1、Technological Change and Economic Growth: the Interwar Years and the 1990s,Alexander J. Field afieldscu.edu All Ohio Eocnomic History Seminar Ohio State Universiyty April 30, 2004,The Most Technologically Progressive Decade of the Century,Field, American Economic Review (2003) Hint #1: Its not what
2、you think Hint #2: It wasnt the 1990s. Hint #3: It wasnt the 1920s,The Main Argument,The years 1929-1941 were, in the aggregate, the most technologically progressive of any comparable period in U.S. economic history.,Y = real output N= labor hours K=capital inputY/N = Labor Productivityy n = Labor P
3、roductivity growth(lower case letters = compound annual average rates of growth)Y = A KN1- = Production function (Cobb Douglas, crts)A = Y/(KN1-) = Multifactor Productivitya = y k (1-)n = Growth Rate of MFPy - n = a + (k - n) = Growth rate of labor productivity,Labor and Multifactor Productivity Gro
4、wth Formulas,Disclaimers - 1,What we measure in the residual contains not only serendipitous or accidental discovery of useful knowledge, but a variety of related influences, including, but not limited to:The outcomes of focused research and development activities The influence of scientific and edu
5、cational infrastructures Economies of scale and network effects Learning by doing Reallocation of economic activity from sectors with low to higher value added per worker New organizational blueprints or managerial practices,Disclaimers - 2,Under certain conditions, the residual will underestimate t
6、he impact of technological change because of linkages between such change and the rate of saving and capital accumulation Biased technical change that increases the return to capital or to highly educated labor may shift income to households with higher saving propensities, increasing the aggregate
7、saving rate Technical change that increases the real after tax return to capital may elicit larger flows of saving Innovations that reduce the relative price of capital goods will, for a given saving rate measured at base period prices, be associated with a larger real volume of investment,Abramovit
8、z David (1999),Gordon (2000b),Conventional Wisdom,The measured peak in MFP growth rates between 1929 and 1948 is principally the consequence of the production experience of World War II: a persisting benefit of the enormous cumulated output as well perhaps of spinoffs from war related R and D.,New A
9、rgument,Peak MFP growth between 1929 and 1948 is primarily attributable to an exceptional concatenation of technical and organizational advances across a broad frontier of the American economy prior to full scale war mobilization Governmental and university funded research, as well as the maturing o
10、f a privately funded R and D system that began with Edison at Menlo Park played a role So too in some sectors, such as railroads, did the “kick in the pants” of cut off of easy credit availability and declines in demand,Why do we credit World War 2 with establishing the foundations for postwar prosp
11、erity?,Sheer volume of output between 1942 and 1945 is exceptional Remarkable successes in such sectors as airframes and shipbuildingBetween 1942:1 and 1944:4 airframe productionincreased by a factor of six, and labor productivity grew by 160 percent. This is a compound annual average growth rate of
12、 output per hour of 34.7 percentIn shipbuilding: in one ten-month period alone, thenumber of hours required to build a Victory shipfell by half (U.S. Bureau of Labor Statistics,1946, pp. 89798). On an annualized basis, this is a growth in output per hour of 83 percent a year,Why we should be skeptic
13、al,Overall increase in labor productivity in munitions sector between 1939 and 1945 was 25 percent far below standout sectors (Brackman and Gainsbrugh, 1949). This is a growth in output per hour of 3.71 percent per year, respectable, but hardly surprising given the more than $10 billion of public se
14、ctor capital invested in the defense sector Short period of full scale war production: roughly three and a half years Spillovers which direction? War drained skilled labor, managers, and capital from civilian sector. Output per hour stagnated in 1942-44 in the civilian sector Swollen productivity nu
15、mbers are partly the result of a temporary shift of output to sectors with traditionally high ratios of value added per worker could not persist,Why 1941?,1937 unemployment 14.3 percent 1940 - unemployment even higher 14.6 percent 1941 unemployment is 9.9 percent (6 percent according to Darby) Only
16、2.5 percent of cumulated war spending 1941-1945 had been undertaken by the end of 1941 1941 is the closest we come to recovery before full scale war mobilization,United States, Private Non-Farm Economy CAAGR of MFP, 1919-1948,Solow (1957),“there does seem to be a break at about 1930. There is some e
17、vidence that the average rate of progress in the years 190929 was smaller than that from 193049” (Solow, 1957, p. 316),Kuznets and War Planning,Simon Kuznets needed to estimate the potential output of the U.S. economy to determine war production plans consistent with planned force levels and civilia
18、n consumption. His estimates came in considerably higher than most had expected, leading the military to multiply their production targets and forcing Kuznets and others to fight a rear guard action to bring them down to a realistic level The outward shift of the production possibility frontier duri
19、ng the Depression years - largely unrecognized until then - was the principal reason potential output in 1942 was so much higher than had been anticipated.,Sectoral Evidence,R and D employment in manufacturing Time path of key innovations Kleinknecht Schmookler Mensch MFP growth in telephone, railro
20、ads, electric utilities Build out of surface road system: network effects impact on productivity growth in trucking and warehousing,R and D employment in US Manufacturing,1927: 6,274 1933: 10,918 1940: 27,777Source: National Research Council data; Mowery and Rosenberg, 2000,Kleinknecht (1987),The Ir
21、ony of Secular Stagnation,At precisely the moment when Alvin Hansen and others were developing theories of secular stagnation, the U.S. economy was experiencing its greatest technological efflorescence, a period of creativity which, in the aggregate, remains unmatched to this day. His Harvard collea
22、gue, Joseph Schumpeter had a better fix on what was going on, although he misjudged the terrain on the road to socialism. Schumpeters homage to “creative destruction” was developed against the backdrop of what in fact has turned out to be the most technologically dynamic epoch of the twentieth centu
23、ry.,CAAGR of MFP, 1919-1948,Table 4: Compound Annual Average Growth Rates of Net Stock of Street and Highway Capital, United States, 1925-2000,1925-1929 6.00% 1929-1941 4.32% 1941-1948 0.08% 1948-1973 4.15% 1973-2000 1.63%,Street/Highway Capital as % of Private Fixed Capital,Street/Hway K Private Fi
24、xed K 1929 $16,415 $253,987 6.46% 1941 $30,861 $289,487 10.66% 1948 $47,892 $582,248 8.22% 1973 $290,389 $2,698,194 10.76% 2000 $1,423,833 $21,464,786 6.63%,Two Waves of Government Investment: the 1930 and the 1940s,Both decades experienced substantial government investment in physical capital 1930s
25、: Infrastructure 1940s: Over $10 billion of GOPO capital in manufacturing, much of it equipment, especially machine tools The 1930s investment generated spillovers that augmented PNE MFP growth, particularly in trucking and warehousing, and, through complementarities, in railroads. The 1940s investm
26、ent was associated with declining MFP growth in manufacturing and for the economy as a whole. What does this imply for the generality of the equipment hypothesis?,CAAGR of Labor Productivity,PDE PNE 1919-1929 2.36% 2.27% 1929-1941 2.48% 2.35% 1941-1948 2.17% 1.71%,United States, Private Non-Farm Eco
27、nomy CAAGR of MFP, 1919-1948,Summary Field (2003),The 1930s were characterized by productivity advance across a broad frontier of the U.S. economy. The expansion of potential output between 1919 and 1941 laid the foundation for postwar prosperity, at the same time that it enabled successful prosecut
28、ion of the war. In the light of this finding we need to rethink our understanding of the defining contours and determinants of U.S. economic growth in the twentieth century.,Field (2004),Compares 1929-41 with 1919-29 on the one hand, and 1995-2000 on the other Disaggregates, by broad sector, contrib
29、utions to MFP growth in these different periods Reassesses ITs impact on economic growth in the 1990s, as well as the broader utility of the GPT concept with which it is closely associated.,Main Arguments,MFP Growth in the 1920s was almost entirely a story about manufacturing In the 1930s, manufactu
30、rings contribution declined, although remaining high in absolute terms. Transport and Public Utilities played a much more important role. The same was true to a lesser extent of distribution. Although MFP growth between 1995 and 2000 was triple what it had been during 1973-95, it was less than half
31、what it was between 1929 and 1941. The IT revolution was responsible for about two thirds of MFP growth, and about a third of labor productivity growth between 1995 and 2000,CAAGR of MFP, PNE, United States, 1919-2000,1919-1929 2.02 1929-1941 2.31 1941-1948 1.29 1948-1973 1.90 1973-1989 .34 1989-200
32、0 .781973-1995 .38 1995-2000 1.14Sources: 1919-48: Field (2003); Kendrick (1961)1948-2000: Bureau of Labor Statistics: www.bls.gov,CAAGR of Labor Productivity, United States, 1919-2000,1919-1929 2.27 1929-1941 2.35 1941-1948 1.71 1948-1973 2.88 1973-1989 1.33 1989-2000 1.971973-1995 1.40 1995-2000 2
33、.43 Sources: 1919-48: Kendrick (1961), Table A-23.1948-2000: Bureau of Labor Statistics: www.bls.gov,Labor Productivity Growth,Roughly comparable over 1919-29, 1929-41, and 1995-2000. The 1930s were exceptional because advance took place in the virtual absence of capital deepening The 1948-73 period
34、 remains the golden age of living standard improvement, because of the combined effects of respectable MFP growth and robust rates of capital deepening.,Labor Quality - 1,How much of the growth in output per hour between 1929-41 is attributable to labor quality improvement? Margo (1991, 1993) has em
35、phasized selective retention of higher quality workers as employment drops in a recession Not relevant for peak to peak comparisons; these composition effects would have been unwound as economy returned to full employment 1941 is much closer to full employment than 1940 or 1937, but still had 9.9 pe
36、rcent unemployment, vs. less than 4 percent in 1929 In retrospect it would have been helpful for my research had Japan delayed attack on Pearl Harbor for another 8 to 12 months, so the U.S. economy could have continued its then rapid movement toward full employment before full scale war mobilization
37、 took place,Labor Quality 2,Goldin (1998) has emphasized rapid rise in high school graduation rates in the 1930s. R and D employment data for manufacturing and Margos analyses reflect strong demand for managerial, scientific, and technical personnel during this period Opportunity cost of high school
38、 attendance dropped as probability of a non high school grad being unemployed rose Build out of surface road network facilitated high school attendance Influx of human capital fleeing Hitlers Europe,Labor Quality 3,Labor Quality did rise between 1929 and 1941, but changes in labor supply were probab
39、ly not a major or dominant influence on growth in output per hour If we calculate rate of growth of output per hour between 1929-41 using adjusted hours, where adjusted hours take into account labor quality improvement, increase in output per hour is 6 percent less If selective retention were the do
40、minant influence on growth as we came out of recession, we should have had a larger increment to output per hour moving from 19.1 to 14.6 percent unemployment between 1938 and 1940 than we did moving from 14.6 to 9.9 percent unemployment 1940 to 1941. But the reverse was true.,1929-41 Manufacturing
41、MFP Calculation,Output: 3.81 percent/year Hours: 1.35 percent/year Capital: .85 percent/year MFP: 2.60 percent/year Hours and output data from Kendrick; capital input data from BEA Fixed Asset Table 4.2,1941-48 Manufacturing MFP Calculation,Output: 2.20 percent/year Hours: 2.17 percent/year Capital:
42、 4.02 percent/year MFP: -.52 percent/year Hours and output data from Kendrick; capital input data from BEA Fixed Asset Table 4.2,CAAGR of MFP, Manufacturing, United States, 1919-2000,1919-1929 5.12 1929-1941 2.60 1941-1948 - .87 1948-1973 1.52 1973-1995 .66 1995-2000 2.09Sources: Field (2004), Kendr
43、ick, (1961); Bureau of Economic Analysis Fixed Asset Table 4.2; Bureau of Labor Statistics, Series MPU300003 (B).,Manufacturing, 1919-1929,Share of PNE (1929): .333 CAAGR, MFP(1919-29): 5.12% Cont to PNE MFP Growth: 1.71% PNE MFP Growth (1919-29): 2.01% Manufacturings Share: 85%,MFP Growth, Durable
44、Manufacturing,MFP Growth, NonDurable Manufacturing,1919-29 1929-48 1949-73 1973-2000 1973-95 1995-2000,Manufacturing, 1929-1941,Share of PNE (1941): .426 CAAGR, MFP (1929-41): 2.60% Cont to PNE MFP Growth: 1.11% PNE MFP Growth (1929-41): 2.31% Manufacturings Share: 48%,Manufacturing, 1995-2000,Share
45、 of PNE (2000): .214 CAAGR, MFP(1995-00): 2.08% Cont to PNE MFP Growth: .45% PNE MFP Growth (1995-00): 1.14% Manufacturings Share: 39%,Transport and Public Utilities 1919-1929,Share of PNE (1929): .14 CAAGR, MFP(1919-29): 1.86% Cont to PNE MFP Growth: .27% PNE MFP Growth (1919-29): 2.01% Sectors Sha
46、re: 13%,Transport and Public Utilities 1929-1941,Share of PNE (1941): .123 CAAGR, MFP(1929-41): 4.67% Cont to PNE MFP Growth: .58% PNE MFP Growth (1929-41): 2.31% Sectors Share: 25%,Transport and Public Utilities 1929-1941,26 percent of sector MFP growth comes from railroads 40 percent of sector MFP
47、 growth comes from trucking & warehousing 10 percent of total private non farm MFP growth comes from trucking & warehousing,Table 7 MFP Growth, Transportation and Public Utilities, 1929-1941Share of Share Share of Subsector SubsectorNI*100 of Covered MFP Contribution1941(1948) T & PU Subsectors Grow
48、th MFP Growth 1929-41 Railroad transportation 3.27 0.372 0.420 2.91 1.22 Local and interurban passenger transit 0.66 0.075 0.084 3.02 0.25 Trucking and warehousing. 1.03 0.117 0.132 13.57 1.80 Water transportation. 0.42 0.048 0.054 1.47 0.08 Transportation by air 0.18 0.021 0.023 14.69 0.34 Pipeline
49、s, except natural gas 0.10 0.012 0.013 4.48 0.06 Transportation services 0.13 0.014 Telephone and telegraph. 1.30 0.148 0.167 2.02 0.34 Radio and television. 0.10 0.011 Electric, gas, and sanitary services. 1.61 0.183 0.104 5.55 0.578.81 TOTAL 4.67,Transport and Public Utilities 1995-2000,Share of PNE (2000): .214 CAAGR, MFP(1995-00): .6% Cont to PNE MFP Growth: .13% PNE MFP Growth (1995-00): 1.14% Sectors share of MFP Growth: 11%,
