Bill Kruskal and the Committee on National Statistics

Statistic al Scienc e 2007, V ol. 22, No. 2, 273– 274 DOI: 10.1214 /0883423 06000000394 Main article DO I: 10.1214/0883 42306000000420 c  Institute of Mathematical Statisti cs , 2007 Bill Krusk al and the Committee on National Statistics Ma rgaret E. Ma rtin In 1 972 I w as working at the U.S. Bureau of the Budget in the Office of Statistical Po licy (formerly the Division of Statistical S tandards). I had b een there for some 30 ye ars and had b een app ointed an assistan t chief to co ve r the areas of emplo yment, p opulation and income statistics. I tol d t he story of ho w I came to b e executive director of the Commit- tee on National Statistics in a 1 994 in terview in Sta - tistic al Scienc e co nd ucted by Miron S tr af with In- gram Olkin in atte ndance. Let me quote from there: Bill Krusk al had called me to ask m e if I would accept the p osition of Executiv e Director to the newly-organized Commit- tee on Natio nal Statistics at the National Academ y of Sciences, and I had turned him d o wn. I wa sn’t immediately attracted to it and I didn’t think I’d d o a go o d j ob. So, months w en t b y , ma y b e nine months, and h e called me again. He w as c hairman of th e Committee. It wa s in the midd le of the bud get season. I had just had one of these a w kw ard incidents with Julie [Shiskin, then director of the Office of Sta- tistical Po licy] which I would h a v e p roba- bly forgotten in tw o or three days, b u t I hadn’t wh en Bill called me. F urthermore, I wa s in the middle of my 29th bu d get sea- son. It was enough, y ou know. Bill called up and said, “I know y ou turned me do wn, but I’m calling again b ecause I hav e a list of p eople . . . .” They wan ted someb o dy Mar gar et E. Martin is r etir e d fr om the p osition of Exe cut ive Dir e ctor, Committe e on National S tatistics, National R ese ar ch Coun cil, and lives in Mitchel lvil le, Maryland 20721 -2734, U SA e-mail: mgr andaunt@ aol.c om . This is an electronic repr int of the orig inal article published b y the Institute of Ma thematical Statistics in Statistic al Scienc e , 2 0 07, V ol. 22, No. 2, 273– 274 . T his reprint differs fro m the or iginal in pag ina tion and t yp ogr aphic detail. who had exp erience with the federal go v- ernment ’s statistical s y s tem, since few on the Committee h ad. He w an ted to read m e a list of n ames, and w ould I p lease com- men t on these p eople. I just heard myself sa ying (I hadn’t th ou ght ab out it at all) “I h ad n o idea the job w as still op en, but if yo u we re serious ab out asking me b e- fore, I’d like to b e considered on the list.” He said, “W ell, that’s it. I’m not going to read the list, then.” So, I never did fin d out who w as on the list. Bill wa s a great p erson to wo rk for. Lo oking bac k from this long d istance I realize th at he had to mak e great leaps b et w een the academic views of s tatistics and the go ve rnment adm inistrativ e view. He was trying to wo rk in tw o differen t cultures. He did a great j ob b ridging the gap. He con tributed a great d eal to the establishment of th e Committee on National Statistics, and it w as tough going. He hired me, but I h ad no exp erience in dev eloping prop osals or getting funds. W e had one y ear’s funding f r om the Russell Sage F oundation, and then one more ye ar from them at a red uced lev el, but they thought the gov ernment should pa y for this activit y . Bill and I s p ent a great deal of time together going to heads of go ve rnm en t agencies. In particular we wen t to the head of the National Sci- ence F oundation (NSF) and to the h ead of the math- ematics group, where statistics was thought to b e housed. T hey couldn’t hav e b een less interested. W e finally got su p p ort from the so cial sciences at the NSF. Bill was alwa ys c heerful throughout this ef- fort. Sur ely h e was very busy at Ch icago, but he called at least ev ery we ek to encourage us. He thought of him s elf as a one-man agency to try to dra w statisticians together thr ou gh h is corresp on- dence and forw arding materials to them. When it w as first established at the National Re- searc h Coun cil (NR C), the C ommittee wa s attac hed to the Assem bly of Mathematica l and Physical Sci- ences. Nob o dy there h ad muc h exp erience with gov- ernment statistics, but there w as a section of the 1 2 M. E. MAR TIN NR C that dealt with th e life s ciences—statistics wa s also imp ortan t there—and a section that dealt with the so cial sciences. T he staff officer in c harge of so- cial sciences wa s lo oking for additional fi elds to add to his group. He though t that reviewing/monitoring national p olicy statistics w ould logically fall there, either in b ehavioral science or in so ciology or in eco- nomics. T his attracted me b ecause I had disco ve red that the tw o things that are imp ortan t b et ween the committee and its h ome base were (1) advice on and appr o v al of pro jects and (2) suggesting appr o- priate p eople to pu t on individual panels for sp ecific pro jects. Und er the original arrangemen t, as staff officer I w as rep orting to geologists and physicists as w ell as m athematicia ns. Th ey w ere kind enough to me, t hey didn’t obstruct, but they did n’t add an ything b ecause they w eren’t equipp ed to do so. So I was pleased to suggest the c h ange, but it w as against Bill’s deep est feelings to lea ve mathematics. I admired Bill for un derstanding the necessit y of the c h ange. Thus it w as painful for Bill to prop ose that w e mov e and explain why , and to ask that nob o dy try to s top us, but Bill did it. He never criticized or assessed blame, whether we fell on our faces or n ot. He w as supp ortiv e, thought- ful, careful to try to direct us in what he though t w as the righ t dir ection, but in very subtle wa ys. He had wide in terests and wide acquainta nceship among statisticians, and these w ere very helpfu l. He laid suc h a sound basis for the Committee that it has flourished for o v er 30 yea rs.