Shiskin Honoree Kennickell Lauded For Finance Survey Work*

Federal Reserve Board senior economist Arthur Kennickell, head of the central bank’s Microeconomic Surveys Unit, will receive the 2007 Julius Shiskin Memorial Award for Economic Statistics.  NABE, one of three sponsors of the award, will honor Kennickell during the annual meeting Sept. 9-11 in San Francisco.

As head of the surveys unit, Kennickell “made significant contributions to the theoretical basis and the methodology for implementation” of the Federal Reserve’s Survey of Consumer Finances (SCF), according to the award citation.  Under his leadership, the SCF has become “the world’s pre-eminent wealth survey.”

The award also recognizes Kennickell for his “achievements as an international expert on the design and implementation of household economic surveys.”   In the international arena, he is currently working with the European Central Bank on a survey of household financial behavior, as well as advising other governments in similar efforts.

The other two Shiskin Award sponsors are the Washington Statistical Society, which honored Kennickell at its annual dinner in June, and the Business and Economics Section of the American Statistical Association.  The ASA section will join NABE in presenting the award in September.

Kennickell said he is “absolutely delighted to be the recipient of the award this year.”  He observed that NABE’s involvement in recognizing his contributions in the field of survey methodology and data collection “sends a strong signal” that business economists value such work.

“For the future of economics, I believe it is very important for major economics organizations, such as NABE, to encourage the study of data collection.  Typically, data issues are treated as regrettable necessities in completing a research project,” he said in a late July interview with NABE News.

Bob Parker, one of the NABE representatives on the Shiskin Award committee, commented that for economists, the Survey of Consumer Finances “is the major, if not only, source of reliable data on both household income and wealth and has been used for the analysis of many topics. For example, the 2004 SCF was used to study income inequality, homeownership, household ownership of types of financial assets such as stocks and IRA's, and the relationship of wealth and education.”

Growth of Survey Accelerated After Redesign

To a significant degree, Kennickell’s 23-year tenure at the Federal Reserve parallels the expansion and improvement of the Survey of Consumer Finances.   He joined the central bank’s Special Studies Section in 1984, soon after completing his Ph.D. in economics at  the University of Pennsylvania.  “My dissertation was a micro-level study of saving behavior, and I was very frustrated by the lack of appropriate data,”  he said.

Working initially under “the direction of the brilliant Peter Tinsley,” Kennickell said he found the research atmosphere one in which “anything creative and useful was able to flourish.”  Kennickell soon became involved in helping to resolve a number of technical issues pertaining to data collected for the 1983 SCF, working with Robert Avery, the survey’s first project director.  When Avery left the board in 1989,  Kennickell became project director.  There are six staff members working on the SCF under his direction—three economists, a statistician, and two financial analysts.

The 1989 SCF represented a major redesign of the questionnaire, the sample, and the processing procedures, Kennickell said.  “Since then the survey unit has been particularly dedicated to two things.  First, we have tried to keep the framework of measurement as steady as possible in the face of a changing financial world and evolving analytical needs.
Second, we have nurtured a process of data quality control that I believe is the key to the success of the SCF,” he said.   “A strong research agenda apart from methodological concerns is also critical,” he said.  “Without the insights that come with using the data, I believe the quality of the data would deteriorate.”

SCF staff analysts are currently reviewing data and reports from interviewers in the field for the 2007 survey, which was about halfway through the field work as of late July, he said.  The Fed plans to release a summary of results for the 2007 survey in early 2009.

Over time, the SCF has become the primary data source for economists on the wide array of issues it covers – from household wealth to demographic patterns of income and pensions.   When he first became involved in the survey in the mid-1980s, Kennickell said that “it seemed that virtually no one outside the Board had heard of the SCF.  It was a major effort to get academics to take the project seriously.”

In contrast, now the Fed often hears from job applicants who have presented seminars based on the SCF, he said. “The survey is widely used across the government, academia, and the private sector.  When researchers want to know about household wealth and financial behavior in the United States, they turn to the SCF,” he said.

Call for More Attention to Interviewers

From the perspective of a survey manager, Kennickell offered a critical comment about what he sees as an overlooked need to pay more attention to interviewers in key government data programs. “In most data collection efforts, interviewers are absolutely essential,” he said.  “Alas, they are often treated as if they were low-skilled workers who are necessary to harvest data.”

In conducting the SCF every three years, for example, the Federal Reserve contracts with the National Opinion Research Center (NORC), a social science and survey research organization at the University of Chicago. About 200 interviewers worked on the 2004 SCF, the most recent survey, in which more than 4,500 interviews were completed, representing some 112.1 million families.

Interviewers at “not a neutral element in data collection,” Kennickell said.  Rather, they face what he calls a “variety of incentives that shape the way they work.  We need to understand better what those incentives are and how they operate, in order to neutralize their behavioral effects more fully.”

When asked to describe these incentives, the Fed economist observed that “traditionally, interviewers have been judged on their ability to complete interviews—and only that.”   If there are difficulties in conducting the interviews, “there would be an incentive to avoid these difficulties just to get them done.”   Kennickell said that in managing the SCF, his staff has developed “stricter monitoring of how the cases are handled by interviewers.  We review the data and give feedback directly to the interviewers.”

Interviewing is a challenging job, especially in an environment where prospective respondents are worried about identity theft and confidentiality, Kennickell said.  Assuring respondents that detailed financial information will be protected as confidential is a top priority of SCF interviewers, he said. When households are identified randomly for possible inclusion in the survey, a letter from Federal Reserve Chairman Ben Bernanke is sent by NORC explaining the requirements and value of their responses.  Early returns in this year’s survey indicate the response rate could well be higher than in 2004.

Commenting more broadly on the question of interviewers involved in government data programs, Kennickell said: “The survey industry is in some ways still resting on its golden years.  In the 60s and 70s, there was a pool of highly educated and motivated women.  People got used to the availability of these women. The thinking hasn’t fully evolved” on how to change the system.   “I strongly believe that we need to treat social science interviewing as a professional activity that is held to a high standard and compensated appropriately,” he added.

Working with European Central Bank to Develop EU Survey

Since 2006, Kennickell has been working as an advisor to the European Central Bank “on a survey of household financial behavior that I hope will come to fruition.”

While he cannot discuss the details due to confidentiality rules, he said that he finds the project very interesting as EU countries work to develop measures that “sufficiently encompass the still large variation in financial services.”   A survey pre-test will be conducted in August in Ireland, he said, followed by testing early next year in Germany.

“Many of the countries in the euro area already have some sort of household financial survey.  I believe the Italians, followed by the Dutch, have been in this market the longest.  The Bank of Spain has a survey that I consider the most advanced.  It is fair to say that the SCF has been an important influence on wealth measurement in Europe and elsewhere,” he said.

In addition to managing the SCF, Kennickell has written extensively about survey design and data collection, contributing to the literature on topics such as sample design, nonresponse effects, weighting design, and data quality. “These articles have substantially enhanced the understanding of policymakers, professional statisticians, economists and the general public.  As a consequence, the SCF is the authoritative source of information on the ownership and distribution characteristics of U.S. household wealth, and the standard of quality against which household surveys around the world are compared,” according to he Shiskin Award committee.

Kennickell was elected a fellow of the American Statistical Association in 1998, served as the member of the Federal Committee on Statistical Methodology since 2002, and as a member of the board of the National Bureau of Economic Research since 2004. He also has been a member of the editorial board of the Review of Income and Wealth since 1994. 
He earned both Master’s and Bachelor’s degrees in the Humanities at the University of Chicago. 

*Views of Arthur Kennickell reported here are his personal opinions and do not necessarily represent the view of the Federal Reserve Board or its staff.