The archival artifact that begins this post is a straight-forward response to a letter requesting possible leads for a junior faculty appointment in statistics at Bryn Mawr. It was written by Harvard assistant professor of economics Edmund Ezra Day (who would later go on to be the president of Cornell University–see link below) to a historian colleague at Bryn Mawr who had likewise done his graduate work at Harvard.
Two persons were identified by Day as eligible candidates, the Radcliffe graduate Anne C. Bezanson (about whom more can be found in an earlier post dedicated to her remarkable career) and a 54 (!) year-old economics graduate student Philip Green Wright. It turns out that Wright (with some collaboration with his son, the statistical geneticist Sewall Wright) is the rightly celebrated discoverer of instrumental-variables estimation. Relevant links to the story of Philip Green Wright and instrumental variables, including those to presentation materials as well as videos from a Tufts University Celebration of the 150th anniversary of Philip Green Wright’s birth, will be found below after Day’s letter.
There appears no expression of irony when Day writes “…if you are ready to appoint a woman, it will repay you to consider Miss Bezanson carefully”. Bryn Mawr was after all one of the so-called “Seven Sisters” (the Ivy League of women’s colleges).
________________________
Copy of Reference Letter from E. E. Day (Harvard) to H. L. Gray (Bryn Mawr)
re: Philip Wright and Anne C. Bezanson
March 21, 1916
Dear Howard,
Your recent letter was most welcome despite its obviously professional intent and largely professional content! I am glad to learn that you find life bearable in Bryn Mawr. That will serve as a preliminary report; in time I expected more exciting and promising announcement!
Regarding candidates for the new position in prospect in your department, I find it difficult to write anything at all definite. [John Valentine] Van Sickle is hardly available yet; he is still a couple of years from his degree and will probably not go out until he can take his Ph.D. with him. Furthermore, there is every prospect that, when he is fully prepared, he will command substantially more than the $1200 you mention.
The two students who would seem to be eligible for the position you describe are Philip G. Wright, rather an instructor than a student, and (if you would consider a woman), Miss Annie C. Bezanson. Neither holds the doctor’s degree, but both are very thoroughly capable students. Both are entirely capable of giving the instruction in statistics. Wright is a man well along in years, who for twenty-odd years taught mathematics and economics at Lombard College, Illinois, and, despite that fact, retains his intellectual vitality remarkably. He is perhaps a bit lacking in aggressiveness in classrooms, but is none-the-less an effective instructor. (You would probably have to pay $2000 to get him)
Miss Bezanson is a Radcliffe student whom I have had in both the elementary theory and graduate statistics courses. In the latter, last year, she did “A” work. She comes this year for her “generals “and any recommendation would be conditional upon her passing the examination credibly; but the staff expects her to pass with a large margin. It seems to me that, if you are ready to appoint a woman, it will repay you to consider Miss Bezanson carefully. Prof. [Frank] Taussig will write further details regarding both Wright and Miss Bezanson if you are interested. [Edwin Francis] Gay, too, has seen a good deal of Miss Bezanson’s work.
Let me know if I can be of further assistance to you, Howard. Mrs. Day joins me in warm regards.
Cordially yours,
[copy unsigned, Edmund Ezra Day]
Source: Harvard University Archives. Department of Economics, Correspondence and Papers 1930-1961 and some earlier. Boxt 26, Williams–Young. Folder “Wright, Philip Green”.
______________________
Biographical sketch of Philip Green Wright
“At Lombard, Philip taught economics, mathematics (including calculus), astronomy, fiscal history, writing, literature and physical education; he also ran the college printing press. Philip had a passion for poetry and used the press to publish the first books of poems by a particularly promising student of his, Carl Sandburg….
“…In 1912, Philip and Sewall [Philip’s son, a statistical geneticist] moved to Massachusetts. Philip took a visiting position teaching at Williams College, and Sewall entered graduate school at Harvard. In 1913, Philip took a position at Harvard, first as an assistant to his former advisor, Professor Frank W. Taussig, then as an instructor. Taussig was subsequently appointed head of the U.S. Tariff Commission in Washington, D.C. In 1917, Philip left Harvard for a position at the Commission, then in 1922 took a research job at the Institute of Economics, part of what would shortly become the Brookings Institution….
“…While at Harvard, in addition to his 1915 review of Moore’s book, he wrote a number of articles in the Quarterly Journal of Economics, and while at Brookings, he wrote several books and published articles and reviews in the Journal of the American Statistical Association, the Journal of Political Economy and the American Economic Review. Some of his writings used algebra and calculus, typically following graphical expositions. Although Philip wrote on a wide range of topics, the identification problem was a recurrent theme in his work (P. G. Wright, 1915, 1929, 1930). In his later years, Philip was particularly concerned about tariffs, and he wrote passionately about the damage being done by recent tariff increases to international relations (P. G. Wright, 1933).
“…In our view, this evidence points toward Philip as being both the author of Appendix B and the man who first solved the identification problem, first showed the role of “extra factors” in that solution and first derived an explicit formula for the instrumental variable estimator. Yet, as historians of econometrics like Christ (1985) and Morgan (1990) point out, a greater mystery remains: Why was the breakthrough in Appendix B ignored by the econometricians of the day, only to be reinvented two decades later?”
Source: James H. Stock and Francesco Trebbi. “Who Invented Instrumental Variable Regression?” Journal of Economic Perspectives. Vol. 17, No. 3 (Summer, 2003), pp. 177-194.
______________________
Three Worthwhile Links
Philip G. Wright, The Tariff on Animal and Vegetable Oils, 1928.
James Stock’s webpage: “The History of IV Regression”.
______________________
Philip Green Wright, Double Jumbo and Inventor of IV Regression
Sesquicentennial of the birth of Philip G. Wright
Tufts University Economics Department, October 3, 2011
Presentations
James Stock’s slides “Philip Green Wright, the Identification Problem in Econometrics, and its Solution“.
Joshua D. Angrist’s slides “Instrumental Variables in Action“.
Kerry Clark’s slides “Philip and Sewall Wright: The Invention of Instrumental Variables Regression“.
“Remembrances” by Philip Green Wright’s Grandchildren.
Video of the event
(Warning: poor audio)
Part 1: https://www.youtube.com/watch?v=INbip-UFluo
Opening remarks by Chairman of the Tufts economics department: Professor Enrico Spolaore
Welcome by the President of Tufts University. Anthony Monaco
James Stock begins at 8:20
Part2: https://www.youtube.com/watch?v=UvcfNk7rBn0
James Stock continues up to 9:20
Kerry Clark (AB Economics, Harvard 2012) begins at 10:30 [Ms. Clark’s other Harvard activities: Women’s Varsity Lacrosse, Center for History and Economics, Quincy Grille Manager, and Harvard University Women in Business. According to Linked In, she works at Citi)
Part 3: https://www.youtube.com/watch?v=4NWWCRaj4_Q
Kerry Clark continues to 5:20
Joshua Angrist begins at 7:30
Part 4: https://www.youtube.com/watch?v=Sp7g-L69MNU
Joshua Angrist continues for entire part 4
Part 5 https://www.youtube.com/watch?v=B3BPifHzex4
Brief Q&A
Grandchildren remember from 7:45 to 22.30
Image: Portrait of Philip G. Wright from James Stock presentation