Report | Wages, Incomes, and Wealth

Missing the moving target

Issue Brief #197

Download PDF

Read a follow-up to this release,  White House backs off CEA prediction of “average” job growth (February 18, 2004)

Missing the moving target

Meager job growth and the poor track record of the administration’s job forecasts

by Jared Bernstein, Lee Price, and Isaac Shapiro

This Issue Brief is a joint release by the Economic Policy Institute and the Center on Budget and Policy Priorities.  CBPP

The Council of Economic Advisers (CEA) writes the Economic Report of the President, the administration’s major annual assessment of economic trends. In each of the last three years, the report has included the administration’s employment forecast for coming years. The following conclusions stand out from an assessment of these forecasts:

Job growth would have to improve dramatically to meet the CEA’s just-released prediction of an average of 132.7 million jobs in 2004. The administration’s prediction would only be met if job growth averaged more than 450,000 new jobs each month, about four times the level of job growth in January.1

Actual employment levels in recent years have fallen far below administration forecasts. In its 2003 report, for example, the CEA predicted that the average number of jobs in 2003 would be 1.7 million higher than its average in 2002. Instead, it was 400,000 lower. (See Table 1.)

Each successive year, the administration has had to lower its starting point for jobs, but it has forecast strong growth just around the corner. For instance, in its 2002 report the CEA predicted there would be 138.3 million jobs in 2004. Now its prediction of 132.7 million (which is also likely to be too optimistic) is for 5.6 million fewer jobs than its 2002 forecast.

Table 1

Among other implications, these outcomes indicate that the administration’s tax cuts and economic policies have fallen far short in terms of expected job creation. This conclusion is buttressed by an ongoing effort by the Economic Policy Institute to compare job growth since the enactment of the 2003 tax cuts to the expected growth predicted by the Council of Economic Advisors. Through January 2004, EPI has found that 1.8 million fewer jobs have been created than was projected by the administration.

These findings also underscore how weak job growth has been during the recovery, according to the administration’s own forecasts. The chairman of the CEA has said that its recent forecast reflects a rate of job growth “that is about average for a recovery.”2 Presumably, a similar approach also was used in previous projections. Thus, the dramatic degree to which actual job growth has fallen behind previous CEA projections underscores how significantly below average job growth has been. And to the degree that the most recent CEA projection turns out to be too high, this would underscore how significantly below average job growth will continue to be.

Employment predictions for 2004

The CEA’s forecasts focus on nonfarm payroll employment, the most reliable measure of the number of jobs in the economy.3 In its Economic Report of the President for 2004, released on February 9, the CEA predicts that such employment will average 132.7 million in 2004, reflecting a 2.6 million increase in jobs over its average in 2003. This estimate provides a benchmark by which to measure whether the administration will reach its own new expected level of job growth.

• In part because job growth has been tepid in the three months since the administration’s projections were made, to achieve the estimate of 132.7 million jobs during 2004, an average of 460,000 jobs a month would need to be created from February through December 2004. In other words, about five million jobs will need to be created between now and the end of the year to hit the administration’s projection.4

• It is quite unlikely that the administration’s target will be reached. In January 2004, jobs increased by 112,000—a modest fraction of the necessary level for the next 11 months. Furthermore, the January 2004 level of job creation was the largest single-month gain since 2000.

Is the administration already backing off its new forecast?

Several news stories have suggested that the administration may already be revising its employment projections downward in the wake of the disappointing job growth that has occurred since its predictions were formulated. A February 10, 2004 story by Alan Beattie in the Financial Times, for instance, said that “The [CEA] forecast, which used data available in early December before figures were released showing disappointing jobs growth at the end of the year, envisaged the average level of payroll jobs rising 2.6 million in 2004 to 132.7 million. That target level has been affected by the weak figures released since, but White House staff admitted their forecast still implied job growth averaging about 325,000 a month this year—well in excess of recent rates.”

When, however, CEA Chair N. Gregory Mankiw was asked during February 10 testimony before the Joint Economic Committee whether the CEA has decided on a new forecast, he declined to provide one or to say if he thought the forecast it published was too high.

If a new forecast becomes official, we will update this analysis accordingly. In the meantime, it is worth noting that job creation of 325,000 a month is also highly optimistic, and would be more than four times the average job creation during the past five months. See the appendix for a detailed analysis of this issue.

Previous projections have been far too optimistic

As shown in Figure 1, the CEA job-growth forecasts have consistently been too optimistic. The figure shows three forecasts for the level of payroll employment in 2004, published successively in 2002, 2003, and just this week. In 2002, their forecast for 2004 was 138.3 million. By 2003, the reality of the weak labor market forced the CEA to downgrade its 2004 forecast to 135.2 million jobs. This year their forecast is for 132.7 million jobs in 2004.

Figure 1

Even though the latest forecast is 5.6 million below the forecast of two years ago, it still seems implausibly optimistic. Figure 2 plots the administration’s last three job forecasts through 2004, along with actual job growth through last year. The top line in the figure is the forecast published in early 2002, which clearly diverged sharply from the reality of the jobless recovery that lasted through mid-2003. For the CEA’s next forecast, shown in the line second from the top, they scaled down their initial prediction but then forecasted strong job growth beginning in 2003. Once again, reality contradicted the CEA’s model, and their most recent prediction begins from an even lower base, and once again forecasts strong growth.

Figure 2

In each case, the CEA told us to expect a prompt and steep rebound in employment. In effect, the administration constantly tells us we a
re about to turn the corner and begin robust job creation, only to push the “corner-turning” out another year when reality shows otherwise.

CEA predictions including the effects of the 2003 tax cut

Since the summer of 2003, there has been some small growth in overall employment since the legislation was adopted, but it has fallen dramatically below administration expectations (see In sum:

• The CEA predicted that, with the passage of its tax cut plan, 5.5 million jobs would be created in the 18 months from July 2003 through December 2004.

• In the first seven months, however, only 296,000 jobs have been created, or just 5% of the administration’s projection. Since the administration projected that 2,142,000 jobs would have been created by now, actual job creation fell 1,846,000 short of this mark.

• To reach the 5.5 million target by the end of 2004, job growth would have to average 473,000 per month over the next 11 months.

• In effect, for the administration’s projection to be met, virtually all of the 5.5 million jobs that it felt would be created in the wake of last year’s tax bill will have to be created over the next 11 months.

The CEA projections and the Bush Administration’s explanation for job growth trends

The administration has been attributing weaknesses in the economy and the job market to a variety of factors, such as the events of September 11, the economy’s downturn, the bursting of the stock market bubble, corporate scandals, and the wars in Afghanistan and Iraq. It has argued that its tax cuts have made job growth and the economy better than it otherwise would have been.

In assessing these arguments, the following points must be considered:

• The first CEA report of the Bush Administration was issued after 9/11, after the bursting of the stock market bubble, after the war in Afghanistan, and after its first (and largest) round of tax cuts, yet their employment projections still proved far too high.

• The second CEA report came after a second round of tax cuts, after Enron and most of the other corporate misreporting scandals were publicized, and with the knowledge that the war in Iraq was, in all likelihood, about to commence. Nevertheless, the Bush Administration’s employment projections again proved remarkably optimistic.

• This year’s predictions have occurred well after all the factors to which the administration has attributed the weak economy, yet its predictions, at first blush, also seem well off the mark.

In short, the CEA should have been accounting for these factors all along in its job growth projections. Among other conclusions, this suggests that Bush’s CEA consistently has been far too optimistic about the positive effects of the administration’s tax cuts and other economic policies on job growth.


Why the CEA’s estimate implies monthly job growth of 460,000 for February-December of 2004.

As emphasized in the text, the CEA forecasts an average employment level of 132.7 million in 2004. In January of this year, the level of payroll employment was 130.2 million. As the Appendix Table A shows, in order to generate a 2004 employment level of 132.7 million, job growth will have to average 463,000 per month from February to December 2004.

Updating CEA’s prediction with more recent data.

In the Economic Report of the President published this week, the CEA predicted that employment would average 132.7 million in 2004, or 2.6 million more than its prediction of 130.1 million jobs for 2003. It based its data on information available through December 2, 2003. As of that date, monthly employment data were only available through October, so the CEA predicted employment growth from November 2003 onwards. However, when we incorporate more recent data, the results are very much the same as in our analysis.

Column one of Appendix Table B provides month-by-month employment figures that are consistent with the CEA’s original prediction (the CEA did not itself publish monthly figures). From January through October 2003, column one reflects the actual payroll employment levels then available. From November 2003 through December 2004, column one assumes that employment would grow by 300,000 per month (see table addendum). This monthly job growth level yields the CEA average levels of 130.1 million in 2003 and 132.7 million in 2004.

The job levels for January through October 2003 have, for technical reasons, been revised slightly downwards since CEA locked in its prediction. Column two of Appendix Table B includes these revised figures for the first 10 months of 2003. It then assumes the same number of jobs would have been created since then—300,000 a month—as the CEA’s prediction appears to assume. In other words, in column two, we use newly revised payroll data through October of last year to show what the CEA’s forecast would look like if they had these data available when they did their analysis.

The results change very little—instead of forecasting a growth of 2.6 million jobs between 2003 and 2004, CEA would have predicted 2.5 million. Instead of jobs averaging 132.7 million in 2004, it would have predicted 132.5 million.

Column three takes the new prediction of 2.5 million jobs, which CEA would have made had they been able to account for the technical revisions to the payroll data, and calculates what is needed in terms of monthly job growth to meet that goal, given actual job creation through January of this year. This column shows that CEA’s basic underlying assumptions mean that we will need 425,000 jobs per month to hit their adjusted prediction for average payroll employment of 132.5 million in 2004. In part, job growth has to be spectacularly high for the next 11 months for the CEA projection to be met because actual job growth for the other three months covered by its projections—November and December of 2003, and January of 2004—has been disappointing.

Appendix Table A

Appendix Table B


1. We include an appendix that incorporates technical revisions to the CEA’s estimates. These revisions reflect a recent downward adjustment of 188,000 to October 2003 employment levels that was unavailable to the CEA when they made their published projections. This results in an estimate of 2.5 million more jobs, on average, in 2004 over 2003. This adjustment changes none of the qualitative conclusions of our analysis.

2. Reuters News Service, “White House advisor says sees big job gains in 2004,” February 10, 2004.

3. For a detailed analysis comparing the surveys, see Elise Gould, Measuring Employment Since the Recovery—A Comparison of the Household and Payroll Surveys at

4. There has been some confusion in the media on this issue, with many stories incorrectly suggesting that the CEA projects a total of 2.6 million jobs to be created this year. In effect, these reports are stating that the CEA has predicted there will be 132.7 million jobs at the end of 2004 when, in fact, the CEA has predicted that the average number of jobs for all of 2004 will be 132.7 million. To reach this average figure, there will have to be many more jobs than 132.7 million in December 2004, as there are 2.5 million fewer jobs than that right now (see Appendix Table A).