YOU ARE HERE: LAT HomeCollections


Note to the Candidates: Sky May Not Be Falling

September 28, 2003|David Friedman | David Friedman, a contributing editor to Opinion, is a senior fellow at the New America Foundation.

At no time in the last several years have California's jobs data, the key to measuring the state's economic health, been less reliable. Statistical inaccuracies inevitably crop up during volatile periods, such as a recession or a recovery, because the numbers are based on surveys that frequently lag behind realities on the shop floor. But these inaccuracies have been magnified by new, untested survey methodologies whose latent errors have not been fully identified, let alone corrected.

This is no esoteric matter. The target of the recall and the replacement candidates have been making strong statements about the condition of the state economy. Gov. Gray Davis has touted evidence that California is doing better than the rest of the country. At last week's debate, the five leading candidates to replace the governor overflowed with conflicting economic data. Problem is, the underlying data aren't reliable enough to support assertions about current economic conditions.

There's another problem. The aspiring candidates offer up economic prescriptions for ailments that may have no factual underpinnings next spring, when the data will be corrected. As a result, those who purport to be "saving" California from a job-loss calamity may be seriously misjudging the political consequences of their message.

Most of the debate about California's economic health depends on information contained in the monthly employment reports released by the state Employment Development Department. Unfortunately, the underlying data used to produce the reports are due for a major revision that won't be completed until well after the Oct. 7 election.

California's employment statistics are derived from a survey of a relatively small number of firms. Results are then used to extrapolate employment for the state and various counties and regions. The accuracy of these estimates depends on how closely the assumptions used to infer the broad numbers actually reflect real-world trends.

The process is most reliable when the economy is stable. When recessions or growth spurts depart from previous trends, however, state employment estimates can go astray. To correct for this problem, the data are "benchmarked," or tested against actual payrolls and tax receipts. The underlying assumptions in the employment model are accordingly modified.

Since late 2000, several factors have unsettled California's reported employment data. First, the recession bit so deeply and so fast, especially in the once highflying Bay Area, that company survey reports could not keep pace with actual job cutbacks and business failures. Once the lag was identified, statewide employment estimates had to be revised downward by nearly 100,000 jobs in 2000, and by an additional 175,000 jobs in 2001. Official reports thus didn't fully reflect the depth of the recession until well after the fact. It's likely that current job data will also have to be adjusted at the next benchmarking, in March.

Confounding matters further, California's employment survey methodology was significantly revamped in 2002. It now uses a new, randomized system designed and implemented by the federal Bureau of Labor Statistics. This approach may prove much more accurate than previous techniques. As with anything brand new, however, it's virtually certain that it will need to be fine-tuned before its accuracy can be assured.

What this means is that since early 2002 -- when the new survey method was fully introduced, and after the last benchmark was completed -- state employment statistics have become increasingly uncertain. From August 2002 to August 2003, for example, the three-month rolling average of reported employment results indicated that California lost about 44,000 jobs, a 0.3% decline. If true, that's less severe than the pace of job losses reported for the nation as a whole. Yet, over the same period, reported regional employment suggests that the state instead lost a net 93,000 jobs, a 0.7% setback that would be roughly twice as bad as the rest of the country.

Many experts believe that jobs will be added to the regional numbers in part because the statewide figures are thought to be based on more complete samples of company surveys. Consistent with this view, the latest California Department of Finance bulletin indicates that state tax receipts for the current fiscal year are up about 10% compared with 2002. That seems to indicate a stabilizing economy.

Los Angeles Times Articles