|
Long-term Productivity Projections
May 25, 1999
The productivity assumptions in the 1999 Trustees Report assume that the slow productivity growth of the last quarter century will persist into the indefinite future, with absolutely no rebound whatsoever. At the least, this is a rather pessimistic view of a future of which we can claim little knowledge. Based on the United States productivity experience in the more distant past, and the current experience of other OECD nations, it is questionable whether it is the most plausible assumption about future productivity growth. It is worth noting that the 1999 Trustees Report effectively revised down the assumption for long-term productivity growth from the previous year. While it kept the projection the same, at 1.3 percent annually the 1999 report incorporated the impact of recent changes that were made in the GDP deflator. These changes should lead the GDP deflator to report a rate of inflation that is approximately 0.2 percentage points lower annually, than the GDP deflator in place at the time the 1998 Report was issued (Economic Report of the President, 1999, p 94). This means that the measured rate of growth of output, and therefore productivity growth, should be approximately 0.2 percentage points higher each year, when measured against the new GDP deflator, compared with a measure based on the old GDP deflator. Since the projected rate of productivity growth was held constant at 1.3 percent, in spite of the new deflator, this amounted to a reduction of 0.2 percentage points in the projected rate of productivity growth. In fact, 1.3 percent annual productivity growth would actually imply somewhat further slowing when measured against the business sector as a whole. The table below shows the published rates of productivity growth in the business sector and the non-farm business sector for both the period of the productivity slowdown, and the longer period since 1947. Bureau of Labor Statistics (BLS) Published Rate of Productivity Growth Non-Farm Business Business 1973-98 1.14 1.28 1947-73 2.83 3.29 1947-98 2.00 2.30 The published data do not include the impact of the use of geometric means for years prior to 1995, nor the changes in the treatment of some medical procedures that were put in place in 1993 and 1994. There have also been other, smaller changes in methodology implemented in recent years (e.g. the use of hedonic regressions for measuring quality improvements in television sets), most of which would have had the effect of lowering the GDP deflator relative to the true rate of inflation, and thereby raising the measure of real output and productivity growth. While it is not possible to determine precisely the impact that these changes would have had if they were applied consistently in the past, an estimate of 0.2 percentage points annually (mostly attributable to the geometric means and the changes in the measure of medical services) would be conservative based on existing research. This implies that the measure of past productivity growth would be slightly higher than those shown above. The table below incorporates this adjustment. Rate of Productivity Growth Adjusted for Measurement Changes Non-Farm Business Business 1973-98 1.34 1.48 1947-73 3.03 3.49 1947-98 2.20 2.50 The numbers in this table suggest that the current assumption on productivity growth implies a small slowdown even from the growth rate of the years 1973-98. The slowdown compared with the longer period since 1947 is almost 50 percent. The ability of economists to project productivity growth rates for the distant future is obviously extremely limited. For example, virtually no economists foresaw the productivity slowdown that began in 1973, even after it had already begun. The difficulty in making such a projection is demonstrated by the fact that virtually the entire slowdown in productivity growth was attributable to a decline in the rate of multifactor productivity growth. For the non-farm business sector the decline in multifactor productivity was 101.9 percent of the decline in labor productivity. For the business sector as a whole, the decline in multifactor productivity was 89.6 percent of the decline in labor productivity. (This comparison is not precise, because the periods are not identical.) By definition, multifactor productivity growth is an increase in output not directly explainable by factor inputs. BLS Published Rate of Multifactor Productivity Growth (all years presently available) Non-Farm Business Business 1973-96 0.16 0.30 1948-73 1.87 2.10 1948-96 1.05 1.24 (source: BLS Multifactor Productivity Trends) In looking to the distant future, there is no obvious reason to assume that productivity growth will either be closer to that of the low growth years from 1973-98 or the fast growth years of 1947-73. It is important to recognize that any projections of multifactor productivity growth are essentially a guess, not an informed judgement. We do not even know what technologies will be driving the economy in twenty or thirty years, just as economists failed to predict the importance of the personal computer or the Internet in the sixties or seventies. If an error in either direction is viewed as equally costly, it would be most appropriate to project that productivity growth will eventually settle between the two extremes we have witnessed in the last half century, near the average for the longer period. There certainly seems little basis for assuming that the rate of productivity growth will settle to a level that is even lower than that of the last quarter century, as the projection in the 1999 Trustees Report implies.[1] International Comparisons While the divergence in productivity growth rates between the early post-war period and the last quarter century may provide no clear basis for projecting productivity growth in the distant future, productivity trends in other OECD nations may provide some guidance. While most OECD nations also experienced a slowing of productivity growth in the last two decades, in most cases their growth rates remain considerably above those in the United States and show no signs of slowing further. Several of these nations now have absolute levels of productivity that are the same or greater than those in the United States. The question that this raises for an assessment of the prospects for long-term productivity growth in the United States is whether it is plausible that the level of productivity in the United States will far behind other nations, and continue to diverge indefinitely. Furthermore, it must be recognized that this divergence would be taking place in the context of an international economy that can be assumed to be far more integrated than it is presently. The table below shows two different measures of labor productivity for the G-7 nations. (This discussion focuses on the G-7 for simplicity. Similar productivity trends could be shown for other OECD nations.)
OECD
Conference Board Canada
1.0
0.7 * Or latest year available: 1994 for Germany; 1995 for Italy and Japan. (source: Schmitt and Mishel (1998) analysis of OECD data and Conference Board 1997)[3] The significance of this table is that the other G-7 countries, with the exception of Canada, are currently experiencing productivity growth rates well in excess of the 1.3 percent annual rate projected for the United States over the next seventy five years. It is important to recognize that this difference can no longer be attributed to a process of catching up to the United States. Several of these nations now have absolute levels of productivity that are the same or even higher than the levels in the United States. The table below shows BLS and Conference Board estimates of GDP per hour for the G-7 countries. The BLS data is for 1996, the Conference Board data is for 1995.[4]
BLS
Conference Board (source: BLS Comparative Real Gross Domestic Product Per Capita and Per Employed Person and Conference Board 1997) Canada's productivity growth is comparable to that of the United States, and the more rapid growth in Japan and the United Kingdom can be at least partially attributed to a process of catching up. However, the three largest economies in the EU, France, Germany, and Italy, all have comparable absolute levels of productivity that are comparable to that of the United States, and a considerably more rapid rate of productivity growth. The implication of the continuation of current trajectories is that the United States will fall far behind the world leaders in productivity growth. If France, Germany, and Italy can maintain their recent productivity path, then by the middle of the next century they will have far surpassed the United States in output per person. By the end of the seventy-five year planning period, if productivity growth in the United States follows the course projected by the Social Security trustees, the United States will no longer rank among the world leaders in productivity. The table below shows the absolute levels of productivity attained by each of the G-7 nations in the years 2035 and 2075, assuming a continuation of trend growth for other nations, and that productivity growth in the United States follows the projections of the Social Security Trustees. The original levels are taken from the Conference Board's data.
OECD
Conference Board * The OECD figure averages West German productivity growth rates from 1979 to 1991 and total German productivity growth rates from 1991 to 1994. The Conference Board data refers to West Germany only. In the data based on the OECD derived measure of productivity growth, France and Italy will both have levels of productivity that are more than 20 percent higher than the United States by 2035. By 2075, France's productivity level will be more than twice as high, while Italy's will be nearly 70 percent higher. The projections based on the Conference Board's data show the United States falling even further behind. By 2035, only Canada is projected to have a lower level of productivity. Germany's level of productivity is projected to be more than twice as high as that of the United States, and Italy's more than 60 percent higher. By 2075, the levels of productivity in Germany, Italy, and Japan would all be more than twice those of the United States, based on these projections. This projected divergence in absolute levels of productivity raises the question as to whether it is plausible that in an increasingly integrated global economy, developed nations can have sustained divergences in levels of productivity. While this may not be an impossibility, it is at least difficult to describe the scenario that would maintain such a divergence. The restrictions on flows of capital among the industrialized nations are already quite limited, and presumably will be even smaller in the future. Similarly, if more efficient techniques are developed in some countries, transfers of technology should allow for the quick diffusion of such processes. If the Trustees are to project a rate of productivity growth that will place the absolute level of productivity in the United States far below that of other industrialized nations, there should at least be a theory that explains how such a divergence can be plausibly sustained. Alternatively, the Trustees could take the view that all nations will eventually experience a slowdown in productivity growth, so that they experience a growth rate comparable to that projected for the United States. While this is possible, there is no evidence to suggest that such a slowdown will occur. As noted above, several European nations have already effectively caught up to the United States in absolute levels of productivity, yet continue to maintain growth rates near 2.0 percent annually.[5] At this point, there is no evidence to support the view that these nations will experience a decline in their productivity growth rates. If there were a very compelling reason to believe that productivity growth in the United States would never bounce back even part of the way to its early post-war rate, then it may be appropriate to postulate that a similar productivity slowdown will eventually hit the other OECD nations, even though no evidence for such a slowdown exists at present. However, as was noted previously, the major factor driving the productivity slowdown in the United States was a drop in multifactor productivity growth, a phenomenon that is still largely unexplained. Since there is no theoretical reason for believing that productivity growth in the United States will remain slow indefinitely, there can be little basis for assuming a productivity slowdown elsewhere, without any evidence that would support such a view. In conclusion, if the panel is to recommend that the Trustees adhere to a long-term productivity growth assumption comparable to the 1.3 percent in the 1999 Trustees Report, then it should also have an explanation of how large divergences in absolute levels of productivity can be sustained through time in an increasingly globalized economy. Unless such an explanation can be produced, it does not seem plausible to assume that the United States will sustain a rate of productivity growth of just 1.3 percent rate for most of the next century. [1] It is worth noting that the system used by the 1991 Technical Panel on Assumptions to weight rates of productivity growth in prior years would yield an average annual rate of productivity growth of 1.55 percent in the non-farm business sector, using data through 1998 (1991 Advisory Council on Social Security, p 22-23). When the impact of recent changes in the GDP deflator are taken into account, this weighting system would produce an average rate of productivity growth of approximately 1.75 percent. [2] BLS Comparative Real Gross Domestic Product Per Capita and Per Employed Person.[3] "An Evaluation of the G-7 Economies in the 1990s." 1998. John Schmitt and Lawrence Mishel, Washington, D.C.: Economic Policy Institute. Conference Board, 1997. "Perspectives on a Global Economy: Understanding Differences in Economic Performance." New York: The Conference Board Report Number 1187-97-RR. [4] BLS only has data on hours worked for France, Germany, and Japan. BLS data on GDP per employed person imply that Italy has also attained an absolute level of productivity that is comparable to that of the United States. In 1996, its real GDP per employed person was $56,173, compared to $54,397 in Germany, and $53,912 in France.[5] It is worth noting that the World Bank simply assumed that real wage growth would average 1.0 percent annually in the United States, and 2.0 percent in other OECD countries, in its recent examination of the impact of aging populations (The World Bank, Averting The Old Age Crisis, Oxford: Oxford University Press, 1994, p 160). |
