Abstract:
Structural change, i.e. the reallocation of economic activity across the broad sectors agriculture, manufacturing, and services, is a stylized fact of modern economic growth. This paper first documents the patterns of structural change over more than 100 years and then tests whether standard preference specifications can fit the data. Consumption expenditure data for the US and the UK show that trends in sector shares differ across time periods in important ways. While we observe a steady decline in agriculture and manufacturing and an increase in services since the Second World War, these patterns are fundamentally different when including the first half of the century. Going back to 1900 reveals a hump shape in manufacturing and a late rise of services in both countries. The choice of the time period has important consequences for whether the traditional theory can fit the data. The standard preferences used in the structural change literature is the Generalized Stone-Geary specification that includes subsistence consumption levels. We structurally estimate this specification using the aggregate time series data for the US and the UK. While the Generalized Stone-Geary form can fit the data after the Second World War relatively well, it cannot explain the late rise in services when extending the sample back to 1900. The subsistence level in agriculture binds in the early sample years, which would imply that the representative household starves. Furthermore, the structural estimates of the Generalized Stone-Geary specification are sensitive to the sample period. We conclude that an alternative preference specification that is in addition consistent with balanced growth may be better able to explain important long-run patterns of structural change.
Photo: Vegard Sjurseike Wiborg