The statistics corner: it takes a long time to change statistical programs - The Business Economist at Work

Business Economics, Oct, 1993 by Joseph W. Duncan

Several significant changes took place. First, data on consumer income and expenditures that were collected every ten to fifteen years as part of the program to update and revise the CPI, were replaced by an ongoing consumer expenditure survey, which could, over time, be rolled up to provide up to date weights. Second, the 1978 revision program implemented modern scientific sampling procedures to ensure that the current CPIs properly reflect the market place on a continuing basis.

The new Index included three separate calculations:

1. A new CPI for all urban consumers,

2. A revised CPI for urban wage earners and clerical workers,

3. The old or "unrevised CPI" for urban wage earners and clerical workers, which was discontinued after the data for June 1978 were published.

The 1978 revision was the first major revision in the Consumer Price Index since 1963. Thus, we see that even for ongoing statistical series, the revision process as well as the implementation process is a lengthy period. Research frequently takes five to ten years and, of course, implementation in the sense of developing an ongoing time series requires even longer.

CONCLUSIONS

The development of a new data collection (the Survey of Income and Program Participation) and the revision of an existing data series (the Consumer Price Index and the measure of poverty) both involve long gestation periods. As a matter of fact, the issue over the home ownership component of the Consumer Price Index, which was particularly contentious in the mid-1970s, remains a controversial issue twenty years later.

It is certainly true that statistical measures should not be capriciously changed to reflect current fads or short-term trends. At the same time, however, in the rapidly changing political and economic environment of the global economy, it appears necessary to find techniques for adjusting our statistical programs more rapidly in order to capture the true context of current economic activity. We will return to that subject in the next issue.

FOOTNOTES

1 A detailed discussion of the project can be found in Mahoney, Bette S., "Review of Poverty and Income Distribution Statistics," Statistical Reporter, January, 1974, page 117-121.

2 For a detailed discussion of the SSA poverty standards, see Orshansky, Mollie, "Counting the Poor: Another Look at the Poverty Profile," Social Security Bulletin, January 1965, and "Who's Who Among the Poor: A Demographic View of Poverty," Social Security Bulletin, July 1965.