Category Archives: Economic education and methods

On the state of economics

Posted on by

I have a long essay on the scientific status of economics in National Affairs. A few excerpts from the conclusion:

In the end, can we really have effective theory in economics? If by effective theory we mean theory that is verifiable and reliable for prediction and control, the answer is likely no. Instead, economics deals in speculative interpretations and must continue to do so.

Young economists who employ pluralistic methods to study problems are admired rather than marginalized, as they were in 1980. But economists who question the wisdom of interventionist economic policies seem headed toward the fringes of the profession.

This is my essay in which I say that academic economics is on the road to sociology.

Facts, Feelings, and Filters

Posted on by

A commenter writes,

Arnold’s argument that economics is about using particular frameworks as lenses for interpretation is also quite postmodern.

Well, sort of.

Consider three statements.

a) Amazon announced its intention to purchase Whole Foods.

b) Amazon should not be allowed to purchase Whole Foods.

c) Amazon’s purchase of Whole Foods damages the prospects of other grocers.

(a) is an example of a fact. (b) is an example of a feeling.

(c) is an example of an observation based on a filter, in that it depends on one’s framework of interpretation. You might think one way if you see Amazon’s move as intensifying competition in the grocery industry. You might think differently if you see it reducing competition and/or as a signal that there is value in national grocery franchises (what if Google or Facebook decide they also want to own grocery stores?). And, yes, the drop in stock prices for other large grocery chains says that investors favor one interpretation more than another. But my point is that the interpretation is contestable.

Some more remarks.

1. In 20th-century philosophy, the Logical Positivists seemed to dismiss the concept of filters. They would regard (c) as an attempted fact-claim. Anything other than a fact-claim or a feeling is a dogma.

2. The Post-modernists take the opposite view. Every statement comes through a filter. This would make every statement contestable, including (a).

3. I wish to take an intermediate position. I believe that there are scientific observations and laws that are not contestable, but I also believe that filters are very important. Synonyms for filters include frameworks of interpretation, models, theories, and paradigms.

4. In Specialization and Trade, I argue against the dominant filter in macroeconomics, which I call the GDP factory.

5. In The Three Languages of Politics, I argue that progressives, conservatives, and libertarians each use distinctive linguistic filters. If you make an argument using terms that correspond to (for example) the progressive’s linguistic filter, a progressive will approve that argument, while it will fail to resonate with a conservative or libertarian.

6. In both books, I am suggesting that people think they see truth, but there are different plausible filters that would change their outlook.

7. However, I do not go so far as to say that there is no truth, and that any belief system is as good as any other. Instead, I am saying that sometimes there is more than one plausible filter. If you are sending a man to the moon or building a computer, you had better use the consensus scientific filters. In other realms, where causal density is high, no filter is robust.

8. If you want to be wise, you need to acknowledge the anomalies that cast doubt on your filters. Otherwise, you end up treating your filter as a sacred tribal doctrine.

9. There is a prominent version of post-modernism that I would term Left Post-modernism. Strictly speaking, post-modernism should lead one to be aware of many possible filters and skeptical of one’s own filters. In contrast, Left Post-modernism puts everything through the filter of race and gender and is entirely lacking in self-doubt. For example, in Sunday’s WaPo, Tung Yin writes,

Mass killings look the most like terrorism when their perpetrators seem the most alien from the Judeo-Christian, white majority.

This is Left Post-modernism treating its filter as a sacred tribal doctrine, ignoring some pretty obvious contrary evidence. Just off the top of my head, the Irish Republican Army and the Baader-Meinhof gang were labeled terrorists.

The WaPo itself has an analysis on line (not in print, that I could see) of Friday’s terrorist stabbing in Jerusalem, which is focused on “who they were working with and for.” That is one distinctive feature of terrorism, which is that the perpetrators claim to act on behalf of an organization that engages in terrorism. But far be it from Yin to admit that the term “terrorism” is anything other than a racist epithet.

Speaking of Friday’s attack, in which an Israeli policewoman was stabbed to death before the attackers were killed, The BBC notoriously headlined the incident “Three Palestinians killed after deadly stabbing in Jerusalem.” This is how they prefer to filter such news (although in this rare instance, following Israeli outrage the BBC later changed the headline). The WaPo filtered the news even more effectively, because I did not see any coverage of the incident at all in its print editions. It might otherwise disturb the narrative that the WaPo put forth prominently in recent Sunday editions, in which the Palestinians suffer from checkpoints for no reason under “occupation.”

The WaPo news and Outlook sections are now all Left Post-modernism, all the time. The editorial page is sometimes more broad-minded, but I have given up on the heavy-handed filtering disguised as reporting and analysis. For news, I look elsewhere.

The decline of GDP

Posted on by

Shane Greenstein writes,

Television has gotten much better over the last few decades, but—for many reasons—total advertising has not grown.

Without more advertising revenue, the contribution of better TV to GDP is zero. A few remarks:

For me personally, the value of television is close to zero. I never turn on the TV. Still, if I did have to watch, I would find prefer new shows to the type of shows that were available when I was a kid.

But Greenstein’s larger point is that free services, like Google Maps, do not get valued properly in GDP.

I would like to make the point larger still. The economy is much less legible today than it was in 1950. The most legible components of the economy are agriculture and manufacturing. In 1950, the majority of people worked in those sectors. Today, if you add up farm labor and manufacturing production workers (not including white collar workers in manufacturing), you get maybe 7 percent of the labor force. Pretty much everyone else works in sectors like finance, government, health care, and education, where we do not know how to measure or value output.

The Department of Commerce hums along, producing a number for GDP. And many economists read a lot into the behavior of this number. In the process, they treat an increasingly illegible economy as if it were still legible.

Yuval Levin on the CBO

Posted on by

He writes,

Rather than produce stark hard-number projections behind a heavy veil, the CBO and JCT could act more as developers and stewards of an open public model, available online to anyone with sufficient technical prowess to use it. The two agencies, with outside help, would create and maintain the model, and would decide on a set of official economic assumptions and policy expectations to be used in modeling for formal budget-process purposes. Anyone could “score” any policy proposal using those official assumptions and could also use the model with other assumptions to project a proposal’s consequences under alternative circumstances.

I do not think that any fix of the Congressional Budget Office or Joint Committee on Taxation gets at the problem.

Alan Blinder once pointed out what he called Murphy’s Law of economic policy. That is when economists agree on something and are confident about it, nobody listens. But when economists are engaged in speculation and biased argument, they get attention.

When it comes to “scoring” health bills for their effect on the proportion of the population with health insurance, the CBO is engaged in speculation, and Levin points out a form of bias.

It is by this point basically a quirk of the CBO model that it judges the mandate to be exceedingly effective — far more effective than the evidence of the past few years would suggest.

If your model tells you that a lot of people purchased health insurance because of the mandate, then it will predict that a lot of people will decline health insurance without it. And, of course, the press will report this as people “losing” their health insurance.

On the other hand, when the CBO puts out its reasonable forecasts that entitlement spending is not sustainable, pundits tell us not to take those seriously. But as far as I can tell, those forecasts are still on track.

My Review of Kevin Laland

Posted on by

The book is Darwin’s Unfinished Symphony, which is still my favorite non-fiction book of 2017. My review says,

Laland weaved together mathematical models, simulation exercises, experiments, and observations in a way that was much more persuasive than most social science. I recommend that economists read the book in order to stimulate thinking on how to improve our research methods.

The Samuelson Story

Posted on by

I am still reading the Backhouse intellectual biography. What a monumental effort! By comparison, it makes Sebastian Mallaby’s biography of Alan Greenspan look like a summer vacation project. But I recommend Mallaby to anyone. Backhouse is for people with a pretty narrow interest in the history of economic thought.

A few more comments.

1. Backhouse sees Samuelson as making a major contribution to clarifying Keynes. I disagree. At most, Samuelson provided the freshman interpretation. But for intermediate courses and above, Hicks was the go-to clarifier, and he preceded Samuelson. Backhouse agrees that the intermediate courses left Samuelson behind.

It seems that a lot of the initial confusion about the simple Keynesian multiplier concerned the process by which it was supposed to play out. You can write down Y = C + I + G and C = a + bY and then solve for Y and C in terms of the other letters, but nobody in the 1930s or 1940s thought in terms of such an instantaneous equilibrium. Instead, they thought in terms of a solution being ground out over time, iterating back and forth between the two equations. There were any number of plausible lag structures, and also a variety of plausible definitions of investment, consumption, and savings.

As Backhouse points out, in hindsight it appears that Keynesian economists were floundering unnecessarily in their attempts to deal with what we now regard as the simplest class of models. Perhaps without Samuelson the floundering would have been worse and more long-lasting. I am not convinced.

But the main reason I do not give Samuelson credit for clarifying Keynes is that Keynesian economics remains unclear to this day. And economists who call themselves Keynesians would be the first to admit that.

2. The conventional narrative is that Samuelson made mid-century economics. His prodigious talent shaped the profession. Backhouse sticks with this narrative.

I would pose an alternative narrative, which is that mid-century economics made Samuelson. Sometimes a movement needs to create a larger-than-life figure. In religion, the larger-than-life figure might be Christ or Mohammed or the Buddha. In a nation at war, it might be Napoleon or Churchill.

In the (seemingly) secular world of Olympic sports, the broadcast network will try to make a few athletes larger than life. And speaking of Keynes, he became a larger-than-life figure in the movement to promote activist macroeconomic policy.

With his unbounded ego, Samuelson moved easily into his heroic role. He was a legend in his own mind, given to overstating both the rigor of his mathematics and the significance of his analysis. Had he appeared at a different point in history, this might have earned him scorn and derision. But when Samuelson arrived as a graduate student at Harvard, there was a strong latent desire, particular among younger economists, to show that they were scientists and to show that they, unlike their elders, knew how to prevent and cure Depressions. So instead of pushing back against his arrogance, his peers heaped praise on his work. In the process, they were raising their own status relative to older economists while creating a new and more exalted public image of the economics profession.

By the way, I think that the same thing happened, to a lesser degree, with some of the stars of 1970s macro, such as Robert E. Lucas, Jr., and Christopher Sims. Some of their peers saw an opportunity to raise their own status by characterizing Lucas and Sims as intellectual giants who towered over everyone else.

My own view is that in economics there are no towering giants, in the sense of economists whose aptitude for the subject far exceeds that of others in their cohort. Some economists are a bit more creative than others, some work much harder than others (Samuelson was certainly exceptional in the efficiency and extent of his work effort), and some play the academic game much better than others. But the distribution of rewards ends up being much more skewed than the distribution of talent.

Why I see academic economics moving left

Posted on by

Don Boudreaux wonders why I make that prediction, and he offers one possible explanation.

economists’ increasing embrace of empiricism that isn’t solidly rooted in basic microeconomic theory of the sort that can be, and should be, taught to undergraduates. That which can be measured and quantified is that which can be seen. Indeed, to quantify something is, in a real sense, to see something – or at least to see some of that something’s physical manifestations. And yet a chief lesson taught by a sound course in basic economics is that many social phenomena are unseen.

What I need is an explanation for what I see as the current and future drift of economics. The difference between the seen and the unseen has been discussed since Bastiat. It could be true that the greater availability of data and the greater emphasis on empirical work creates a bias toward seeing the benefits of intervention and understating the costs. But that is not one of the explanations I had in mind.

I am thinking primarily in terms of the culture of the academy. Disciplines can easily become dominated by narrow ways of thinking. As of the late 1970s, when I was in graduate school, mathematical optimization and multiple regression crowded out everything else. Soon, rational expectations theory and dynamic stochastic optimization had crowded out any sensible thinking about macroeconomics. As those of you who have read my macro memoir can perhaps appreciate, my experience leads me to believe that incredibly dumb ideas can become dominant in an environment in which a handful of professors effectively control the academic job market in a field.

I actually think that in terms of methods, academic economics is getting more broad-minded and less stupid. The generation that believed that progress comes from solving more difficult math problems and using fancier econometrics is finally giving way to younger economists with less rigid methodological doctrines. But young economists are less interested in the fundamental issues of social organization. They prefer looking at narrower, technical questions. They are happy to write papers that suggest potential policy improvements, without worrying about how the political process conducts itself in practice.

If contemporary America is embroiled in a war between the Concrete class that works with stuff and the Abstract class that works with words, numbers, and computer programs, then economists, like other Academics, will side with the Abstract class. The Abstract class believes that in the natural order of things it ought to be managing and governing. I suspect that young economists today are not going to want to criticize this natural order of things, for fear of being shunned as class traitors.

An Introduction to Economics

Posted on by

Economics is the study of human interaction in the production and exchange of goods and services in the context of large-scale societies.

In human interaction, our thoughts affect our behavior. Among all animal species, humans are unique in the extent to which we record information and communicate with one another. Our beliefs, and especially our shared beliefs, vary across cultures and evolve over time. This makes the causal structure in human interaction more complex than that in natural science. We are limited in our ability to predict and control human interaction, including economic behavior.

Economics is focused on human interaction in the production and exchange of goods and services. Political science explores human interaction in the exercise of government power. Sociology explores human interaction in the maintenance of status hierarchies and social norms. Psychology and anthropology explore more fundamental aspects of human interaction, often with a focus on interaction within families or isolated small groups.

Economics is most useful in studying large-scale societies, even though there are some economic concepts that can be applied to just one individual (the so-called “Robinson Crusoe economy”). A large-scale society differs from a family or a small isolated tribe in two important ways. First, it is possible in a large-scale society to have a much higher degree of specialization. Second, in a large-scale society, exchange takes place among people who do not know one another and who do not engage in repeated interactions with one another. To coordinate complex specialization and anonymous exchange, large-scale societies require mechanisms that are not needed in small groups. Economists explore these coordination mechanisms.

Many economists employ a technique in which they set up a simplified version of an economic problem, make some basic assumptions about the goals of individuals, calculate the optimal strategy for each type of individual, and arrive at a mathematical prediction for the outcome. This optimization modeling has come to dominate the research and pedagogy of economics. However, for the purpose of describing human interaction, optimization modeling is rigid, limited, and often misleading. It is best to treat optimization modeling as just one tool of economic analysis, rather than as the essence of economics.