Council of Economic Advisers’ Report Reveals Fallacies of Optimistic Forecasting
When President Bush put forth a set of modest tax proposals in his State of the Union address, he surrounded them with rhetoric borrowed from his glory days a year earlier when he was threatening Saddam Hussein rather than congressional Democrats.
Perhaps such hyperbole by politicians should not be a surprise. But it was disappointing to see the annual report of the Council of Economic Advisers join the chorus.
The CEA has the respect of economists at large for its professionalism and for the high quality of analysis. Most of this year’s annual report continues these high standards. But the section titled “The President’s Policies or Business as Usual” supports Bush’s tax proposals with projections that are more exaggerated than the President’s rhetoric.
The CEA projects real gross domestic product, unemployment and interest rates out to 1997 on two alternative assumptions--”Business as Usual” and “Policy,” meaning the proposals outlined in the State of the Union speech. The main proposals are a temporary, counter-cyclical stimulus to housing and to business investment through accelerated depreciation, plus permanent cuts in the capital gains tax and liberalization of individual retirement accounts.
The policy path has faster real GDP growth in each year, and by 1997, estimated GDP is 2.9% higher than the business-as-usual path. A small part of this difference is attributable to a stronger cyclical expansion along the policy path. The remainder, amounting to 2.2% of GDP in 1997, is the increase in the trend of GDP projected to result from the policy changes.
This would be an astonishingly large difference in the GDP trend to attribute to even a revolutionary change in economic policy. In fact, whether the permanent cut in the capital gains tax and liberalization of IRAs would have any positive effect on investment, and hence on the trend of GDP, is a matter of dispute among public finance economists because both measures add to the long-run budget deficit. But even granting the CEA’s belief that, on balance, these measures would add to investment spending, the size of the effects projected by the CEA still appears outside all reasonable grounds.
To see why, consider the section in the CEA’s report titled “Causes of the Slowdown in Productivity Growth.” There it reports that, according to generally accepted economic analysis, 1% more capital stock should produce between a quarter and a third of a percentage point more GDP. To get the 2.2% higher GDP attributed to the proposed policy would thus require a 6.6% to 8.8% larger capital stock. That, in turn, would require an unprecedented surge in investment.
The U.S. fixed capital stock, consisting of business plant and equipment and housing, was about $17 trillion in 1991. Along the business-as-usual path, by 1997 it could be expected to grow to about $19.5 trillion (in 1991 prices). The additional 6.6% to 8.8% required by the standard analysis amounts to a further $1.3-trillion to $1.7-trillion addition to the capital stock. To achieve this would require a steady rise in the fixed investment share of GDP to between 24% and 30% by 1997--a rise of 8 to 14 percentage points from the 16% investment share of 1991.
To see just how unprecedented such an investment surge would be, consider that over the past 30 years, the share of fixed investment in GDP has ranged between 14% and 17%. In the long and strong economic recoveries of the late 1960s and 1980s, the investment share of GDP rose briefly by about 2 percentage points from its cyclical low point before falling back again. And the latter period was marked by vast changes in the tax system, including the introduction of IRAs, investment credits and accelerated depreciation, changes in tax rates on corporations and individuals and changes in the taxation of capital gains.
Since the investment surge implied by the “generally accepted economic analysis” is not achievable, the CEA must be projecting the output gains from the President’s policy proposals on a different basis. In fact, the report does suggest that the standard analysis may be too conservative and refers to some hypotheses about how the payoff to investment may be greater than that analysis predicts. These hypotheses, referred to within the profession as new-growth theory, may prove to have merit and could modify the profession’s best estimates of the return to more investment. It would be good news if they did. But it is hard to imagine that they could modify the standard analysis enough to validate the CEA estimates of policy response. When most growth experts discuss how we might improve performance in the future, they emphasize the possibility that investment in education and the public infrastructure might be exceptionally rewarding rather than holding out the hope that the standard analysis vastly understates the returns to conventional investment.
