The historian's craft and economics

My paper (with Mary Eschelbach Hansen) “The historian’s craft and economics” is now available on First View at the Journal of Institutional Economics:

Abstract

History refers both to the past and to the systematic study of the past. Attempts to make a case for history in economics generally emphasize the first definition. There are benefits from increased attention to the past. This paper argues that significant benefits can be gained from increased attention to the systematic study of the past, the historian's craft. The essence of the historian's craft is the critical evaluation of sources. Failure to critically evaluate sources has the potential to lead to erroneous conclusions, whether one is using historical documents or more recently created data.

Economics really needs better critics

Per Byland recently complained that economists had killed economics

“What we have seen over the course of the last eighty years is a systematic dismantling of the contribution of economics to our understanding of the social world. Whatever the cause, modern economics is now not much more than formal modeling using mathematics dressed up in economics-sounding lingo."

I’m not sure that Bylund and Michael Lind would agree on much, but Lind also has seen the destruction of economics

 “Before World War II, economics — the field which had replaced the older “political economy” — was contested between neoclassical economics, which sought to model the economy with the methods of physics, and the much more sensible and empirically-oriented school of institutional economics. Another name for institutional economics was the Historical School. After 1945, the institutional economics associated in the U.S. with John Kenneth Galbraith was purged from American economics faculties, in favor of the “freshwater” (Chicago) and “saltwater” (MIT) versions of mathematical economics, which focused on trying to model the economy using equations as though it were a fluid or a gas.”

Either Bylund and Lind are completely out of touch with what economists are doing now or I am. Their critique of economics is an old one. I’m not sure it ever really applied, but it does not now. The American Economic Review and other top journals are full of empirical research, not lots of new papers about General Equilibrium.   

What sort of work do economists admire? Here are the John Bates Clark medalists since 1990. How many are known as pure theorists and how many are known for the empirical research?

1991	Paul R. Krugman	Massachusetts Institute of Technology	United States
1993	Lawrence H. Summers	Worldbank	United States
1995	David Card	Princeton University	Canada
1997	Kevin M. Murphy	University of Chicago	United States
1999	Andrei Shleifer	Harvard University	United States
2001	Matthew Rabin	University of California, Berkeley	United States
2003	Steven Levitt	University of Chicago	United States
2005	Daron Acemoğlu	Massachusetts Institute of Technology	Turkey United States
2007	Susan C. Athey	Massachusetts Institute of Technology	United States
2009	Emmanuel Saez	University of California, Berkeley	France
2010	Esther Duflo	Massachusetts Institute of Technology	France
2011	Jonathan Levin	Stanford University	United States
2012	Amy Finkelstein	Massachusetts Institute of Technology	United States
2013	Raj Chetty	Harvard University	United States
2014	Matthew Gentzkow	University of Chicago	United States
2015	Roland G. Fryer, Jr.	Harvard University	United States

Bylund and Lind seem to think that economists all aspire to be Samuelson, Arrow or Debreu. Yes formal models with lots of imposing math are still to be found, but more often than not they lead in to empirical research.

Are there things that economist can do better? Yes. I, for instance wish that economists would give as much attention to the evidence that they use as they do to the formal model and the choice of econometric techniques. On the topic, Mary and I have a paper on “The Historian’s Craft and Economics” that I am happy to say was just accepted by Journal of Institutional Economics. I also wish they would give more attention to history generally, but I’m not really an unbiased source on that topic. 

Micro economic history and some digital history too

William Easterly, Laura Freschi and Steven Pennings argue for the benefits of micro economic history of development by examining the development of one block in lower Manhattan over more than three hundred years. They have produced a fascinating paper and website.

Round table on Edward Baptist's Half has Never Been Told

The September Journal of Economic History has a round table of reviews of Edward Baptist’s book The Half Has Never Been told, with reviews by Alan Olmstead, Jonathan Pritchett, Trevon Logan and Peter Rousseau. They each address different aspects of the way that Baptist misrepresents the historiography of American slavery and makes things up. Thanks to Alan Olmstead for mentioning one of my blog posts on the book. Many of the points noted in these reviews  are similar to ones that I and Pseudoerasmus made about the book shortly after it came out, around the same time it was getting glowing reviews in places like the New York Times Book Reviews. I found Logan’s review particularly interesting when it stepped away from what is typically thought of as economic history. He concludes

  

I think as an economic historian I was so offended by the books portrayal of economic historians I may have missed some of the bigger problems. 

Tariffs and the Civil War, or 95% of All Statistics Are Made Up

A recent letter to the editor of our local paper argued that secession and the Civil War were caused by high tariffs not slavery. The Confederate states were rebelling against high taxes and big government. Apparently, they were really just Reagan Republicans or maybe even libertarians (slaveholding libertarians). The author of the letter made the claim that the South paid 75 percent of the tariff revenue in 1859. I thought the claim was so outrageous  he must have just made it up. It turns out you can find this claim all over the internet. It turns out it even has academic credentials behind it. Some people attribute it to Walter Williams, but he appears to have gotten it from Thomas Di Lorenzo, who attributed it to Frank Taussig’s The Tariff History of the United States. Di Lorenzo, however, did not provide a page citation. I suspect that he did not provide a page citation because one does not exist. If someone can find this in Taussig please let me know.

In any case, it is not true that most revenue came from Southern ports. A small fraction of tariff revenue came from Southern ports. In 1860 the Secretary of the Treasury reported the amount of revenue collected in each collection district between 1854 and 1859. (Sen. Ex. Doc. No. 33 36^th Congress 1^st Session). Looking at 1857, for instance, one finds that total revenue was $64,171,034. Most of the revenue, $42,510,753, came as it did every year from a single port: New York. The most important port in the South was New Orleans, which brought in a little more than $3 million, less than half as much as Boston. Southern ports were not even close to being the most important source of revenue.

There is no mystery as to why Southern states seceded. They issued secession proclamations explaining their actions. South Carolina was the first to secede, and the state’s proclamation does not mention tariffs. It is entirely about the perceived threat to slavery. It declares that 

“A geographical line has been drawn across the Union, and all the States north of that line have united in the election of a man to the high office of President of the United States, whose opinions and purposes are hostile to slavery. He is to be entrusted with the administration of the common Government, because he has declared that that "Government cannot endure permanently half slave, half free," and that the public mind must rest in the belief that slavery is in the course of ultimate extinction.” 

Apparently we are to believe that they were simply hiding their true motivation, opposition to tariffs. I wish modern defenders of the Confederacy were as honest as its original defenders.

History, Facts and Life Expectancy

Earlier this week on twitter Peter Bent mentioned Richard Yeselson’s review of Steve Fraser’s Age of Acquiescence: the Life and Death of American Resistance to Organized Wealth and Power in Dissent. One of the claims made by Fraser, and repeated by Yeselson, is that, although life expectancy increased during the Gilded Age, “it is also the fact that the life expectancy of white males born during or after the Civil War was ten years less than it had been a century earlier” (Fraser, 2015: 39). He provides a citation to Centers for Disease Control, National Center for Health Statistics. That is the entire citation. It is not clear whether it refers to a publication, a website, or personal correspondence. I checked the website for the Center. They do have statistics on life expectancy, but I only saw ones that went back to 1900. Historical Statistics of the United States has estimates of life expectancy, but they only go back to 1850. They show that life expectancy at birth increased from about 38 in 1850 to 40 in 1860 and 50 by 1900. If these estimates are reasonable and Fraser is correct, life expectancy at birth would have been between 50 and 60 years in the late 1700s.

There is one estimate that I know of life expectancy in the 1700 that is this large: Fogel, using family histories, estimated that life expectancy was greater than 55 years in the mid-1700s.(Robert William Fogel, "Nutrition and the Decline in Mortality since 1700: Some Preliminary Findings," in Engerman and Gallman Long Term Factors in American Economic Growth.

Fogel’s graph appears to indicate that life expectancy did not return to its mid 1700s level until the middle of the twentieth century. Personally, I’m skeptical of the accuracy of these estimates. They are much higher than other estimates. In the late 1700s, Wigglesworth estimated life expectancy in the mid 30s in Massachusetts in the late 1700s. Recently, Becker estimated life expectancy in the 1700s to be around 40, using data on people who attended Yale. In addition, Fogel notes that members of the British peerage had a life expectancy of only about 40 years in the late 1700s. It should also be noted that Fogel’s estimates of large decreases in life expectancy are consistent with estimates of large decreases average height, but there are good reasons to question the validity of that conclusion as well. If there were no large decreases in welfare reflected in average height, does it make sense that there would have been large decreases in life expectancy. In short, much of the available evidence seems hard to reconcile with very high life expectancy in 1700s America.

 I do find it plausible that there may have been a number of factors in the early nineteenth century that could have adversely affected health. Increased urbanization almost certainly increased the spread of disease. In addition, there were new diseases to spread, like cholera.

With some luck and a lot of work we will probably have more confidence in our knowledge of health and welfare in the eighteenth and nineteenth centuries. In the meantime, I am inclined to believe Becker’s estimates for the 1700s. That would mean that life expectancy increased very slowly during the nineteenth century, and then more rapidly after about 1900 as cities began to invest in sewage removal and water purification.  Chapter 3 of Higgs Transformation of the American Economy (still my favorite book on American economic history and now free from the Mises Institute) describes the impact of these improvements.

What is the point of all this rambling on about what we don’t know? The point is precisely that, we don’t know. I know it’s a lot to ask, but historians should take a critical approach to the evidence. Let people know when something is still up in the air. There is really nothing resembling a fact regarding mean life expectancy in the 1700s in America. There are a number of widely varying estimates. Don’t tell people we have “facts” that we don’t have. There are more, and more important, puzzles in history than what happened to the Roanoke Colony. Perhaps I’m getting old and cranky, but it seems to me that I have seen a lot of historians lately playing fast and loose with the evidence in order to make their point. And many of their reviewers do the same: they evaluate the book on how well it conforms to their preconceptions. 

Some random stuff

Did heights actually decline during the industrial revolution in the United States? Probably not.

New website about research on bankruptcy.

The history of Kool Aid at the Hastings Museum.

Until I was 9 I lived a block away from the Hastings Museum. My Grandma Schneider bought my brother and me annual passes. We spent a lot of time there as kids and went back for the first time in over thirty years last week. It is still one of my favorite museums. Some of my other favorites are Pioneer Village in Minden, NE, the Deutsches Museum in Munchen, the Frontier Culture Museum in Staunton, VA, the Royal British Columbia Museum in Victoria, B.C., and the National Museum of American History in D.C.

 

By the way if you are near Kearney, NE and want some good Mexican food go to El Maguey

How are prices determined? The case of statistical consultants

How are prices determined? AnnMaria De Mars offers advice to statisticians on how to price their services. It comes down to this

 “So, that’s it, decide a fair rate based on what the market is paying, where, based on objective criteria, your skills and experience fall compared to the general population of whatever-you-do and figure in what non-monetary requirements you or the employer have .” 

Dr. De Mars’ offers good advice and good economics. This is pretty much what I tell students regarding how businesses set prices, except I throw in a little economic terminology. She essentially describes a price that is a function of the price elasticity of demand. The price elasticity of demand is the percentage change in the quantity demanded in response to a one percent change in price. Other things equal, when the price of a good increases people buy less of it. Consequently, the more inelastic the demand for the product you sell, the greater your ability to mark up the price above the cost of production.

What determines elasticity? Elasticity is determined by the availability of close substitutes. The more close substitutes for the good you sell (the more elastic the demand), the less control you have over the price; the less close substitutes there are for the good you sell (the more inelastic the demand), the more control you have over the price. In other words, if you are pretty much like the other statisticians out there you need to charge what they are charging; you can only charge more if you can convince people that you are superior in some way. And, in the long run, you can probably only convince people that you are better than others if it is true. In other words, businesses that do not generally follow De Mars’ suggestions are unlikely to survive.

Understanding how prices are determined also provides a better understanding of business strategy. I tell students that if they plan on starting a business they should aim to be a monopolist. The essence of being a monopolist is that you are the only seller. To be the only seller, you need to convince customers that other goods are not a substitute for yours, and you need some barriers to entry, things that keep people from copying what you do. Fortunately for statisticians, they already have somewhat of a barrier to entry in that most people think that math is a lot of work and not much fun.

The other good point that she makes is that people should not just focus on the money. A lot of people think economists are totally focused on money. Nothing could be further from the truth about good economics. Economists assume that people maximize utility, which means satisfaction. People can get satisfaction from a lot of different things.

Two related things:

1. De Mars is trying to make math more fun.

2. One of De Mars’ daughters has done an extraordinary job of demonstrating that none of her competitors provide a close substitute for what she does.  

How much are auto workers paid in Mexico?

The Washington Post reports that “The Center for Automotive Research, a Michigan-based think tank, found that in salaries and benefits, car companies pay an average of $8 an hour for Mexican workers, while in the United States that figure would be four to seven times as high.” A few paragraphs later it reports on a walkout at a Mazda plant where the supervisor was abusive to the workers, stating that “For a job with 12-hour days, often including weekends, that paid about $75 a week — with $3 of that disappearing into union dues — some decided it was not worth it.” Forget about the weekends, $75 for twelve hour days five days a week would come out to $1.25 an hour. That is a lot less than $8. To reconcile the two either workers would have to get about $6.75 an hour in benefits or there would have to be a very high variance in wages. It is possible that both numbers are accurate. One number is an average while the other refers to a particular factory. The large discrepancy does, however, raise a lot of questions that the author and editors do not even seem to notice.

I really don't get Richard Thaler

I was listening to Here and Now yesterday and there was a discussion with Dan Gilbert and Richard Thaler about Thaler’s new book. In the discussion Thaler brought up the story of how he had told an audience of psychologists at Cornell about something like the life cycle theory of saving and how they had all laughed “hysterically.” He seemed to think it was another great example of how everyone else can see how getting a Ph.D. in economics subtracts “common sense” from economists. He probably hadn’t told them about the numerous empirical studies that found some degree of consumption smoothing. But haven’t they at least heard about the debt their students are taking on in the expectation that their future earnings will be higher. Haven’t they met anyone saving for the retirement they are looking forward to? Do they all really live as if there is no tomorrow? Really? Surely he can come up with a better example of the problem with economics than a theory that fits with common sense, casual empiricism and careful statistical analysis.

Thaler also said that the first sentence in every economics textbook is something like “People maximize utility.” Name one. It’s not in the versions of Mankiw, or Krugman and Wells, or Frank and Benanke, or Cowen and Tabarrok. I'm sorry. I really shouldn't keep letting the evidence get in the way of a clever story.

The Panic of 1907 and the Analysis of Financial Crises

I started to research the Panic of 1907 late in 2009. I came to the topic by a rather circuitous route. While working on my dissertation on the origins of the 1898 Bankruptcy Act, I also started to study the evolution of corporate reorganization, which wasn’t covered by the Act. That research ultimately appeared in Business History Review. Several important reorganization cases involved the Farmers’ Loan and Trust Company. The name was familiar to me from teaching American Economic History because of the income tax case, Pollock v. Farmers’ Loan and Trust Co., and two important railroad regulation cases, Reagan v. Farmers’ Loan and Trust Co. and Stone v. Farmers’ Loan and Trust Co.  I was curious what this company did that left its fingerprints all over nineteenth century legal and economic history. So I wrote a book about the Farmers’ Loan and Trust company and its influence on the law.

About the time that I finished the book there was increased attention to the Panic of 1907. The descriptions of New York City trust companies as novel, unregulated and reckless did not fit with what I had been reading and writing about trust companies like the Farmers’ Loan and Trust Co.  So I ended up writing a paper that argued that the panic was not the result of inadequate regulation of trust companies and that to understand the Panic one has to understand that not all trust companies were the same.

 What is really remarkable is that we know so much more about the panic of 1907 than when I started my work in 2009.

Rodgers and Payne have shown how gold shipments from France played a role in ending the Panic.

Hilt, Frydman and Zhou show how the Panic impacted the companies doing business with the trust companies that experienced runs.

Moen and Tallman, who were ahead of the curve in re-examining the Panic of 1907, have examined the transmission of the panic as well as  how the Clearing House Association helped to alleviate the Panic.

Most recently, Fohlin, Gehrig and Haas have shown the role that lack of transparency played in the panic in the stock market.

I believe that we have a much more about what happened in 1907 than we did just a few years ago, but these additions to our knowledge about financial crises in history should also promote caution. I like to think that my work will stand up to the test of time, but I’m sure previous authors did as well. It seems to me that the fact that we are still learning about the Panic of 1907 should cause economists to speak with some caution about the current economic events.

Stoller on Goffman and Ethnography

Paul Stoller examines the Goffman controversy and the future of ethnography. He recognizes that there are really two different sorts of issues involved. The first has to do with her interactions with her subjects. Stoller argues that emotional involvement with one’s subjects is likely to occur in ethnographic research and that ethical dilemmas can arise from getting close to one’s subjects.

“doing ethnography, like living life, involves love and hate, fidelity and betrayal, and courage and fear. Sometimes ethnographic experience brings us to face to face with issues of life and death--the real stuff of the human condition.”

 This seems reasonable, though, if in the process of research someone commits a crime, I think they should be prepared to accept the consequences.

 

Unfortunately, when he gets to the second issue, which has to do with methodology, I think he throws up a straw man. He asks

“But can we trust ethnographic accounts? Can ethnographers get "it" right? Given the infinite complexities of the social laboratory "the quest for certainty," as the philosopher John Dewey put it, is an illusion. If ethnographers cannot provide a perfect, scientifically verifiable representation of reality, how can anyone judge the contribution of an ethnographic work? This question, which has been raised by some of Goffman's critics, fails to fully appreciate the aim of ethnography.”

I believe we should try to get it right, but I think most of recognize that out understanding of the world is always incomplete, we can only have varying degrees of certainty depending on the degree to which the available evidence appears to support or contradict a particular belief.  I certainly do not want everyone to follow some supposed model of what is “scientific.” I don’t even know what “scientifically verifiable” means.

What I do ask is that a scholar’s attempts to persuade me involve more than saying “trust me.” What appears to be lacking in Goffman’s work is a means by which one can determine whether or not her interpretation is based upon empirical evidence, her observations, or on her imagination. This is particularly problematic because of the numerous inconsistencies within the story that she tells and the inability to find evidence consistent with some of her claims, described here and here and here and here.

In his own work on sorcerers, Stoller reported which villages he worked in. If I thought his stories of sorcery were a little far-fetched, I could visit Tillaberi and see if my observations of sorcerers resembled Stoller’s; I could even ask people if they had any recollection of Stoller. Anthropologists have done this and, occasionally, challenged the validity of earlier ethnographies: Mead on Somoa, and Chagnon on the Yanomami. It doesn’t seem to me that this sort of follow up is possible for Goffman’s study. Goffman writes about an anonymous group of people in an unidentified neighborhood in Philadelphia. Yes, I could go to Philadelphia, but if my experience was completely different than Goffman’s should could just say I got the wrong neighborhood. The problem is not that her work isn’t verifiable; the problem is that her work does not appear to be falsifiable. Any evidence that appears to contradict her work will be explained away.

I want to know how an impartial, or even critical, observer can evaluate her evidence. Michael LaCours and Michael Bellisales, just to name two, have shown that “just trust me” is not enough.