Peruvian construction statistics and productivity changes

Summary

Peruvian statistics were examined to see if construction remained labour‐intensive during 1955–67. Data on output, employment, capital, and materials consumption were either lacking or had defects that made a direct comparison of trends impossible. But conclusions could be drawn from relative price and wage trends. If materials and capital had not been substituted for labour, average construction costs would have risen substantially more than they did. Our estimate was that by 1967 they would have been 25 per cent higher.

Evidence from a sample of firms, however, suggested that adoption of labour‐saving techniques cannot be tied simply to changes in wages compared with other costs. All eleven innovations studied, it is true, were markedly labour‐saving; but the rate of adoption did not closely follow wage changes. Adoption came most often during those less tense years when builders were neither overstrained with orders nor lacking credit and clients. One must conclude that both rising wages and innovations can limit employment expansion in relatively poor countries.

Since both agriculture and manufacturing have been unable to absorb the growing labour force of poor countries, economists have turned their attention to other sectors for supplementary employment expansion. One of these is construction. The construction sector creates not only jobs but builds capital goods with a desirable low import content. In association with carefully structured financial institutions, it may even generate savings. How much employment a given expansion of construction will provide depends on the production functions of the sector: their slopes and their potential shifts. One must find the changes in labour productivity that go with likely changes in volume, capital accumulation, trends in material supplies, and alternate technologies.

All these questions will get imprecise answers without good statistics on output, labour, capital, and materials, both unit prices and volume. If output varies in composition and in the relative quality of components, problems of weighting and aggregating arise. Because of the sector's • instability and footloose nature, data on construction remain inferior compared with other sectors even in the most advanced and statistics‐rich countries. Can one make anything of the sorts of data available in poor countries where the sector must play its most crucial role?

In this article, using rather limited data, we note with considerable alarm that steady labour intensity and corresponding employment expansion cannot be taken for granted. Where daily wages are but a fraction of hourly American wages, and where interest rates are a multiple of American rates, the more lavish use of materials and machinery compared with labour can nevertheless begin early.

Peru is the country selected for our study. During 1955–67 Peru had a relatively high national output growth rate of 5.6 per cent together with a moderate rate of inflation of about 9 per cent. Except for the devaluation years of 1958 and 1967, flourishing exports of fishmeal, copper, cotton, and sugar helped carry this rate of growth. Lima was one of the continent's fastest‐growing cities in population and building, particularly in squatter barriadas. But in the 1960s commercial conduction also grew at more than a 15 per cent compound rate, measured in square metres built, according to some estimates.1 During 1964–68, construction had high priority under President Fernando Belaunde, a professional architect. More U.S. foreign aid for housing (direct and guaranteed private loans) went to Peru in absolute terms during this period than to any other country. We shall analyse the consequences by examining (1) trends in relative costs, (2) relative output and import trends, and (3) data about receptivity to innovations in a sample of firms.

Notes

D. G. is an Economist at the International Bank for Reconstruction and Development. W. P. S. is Professor of Economics at Michigan State University. This study was partially financed by a Midwest Universities Consortium-Ford Foundation Grant, The findings do not necessarily reflect the views of the Ford Foundation, the Agency for International Development, or the International Bank for Reconstruction and Development.