The families that left tended to be better educated and earn higher incomes than those that stayed.

None of these effects were intended effects of Bill 101. The purpose of the bill was simply to protect the French language in Quebec (the bill also had a bunch of other provisions to protect French as well). Nevertheless, the bill improved the long-term outcomes of immigrant children but also cost Quebec some extra `high value’ immigrants. The lesson here is that new laws can have unintended consequences that can last decades. Understanding these effects is interesting but also extremely important, particularly for policy makers who might be considering similar laws in other jurisdictions.

 

Job market seminar: Galizia

Leave a reply

Last week,  Dana Galizia presented his job market paper to the department. Once again, this paper is quite different from the two previous job market presentations I have discussed. Galizia is a macroeconomist, working on the big picture issue of business cycles.

Macroeconomics is possibly the most visible branch of economics, but it is also probably the most misunderstood by non-economists. Macroeconomics really takes the motto “all models are wrong, but some models are useful” to heart. A macroeconomist will build a highly stylised model that attempts to explain a particular issue in the economy (trying to explain everything at once is just too difficult!).

It is often thought that productivity shocks play a large part in driving business cycles. Productivity shocks can be thought of as things like weather shocks (drought, floods etc.) or international developments (revolutions and wars, for example) that affect global prices and resource supplies. The problem is that when you try to build a model where business cycles are driven by productivity shocks it normally requires productivity shocks that are unrealistically large (by a factor of 5 or more) to get the models to work.

Why is this? Well, the basic macroeconomic models are built around a steady state, which means that in the absence of any shocks the model produces  a constant rate of growth. So to generate business cycles we need a lot of shocks to push the model away from the steady state.

Galizia takes a different approach. He builds on underlying model that already has some cycles built in (these are called limit cycles). In the absence of any shocks, Galizia’s model will exhibit perfectly regular business cycles, which are generated by changes in the amount of purchases made in response to unemployment risks.

Obviously the real economy does not exhibit perfectly regular business cycles, so Galizia adds in some productivity shocks. Because Galizia’s model already has cycles built in, the productivity shocks that are required to generate realistic business cycles are of a much more realistic size (in fact, the productivity shocks in Galizia’s model are actually slightly smaller than those observed in real world data).

Galizia’s model would not be very useful for predicting future economic trends (that is not what it was built for), but it does demonstrate that we can build a model of business cycles with realistic productivity shocks. This is a foundational work, which introduces some key techniques which can now be applied to other, more realistic models.

 

Job market seminars: Yu and Cosman

Leave a reply

So far I have attended two practice job market seminars; those by Zhengfei Yu and Jacob Cosman. Although their work is in rather different fields I will address them both in this single post because there are some interesting juxtapositions between their work.

Yu’s job market paper is a classic job market paper: it is ambitious and attacks a “big” problem in theoretical econometrics (the study of statistical techniques as applied to economic problems). Econometricians often use large data sets that have been collected by national statistical agencies (containing, say, quantities and prices and, if you are lucky, production costs for an industry). One of the assumptions that is usually made to facilitate data analysis is that the industry is in an equilibrium, and this equilibrium is unique.

Economic theory tells us, however, that often there are multiple equilibria that can exist. If there are multiple equilibria, but we assume that the equilibria is unique, then our conclusions will be wrong. Yu asks the question: can we tell, just from looking at the data, whether the assumption of a unique equilibrium is correct?

Yu finds that the answer to this question is yes, and he proposes a method to test whether there is a unique equilibrium or not. This is a great example of big picture econometric research.

Cosman, on the other hand, takes an approach at the other extreme. Instead of attacking a big problem in the abstract, Cosman attacks a very specific problem. In some sense, the answer to Cosman’s question is secondary: just as important are the advances in the techniques that Cosman uses.

Cosman works in urban economics and industrial organisation (the study of firm and industry level behaviour and outcomes). One of the hallmarks of empirical industrial organisation studies is getting a whole lot of mileage out of a paucity of data, usually by making a series of strong assumptions. Empirical industrial organisation advances by scholars pushing the limits and trying to find the minimal level of assumptions needed to get the most answers from a given data set.

Cosman studies night life in Chicago. Using a data set which is built chiefly from liquor licence applications and cancellations (which can be seen as entry of new bars, or old bars closing down) Cosman is able to estimate the costs of starting a bar in Chicago and how much consumers value new bars. If you are interested in social planning, or in Chicago in particular, then you are probably very interested in Cosman’s estimates.

For others though, the methodological innovations that Cosman introduces are more interesting. The main innovation is the introduction of a dynamic estimation technique that both simplifies the computational burden of solving the model and allows (realistically) for sequential decision making by potential bar owners.

Both Yu and Cosman have very interesting papers, and they showcase the broad range of techniques that can be applied to modern economic research. It doesn’t matter if you are answering broad abstract questions, or using novel techniques to  estimate demand functions for a single industry in a single city; it’s all economics, and it’s all pushing forwards the frontier of knowledge. But best of all, it’s all fun!

Job market season

Leave a reply

Economics is, as far as I am aware, the only academic field which has developed the infrastructure to facilitate a large scale coordinated market for hiring new PhDs into academic positions. Every year, in November, over a thousand PhD graduates apply to over a thousand jobs. In January all the North American universities, and most top international universities, conduct interviews in parallel with the annual American Economics Association meetings. Second round interviews run from January until March (or even April) and  usually involve flying the candidates out to visit the hiring university.

This year’s candidates from UBC are listed here. In November, the candidates give presentations to the department outlining their job market papers (generally the best piece of research that the candidate has produced during their PhD).

What is so interesting about all of this? Well, if you want to get a good handle on where economics is heading, looking at the research that the next wave of young researchers are producing is a pretty good place to start.

Throughout the rest of the month I will post up brief summaries of the job market presentations that I attend. I most likely won’t be able to attend all of them, but I will attend as many as I can.

So, check back throughout the month to get a taste of modern economic research across a broad range of fields and from a great group of young researchers.

The 1955 automobile price war?

Leave a reply

 

Today’s offering is another one of my favourite papers. Tim Bresnahan has a fantastic paper that attempts to explain the price and quantity of US automobile industry from 1954 through 1956. When you see the footnote that mentions that the paper was the second chapter of his PhD thesis (and, presumably, his job market paper) it is all the more remarkable.

I’ll begin with a caveat. The conclusions of this paper may not be correct. In fact, there is a much more recent paper by David Rapson that makes exactly that claim. But this doesn’t detract from the contribution of Bresnahan’s work. Bresnahan developed a novel and elegant methodology to answer his stated question but, given data limitations and the state of econometrics at the time he was writing, Bresnahan needed to make several assumptions in order to complete his analysis. The more modern, non-parametric, approach taken in Rapson’s paper suggests that some of Bresnahan’s assumptions may be violated, but Bresnahan’s contribution should be considered within the context of the state of economics in the late 1970’s.

There was an increase in quantity and decrease in price of cars sold in the US in 1955, relative to both 1954 and 1956. Why did this occur? At the time that Bresnahan wrote his paper, this was a puzzle to which that no one had found a satisfactory answer. Bresnahan’s conclusion was that the automobile industry was in a state of tacit collusion in both 1954 and 1956, and that there was a price war in 1955. Bresnahan was able to answer this question using only data on prices and quantities sold (broken down by make and model), and data on the characteristics of the different models.

The first step that Bresnahan took was to build a model of demand for automobiles. To do this, he needed to aggregate the characteristics of each model into a single quality dimension, over which he could estimate preferences. The quality weights assigned to each characteristic are determined via simultaneous estimation with the supply and equilibrium conditions. In other words, Bresnahan allows the data to tell him how consumers are willing to trade off horsepower and vehicle size, for example.

Once each car has been assigned to a point on the quality scale we can start to think about how much different cars are competing with each other. For example, a small hatchback doesn’t really compete with luxury sedans as much as it competes with other small hatchbacks. Now, consider two cars that are very similar to each other. If they are in competition with each other then their prices will be driven down close to marginal cost. If they are colluding with each other, then the prices for both cars will be significantly higher. For cars that are very different from all the other types of cars the effect of competition is not as important.

So, which cars are priced cooperatively, and which cars are priced competitively? Obviously, cars that are made by the same company will be priced cooperatively against each other. For cars made by different companies, Bresnahan estimates what the prices would be if they were cooperating and what the prices would be if they were competing. Then we can check which model (cooperative or competitive pricing) fits the data best. In 1954 and 1956 the cooperative model fits the data best, but in 1955 the competitive model fits the data best.

Bresnahan did an amazing amount with very limited data. But to do so, he needed to make some strong functional form assumptions. For example, he assumed that marginal costs were log linear in quality, and that the relationship between product characteristics and quality has a square root form. These are rather arbitrary functional form assumptions, but were necessary to turn the raw data into an estimable set of equations. As the sophistication of non-parametric econometric techniques has increased over time, it is now possible to get a lot more mileage from a given set of data without needing to make such arbitrary assumptions.

Bresnahan’s paper has made two important contributions to my understanding of empirical industrial organization (as I am not really up to speed with the modern empirical IO literature, I will refrain from commenting on the impact the paper had on the broader literature). Firstly, it demonstrates just how much we can achieve with limited data and a rigorous, theory driven, empirical approach. Secondly, and perhaps even more importantly, it shows the limitations of econometric specifications that impose arbitrary functional forms.

It seems to me like empirical industrial organization is one area of economics that has a lot of potential to take advantage of advances in non-parametric statistical techniques. Reading more empirical IO papers is one of those things that is always on my to-do list, but I never get around to actually doing… hopefully this will change.

 

 

World Cup betting pool: the outcomes

Leave a reply

I have had a request to follow up on my previous post on the World Cup betting pool that I ran in my department. Specifically, I was asked to address whether the betting satisfied the four desiderata that I outlined in the previous post.

For a brief refresher, I used a model where we treat each of the teams (or a group of teams) in the World Cup as a room in a share house, and each of the participants in the betting pool as a tenant in the house. Then the goal is to match rooms to tenants, and allocate shares of the rent across tenants, so that:

  • the outcome should be efficient
  • no one should envy anyone else’s room/rent combination
  • the sum of the rents should be equal to the total rent payable for the house
  • the mechanism should be incentive compatible (i.e. no one should be able to manipulate the outcome by lying about their preferences)

So, did my betting pool satisfy these four criteria? The short answer is that it is impossible to guarantee all four at once.

The longer answer is that, in summary, if we can assume that no one lied about their preferences, then the other three conditions are automatically satisfied. If we think this assumption might be violated, then the first three conditions will still be satisfied if people have misrepresented their preferences in an optimal fashion. If we think that people might have misrepresented their preferences sub-optimally, then condition three will still be satisfied, but there is nothing that we can say about the first two conditions.

For more details, read on!

Continue reading →

Quantity precommitment and Bertrand competition yield Cournot outcomes

Leave a reply

This classic paper by David Kreps and Jose Scheinkman is quite possibly my favourite economics paper of all time. It’s easy to explain and understand, but still makes a very deep point that connects two of the most famous economic models in a simple way.

Way back in 1838, Antoine Augustin Cournot wrote what is, I believe, the first mathematical model of competition between firms. The model makes reasonable predictions that are, in a broad sense, supported by empirical observations. One of these predictions is that as more firms enter a market it should become more competitive and prices should decrease.

However, the Cournot model has one deeply unsatisfactory dimension: firms set the quantity that they wish to sell, and then the market determines the price at which that quantity can be sold. This is most certainly not how firms actually make decisions; when a customer goes shopping, the store posts a price and the customer decides how much to buy.

In 1883, Joseph Bertrand came along and wrote a model where firms set prices, and then the market determines the quantity that will be sold at that price. This is a much more satisfactory foundation for a model of firm behaviour. Unfortunately, the Bertrand model generates very poor predictions. One of the implications of the Bertrand model is that two firms are enough to generate extremely intense competition and low prices. Another implication is that adding more firms to the market doesn’t change the outcome. Neither of these implications are compatible with empirical observations.

So we have one model with good assumptions but inaccurate implications, and one model with poor assumptions but reasonable implications. Is there a way that we can resolve this tension?

It took 100 years, but in 1983 Kreps and Scheinkman found the resolution: they key is to use a two stage model. In the first stage, firms install production capacity. Then, in the second stage, the firms compete over prices ala Bertrand. But here’s the kicker: the outcomes that are produced by this model are exactly the same as the Cournot model.

So we now have a model with both good assumptions and reasonable implications. Of course, this model is still highly stylised and leaves a lot of potentially important features unmodelled, but it does provide an extremely compact way of reconciling two very important economic models. Pretty neat.

Additional notes for economists

Continue reading →

Loss aversion and reference points. Or why mathematical formalism is important for behavioural economics.

Leave a reply

UBC and the University of Hong Kong hold a joint economic theory workshop every summer. The 2014 edition was held over the last week; unfortunately I was only able to attend a couple of the presentations, but one in particular was very instructive.

Collin Raymond, a postdoc at Oxford, presented his work on Stochastic Reference Points, Loss Aversion and Choice Under Risk (joint with Yusufcan Masatlioglu) [1].

Raymond’s presentation focused on the relationships between a bunch of different economic models of decision making. The “standard” model of economic decision making, the Subjective Expected Utility (SEU) model, works pretty well in most situations. But, there are some situations where some people regularly violate SEU. In response to this, economists have developed a number of models that generalize (expand) the SEU model. In particular, Raymond focused on one particular model, the Koszegi-Rabin model, and how it relates to other models.

Now, if you are not familiar with decision theory, this might sound a little strange to you. Why would we need a presentation on how these different models relate to each other? Shouldn’t we know this already? Well, we often don’t because these things are pretty complicated.

In decision theory, a model has two parts. An axiomatization and a representation. Generally, we think of a model as starting with a bunch of axioms – statements about how people might behave that are (at least) plausibly true. An example of an axiom is: If you prefer A to B, and you prefer B to C, then you must prefer A to C. Starting from a bunch of axioms, we can then derive a representation – a formula that “represents” the axioms in a simple and useful way. Often axioms are rather complicated and only subtly different between models, and it can sometimes take years to work out how the axioms from one model relate to the axioms of another model.

Often we can have strong intuitions about how we think models should be related. One of the most striking things about Raymond’s presentation was just how wrong our intuition can be. In other words, Raymond’s presentation demonstrated just how important it is to go through the formal mathematics when dealing with decision theory models.

The Koszegi-Rabin model that Raymond considered is a model of reference dependence. In these types of models, the decision maker (DM) has a reference point which they use to evaluate all of their options against. The DM likes getting things that are better than the reference point, and dislikes getting things that are worse than the reference point. The key point is that the DM dislikes things that are worse than the reference point more than they like things that are better than the reference point. We say that the DM is loss averse. 

There are many different models of reference dependence. The key difference between them is how the reference point is determined. In the first(?) model of reference dependence, which won Dan Kahneman the Nobel Memorial prize[2] the reference point was considered to be exogenous (i.e. determined outside of the model). Modern models of reference dependence include a way of determining what the reference point is inside the model.

Another class of models don’t have a reference point, but instead the DM is pessimistic in the sense that they overweight bad outcomes (i.e. they behave as if bad outcomes are more likely to occur than the actually are).

Raymond compared the Koszegi-Rabin model with two other reference dependent models and two pessimistic models. Amazingly, the behaviour that is allowed by the Koszegi-Rabin model is actually a subset of the behaviour allowed by the two pessimistic models, but only overlaps with the other two reference dependent models at a single point.

What this tells is that, depending on the exact form of the models, sometimes a reference dependent model acts just like a pessimistic model (and not like other reference dependent models). This is a very counter-intuitive result, and it shows the usefulness of being careful and thorough when working with decision theory models. Representation theorems are very powerful mathematical tools, but they are full of subtleties that can take a lot of study to fully comprehend. In this case, as in many others, formal mathematics has taught us something that we would never have discovered without it.

 

[1] I’ve only skimmed the paper, so most of what I write here will be based off what I picked up during the presentation.

[2] His co-author Amos Tversky passed away before the prize was awarded.

Cheap talk can be valuable

Leave a reply

I was a huge fan of the blog Cheap Talk. I say was only because the posting rate there has trickled down to almost nothing (sure, you could say the same thing about this blog, but I also have practically no readers). But, occasionally they still put up a new post that, more often than not, includes terrific passages such as this one:

So, is Hachette, a French company, confused because in France they put price on the x-axis and quantity on the y-axis so marginal revenue is upside down? Surely Jean Tirole can sort that out for you.

 

An economist’s betting pool

Leave a reply

I am running a betting pool amongst some members of the department for the upcoming World Cup. Now, I suppose that the normal thing to do would be to assign teams randomly amongst the participants sweepstakes style. But for an economist, this seems like a rather inefficient thing to do. What if one person really wants Brazil but draws Germany, and someone else really wants Germany but draws Brazil?

Fortunately, there is an entire literature on what is known as the “rental harmony problem.” In the rental harmony problem, there are a group of house mates who are renting a new house and they wish to work out who should have each room and what share of the rent each should pay. We want the outcome to satisfy a few properties:

  • the outcome should be efficient
  • no one should envy anyone else’s room/rent combination
  • the sum of the rents should be equal to the total rent payable for the house
  • the mechanism should be incentive compatible (i.e. no one should be manipulate the outcome by lying about their preferences)

Unfortunately, it is impossible to satisfy all four of these conditions at once (this can be demonstrated as an application of the Vickrey-Clarke-Groves mechanism; if we enforce strict budget balance no mechanism can be found). Therefore, we must relax one of the conditions, and the usual thing to do is to relax the last condition.

At this point, mathematicians and economists differ in their approach. Mathematicians tend to simply assume that people will tell the truth; this is the approach taken in this extremely elegant paper by Su [1]. Economists, on the other hand, tend to assume that people will lie if it is profitable, and attempt to minimize the damage caused by the lying; this is the approach taken in this paper by Abdulkadiroglu, Sonmez and Unver.

Therefore, for my World Cup betting pool I will be using the mechanism described in the Abdulkadiroglu et al paper. Each team (or group of teams) will be treated as a room in the house, and the total prize money available will be treated as the total rent for the house. Each participant will need to submit a vector of bids (one bid for each team or group of teams), and the algorithm will allocate teams and entry costs to the participants. In a follow up post I will outline the bids that were submitted and the outcome of the algorithm.

 

Additional notes for economists:

Continue reading →