We are gathered here to recognize three of the best economic history dissertations on North American topics completed over the last year. These three are competing for the “Nevins” Prize, named in honor of Allan Nevins. Given that Allan Nevins was born in 1890, many of us are too young to know who he was or what he did. Recently, I read for the first time a few of the 50 plus works of Allan Nevins listed in my library's card catalogue. I thought it might be most proper and instructive here to take a moment and hear a few of his words.

On 16 January 1967 at age 76 Allan Nevins delivered an invited address to the Library of Congress, which was published under the title “The Old History and the New.”¹

I will read some words from that address—shamelessly taken out of their thirty-plus-year-old context—that should still resonate with us today. Nevins said: Our three panelists here, and hopefully all of us, have in some way tried to lived up to and I hope will continue to aspire to live up to Nevins's words.

I do not intend to give lengthy comments on each dissertation. The panelists have presented their key methods and results, and I leave the painful dissection of the minutiae to future journal and book referees and editors—of which I am sure these panelists will soon receive more than their fill. Instead, I will comment on the mysteries of the selection process, and so in a roundabout way highlight the superior qualities of the three dissertations before us.

Overall, ten dissertations were received and deemed eligible for this year's Nevins Prize—four from West Coast universities, five from Mid-West universities, and one from an East Coast university. None were submitted from Southern or non-U.S. universities. In total, they included 233 tables, 124 figures, and 1,670 references embedded within 2,760 written pages, and I read every single page—well ok, I looked at every single page. The sample of submissions varied from a low of 133 pages to a high of 687 pages, from a low of four tables to a high of 46 tables, and from a low of 53 references to a high of 550 references. Quality, however, does not necessarily correlate with high numerical values.

Most of the dissertations focused on the late nineteenth and early twentieth centuries. One dissertation dealt with the late twentieth century and one with the first half of the nineteenth century. Alas, no dissertations on colonial or pre-Columbian economic history were submitted. Nor were any dissertations on Canadian, Mexican, Caribbean, Central American, or Native American economic history submitted. Although field topics varied, there did seem to be one prominent theme, namely some type of interface with issues relating to institutions and law and how they affected social welfare.

Being in charge of giving the Nevins Prize is truly an edifying experience, and is recommended to all—though perhaps an experience you should only be blessed with once in a lifetime. It forces you to see, perhaps for the only time in your career, a full cross-section of what the new research in economic history on North America is. I personally learned many new things, such as the techniques of “Generalized Entropy” for measuring inequality, “Social Structure of Accumulation” Theory, and “Principle Components Analysis.” I also learned that what I experienced first-hand during my college days is now considered fodder for historical study—a somewhat disconcerting revelation. And I learned that a dissertation should never have a “third” volume.

Reading the acknowledgments in a dissertation can also be curiously edifying. For example, one dissertation submitter acknowledged his “… great intellectual debt to Karl Marx who opened my eyes to the reality of exploitation.” This made me conclude that the life of a graduate student is not what is used to be, if the reality of exploitation is only experienced when reading Marx. Another dissertation submitter acknowledged the blessings of “God.” This in turn made me ponder the metaphysical source of standard errors and whether they were really necessary.

Because no dissertation on colonial America was submitted, there was no obvious winner, and so I had to invent some criteria, four in total, for ranking the submissions. First, I wanted to see some original data work—something more than just grabbing a canned data set, such as IPUMS or a data set from ICPSR or the BLS, and slicing it in yet another way, or more than simply reporting what was read in an archive like so much travelogue. I wanted to see either original source evidence sorted, coded, and built into a new quantitative data set from scratch, or two or more existing canned data sets merged and extended in an ingenious way not previously done, or original nonquantitative archival evidence not previously used brought to bear in some new and ingenious way.

Second, I wanted to see economic theory used as the driving force behind crafting the historical questions asked and techniques applied, not necessarily post-hoc math modeling or the latest econometric techniques, but hypothesis testing informed by the application of reasoning gleaned from core economic theory. Altough some dissertations were very good at applying the latest statistical and econometric techniques, this by itself is not economics. Thus even dissertations that applied the most current and sophisticated statistical techniques and math-modeling still needed to incorporate some strong hypothesis testing motivated by core economic theory in order to receive superior marks. I would hazard to say that as a profession we are doing a good job of teaching our students the current quantitative techniques and modeling methods, but we need to do a better job of teaching our students how to use basic economic theory, counterfactuals, and hypothesis testing to motivate their application of said techniques and methods. Economics as a useful science, as a research method, as an investigative tool is not just math. It is not just statistical estimation. It is not just models of post-hoc rationalizations. It is not just storytelling about the economy. Because grasping these notions and making use of them is probably the most difficult lesson to be learned in applied economics it received substantial weight among my criteria.

Third, I wanted to see a substantive tome—something with both breath and depth—that attacked its topic from many different angles and thoroughly explored the robustness of the techniques used and results found. Having only one substantive essay was not enough to receive superior marks on this criterion—even if that essay was good enough to be publishable in a top journal, as I believe some among the submissions clearly will be. In addition, I gave little weight to introductory essays that merely reviewed the literature uncritically and to concluding essays that just re-summarized the project's findings unless they were accompanied by more than one substantive essay sandwiched in between. This is not to say that I discounted the standard three-essay dissertation so popular in economics or treated them differently than the single large monograph dissertations, but it meant that each of the three essays in such dissertations needed to be substantive and the three needed a common theme linking them together into a coherent whole.

Fourth, the dissertation needed to make a significant new contribution to history—substantially advancing its sub-area of study. This criterion was harder to apply as I am not an expert in all fields and in all sub-areas. Here, I relied somewhat on the writers themselves to sell their own contributions as new and important in their introductory and concluding chapters. I think it is important for a writer to be convincing on this score and so it deserves some weight among the criteria.

Now, how I weighted the four criteria is hard to say. The weights I used are probably not time-consistent. If I evaluated the same lot five year ago or five years from now, I probably would generate a different rank ordering, at least among the upper half of the submissions, than I did this year. Maybe that is not a lot of consolation to those not selected or exceptional praise for those so selected, but such honesty should not be used to diminish the accomplishments of those before us now. For all three have brilliantly satisfied these criteria.

And now for a little summary and criticism of our panelists' dissertations—namely what do they most need to do to turn what is now an outstanding dissertation into a truly powerful contribution to the field. In alphabetical order, Rebecca Ann Holmes uses principle components analysis to condense into a manageable data set the seeming chaos of heterogeneity in the massive number of labor laws passed in the various states in the early twentieth century. This data set is then used to estimate the cumulative impact and evaluate the subcomponents of what was thought maybe to be unmeasurable, namely the impact of the totality of state labor laws on the country's overall labor demand and on the individual input demands for materials, salaried workers, and wage workers. Although thorough and extremely detailed, the dissertation could be made far more compelling with a livelier narrative and some illustrative storytelling about its subject. Its potential to reach the broader audience that it deserves to reach will be lost without such additions.

Duol Kim, using case studies and new detailed micro-level data, provides a significant advance in our understanding of the selection of firm ownership structures, namely partnerships versus proprietorships, and how these structures performed and survived conditional on various factors. I particularly liked the application of not only standard estimation techniques but also of ingeniously simple techniques, such as separating samples into partnerships where the partners were only family members versus partnerships where the partners were not family members, and then modeling, estimating, and interpreting the different outcomes. While the dissertation is fascinating reading and while I learned a lot, at the end of the day I am still not sure what it all means and how much it matters, i.e., what the true opportunity costs were. The dissertation will have far greater impact if such things can be addressed more adequately.

James Ireland Stewart constructs new merged or linked household samples thereby making important contributions to our understanding and ability to directly measure the extent and nature of self-selection in frontier migration to the Great Plains. These new data allowed Stewart to then examine the role that individual experience, knowledge, and information played in coping with regional climate variations, namely drought, and how that in turn affected the credit markets and feed the populist politics of the times. This three-essay dissertation, however, seems to lose steam in the third essay, which applies evolutionary games theory to mining camps in the far west. How this third essay coheres with the first two essays on frontier farming on the Great Plains is not well articulated. In addition, the model used in this third essay does not seem to lead to a clear set of well-executed or executable tests. The dissertation as a whole would be improved if the third essay on mining camps could be integrated more with the first two essays on frontier farming, and if the tests executed in the third essay were as neat as those executed in the first two essays.

In conclusion, we should all look forward to seeing the important contributions made by our three panelists to our discipline in print, hopefully, in the near future. Please welcome them into the profession.