Every student of economics is required to study these fields, at a level appropriate to the degree they are pursuing, before they can graduate. At the PhD level, there is typically a core exam to assess the student’s proficiency in the core subjects, which the student must pass to receive their degree. In subsequent posts to this blog, I will describe the fundamental theories of the core fields in some detail, complete with mathematical notation and figures.
1) Microeconomics, also called economic theory.
The basic principles of economics: What are consumers, and how do they behave? Same question for firms and markets. Conceptual and mathematical tools for the analysis of firms, consumers, markets, trade, and strategy. Game theory and its applications in economics. Incorporating the findings of economic specialties into the central framework.
The intersection of economics with mathematical statistics. Econometricians study the methods of data analysis, techniques for regressions and estimation, as well as causal inference. They especially emphasize the use of data easily available to economists (surveys, censuses, market research, occasionally experiments in a lab or in collaboration with an external organization) to answer questions of interest to economists, such as the generating process for income, the returns to education, or a method for estimating the demand curve for any good. Econometricians develop methods to deal with some of the difficulties faced by economic research—the processes we’re interested in are complicated, heterogeneous, and depend on many unobservable factors; it’s often not feasible to conduct the experiments we would like, but some information can be obtained by investigating natural data, etc. It’s trivial to find statistical relationships in the data (correlations, conditional means, best-fit lines etc), and in a large random sample, these indicate similar relationships in the population. To find causal effects, ideally you would run an experiment with a treatment group and a control group; however, it is not always feasible or ethical to experimentally find things like whether depriving a person of education reduces their lifetime earnings; and experiments on the macroeconomy (would the USA have been better or worse off without the stimulus?) are often impossible. Even so, with clever use of economic theory, statistics and logic, it’s possible to make reasonable claims about the causal effect of theoretically exogenous interventions even on natural data, even if you cannot perform experiments.
(Using natural data sounds unscientific, and it certainly carries risks—a naïve researcher can make very basic, very serious mistakes, and a sophisticated researcher can make very subtle and sophisticated mistakes. However, there are respectable scientific fields such as astronomy that have succeeded despite the impossibility of experimentally manipulating the stars. Astronomers can make many inferences about the stars and planets, inferences good enough to allow us to gently land probes on many bodies in the solar system, using only natural data and theories developed by physicists (about forces governing the motion of objects). Empirical economists applying the techniques of econometrics, similarly, must rely on theories developed by economists, about incentives governing the choices made by individual participants in the market. It’s still necessary to use caution with the findings of any economist—human behavior is complicated and cannot be fully explained by any theory of rational optimization, or in fact any theory yet developed; and worse, since economists study incentives, and since incentives may change once they are widely known, economic findings sometimes invalidate themselves—it’s possible for something to be true until it becomes well known, at which point it is no longer true. (For an example, consider an amazing restaurant which has not yet been discovered. You love it for the cozy environment and the lack of crowds, but if you tell everyone you know, and they tell everyone they know, the restaurant will no longer be an undiscovered gem. Similarly, if you find a little-known road to go downtown, which is never congested because no one knows about it… ))
The study of the whole economy, and aggregates that make the most sense on a large scale: GDP, economic growth, aggregate demand, aggregate supply, unemployment rates, price levels, inflation, money supply, money demand. Macroeconomics also studies the effect of government policy on these things.
In the early days of the field, macroeconomists studied historical trends in the relationships of key macroeconomic variables. For instance, William Phillips, a New Zealand-born LSE economist who invented the MONIAC hydraulic computer (a friend of mine built one for the latest UChicago Scav Hunt a few weeks ago), also discovered in the 1950s that there was historically a negative relationship between inflation and unemployment. A strong economy, with rapid growth, near-full employment and rising wages also tended to have rising prices. This relationship came to be known as the Phillips curve.
The problem with historical analysis soon became apparent. When governments tried to print money and induce inflation in order to combat unemployment, it quickly became clear that the inflation-unemployment coupling was not simply a mechanical linkage—a strong economy can cause inflation and full employment, but inflation itself does not cause full employment.
Better methods were soon developed, following principles such as the Lucas Critique, named after University of Chicago economist Robert Lucas. In 1976, Lucas wrote that it was naïve to try to predict the effects of a policy based on aggregated historical data gathered before the policy was implemented. (A similar principle, known as Goodhart’s law, says that as soon as a measure becomes a target of policy, it ceases to be a good measure. Campbell’s law, almost identical to Goodhart’s, makes the same prediction and further explains that it is due to “corruption pressures,” ie to self-serving changes in behavior in response to the new incentives.) The problem is the same one we saw when discussing causal effects in econometrics—a government policy changes incentives, and economic behavior consists of people responding to incentives; therefore a government policy that aims to exploit a historical relationship will change the very incentive structure that created the relationship. (Wikipedia’s entry on the Lucas critique gives the example of Fort Knox—the fort contains much of the US Government’s vast gold reserves. It is, of course, heavily guarded. Historically, the fort has never been robbed, but this does not mean the guards can all be retired—the fort is safe from robbers precisely because it is so heavily guarded.)
Instead, Lucas recommended that macroeconomic analysis be based on “microfoundations” – the fundamental, structural, microeconomic parameters that govern consumer behavior, such as consumer preferences, resource constraints, firms’ production technology, etc. Having built a foundation for analyzing economic behavior, we can use a dynamic model of the whole economy to study the emergence of macroeconomic parameters like growth and inflation. On top of this basis, we can predict the effects of government policies by seeing how they change the incentives faced by consumers and firms.
Modern macroeconomics heeds the Lucas critique, and thus can only be studied after we have discussed microeconomics. It uses a dynamic general equilibrium model, based on foundations from microeconomics, and studies how such an economy would grow, and how it would respond to the changes in incentives and resource constraints induced by government policy.