Takeaways from my MPhil dissertation

The last academic year was a whirlwind dominated by researching, writing, editing, and submitting my MPhil dissertation, "Normal Science and Anomaly: Responding to Unemployment in Economics in the 1920s." My motivation for the work was to explore the relationship between empirical evidence and economic theory, which is important for understanding how academic economic research relates to the "real world" -- a perennially interesting question.

To get at these bigger questions, I did a case study of a specific historical episode: persistently high unemployment in Britain in the 1920s. Unlike the United States and some other European countries, Britain experienced high unemployment (on average 12%) for the years preceding the early 1930s Great Depression. While now we have macroeconomics proper and can debate Keynesian, classical, monetarist, Austrian, and dozens of other flavors of macroeconomic models of the world, the 1920s was before Keynes's landmark General Theory and all of these developments. Economics in the 1920s was dominated by the market-clearing microeconomic models of Alfred Marshall from the 1890s, with cyclical business cycle theories developing on the edges. National income accounting was only in its infancy, and GDP had not even been invented yet. So there was little existing research or tools to help economists in the 1920s explain why the British economy had not rebounded from the post-WWI employment slump.

Persistent unemployment, then, presented an empirical anomaly that did not fit in with the major theoretical framework the economists of the 1920s were working with. What did they do? Surprisingly, the data I collected suggests that most economists simply ignored the problem. I made a text corpus of all the articles published in the Economic Journal, Britain's main academic economic journal, from 1921 to 1930. Of the 198 unique authors published in the Journal over those years, almost 90% did not write an article on the unemployment problem. A similarly small percentage of theses were written on the topic mid-decade. Even more surprising, unemployment received the same amount of attention in the Journal in the 1920s, during the unemployment crisis, as in the 1910s, when unemployment was low. Whether due to disinterest, discouragement, or simply a focus on other research goals, many economists were not publishing new ideas and new theories to explain the unique unemployment situation in 1920s Britain.

The rest of my dissertation focused on the 21 economists I identified who wrote on unemployment in the Economic Journal during the 1920s. For these economists, listed in the table, I read everything they wrote during the decade to understand the role that the issue of unemployment played in their work. Were they concerned with the level of unemployment in the economy? If so, did they believe that the new empirical evidence of unemployment required new theories to explain it? I also tried to understand how their views on the issue shifted throughout the decade, as the problem of unemployment grew increasingly persistent and urgent.


Broadly, what I found was that most of these authors tried (unsuccessfully) to use Marshall's microeconomics to explain the unemployment problem, pointing to immobile labor or high wages as an explanation for the unemployment. The new empirical evidence of persistent unemployment -- labor markets failing to clear for many years -- did not cause them to question the old models that they were used to working with, despite these models only being able to explain shorter-term unemployment. While I don't think we can infer too much from this single episode, it is interesting that in this case, many economists in my sample chose to stick to deficient theories rather than take the changing evidence about the "real world" into account.

The four main exceptions to this general conclusion were John Maynard Keynes, Arthur Cecil Pigou, John Rotherford Bellerby, and Henry Clay. I was especially interested in their work because each of them engaged with the unemployment issue extensively and developed new perspectives throughout the 1920s. But, I found that even in these cases, the connection between the reality of persistent unemployment and the development of theory was still pretty loose. While Keynes did make significant theoretical developments in the 1920s, especially the development of an early form of the multiplier, his work in this period was driven by his policy advocacy for price stabilization and state-sponsored investment rather than by a desire to reconcile the new experience of persistent unemployment with existing economic theory. Pigou recognized persistent unemployment as anomalous and inexplicable by the Marshallian economics he knew well, but instead of attempting to revise this economic theory, he detached his work from real-world verification and application. Bellerby, who has become an unknown, wrote passionately on the unemployment problem and recognized its inconsistencies with theory too, but chose to leave the discipline and pursue politics rather than contribute revisions to that theory. Finally, Clay explained persistent unemployment by rejecting unified economic theories entirely and drawing instead on diverse industry-specific experiences.

So after considering the corpus of articles published in the Economic Journal in the 1920s, these four economists in a lot of detail, and 17 other economists in less detail, I concluded that in the 1920s, economic research on unemployment was not reflecting and responding to the changing nature of unemployment in the real world. Though I was way, way out of words for my 30,000 word MPhil dissertation by this point (the first draft was over 65k!), I am hoping to build off of some of these findings in the DPhil dissertation I'll be working on for the next two years. For example, it's easy to draw parallels between the historical arguments I uncovered in this project and macroeconomic ideas currently being tested econometrically in the secondary literature. But, in both the 1920s and in the past few decades, the disaggregated export-industry-focused arguments of Henry Clay have been overlooked. By digitizing a trove of industry-level unemployment data from the Ministry of Labour for the interwar period, I'll be able to test claims like Henry Clay's as part of a larger story about labor market mismatch and immobility in local economic regions. But more on that coming soon!