Economic development and aggregate economic growth are inevitably associated with a changing sectoral composition of the economy. For the period since the onset of the industrial revolution we observe a very specific pattern of the structural change among the three main sectors of the private economy. According to this pattern the primary sector (agriculture, mining) is dominating the economy before the onset of industrialization, then the secondary sector (manufacturing, construction) begins to gain in importance and grows until the tertiary sector (services) starts taking off. This leads to a declining share of the primary sector, a rising share of the tertiary sector and a hump-shaped share of the secondary sector which can be observed for many countries since the 19th century.

In this paper we specialize the Ngai-Pissarides model of growth and structural change [American Economic Review 97 (2007), 429-443] to the case of three sectors, representing the primary, secondary and tertiary sectors. On that basis we explore the dynamic properties of the model along the transition path to the steady-state equilibrium by numerical methods. Our explorations show that the model misses several stylized facts of structural change among these sectors.

We propose several extensions of the model to bring the model more closely to the facts. The introduction of sectoral converging TFP growth rates combined with the introduction of nonhomothetic preferences yields essential improvements of the model. Now the model does not only replicate the monotonically decreasing primary sector share and the monotonically increasing tertiary sector share but most notably also the hump-shaped development of the secondary sector share. The ntroduction of human capital accumulation is a reasonable way to lower the share of the secondary and to increase the share of the tertiary sector to bring the model even closer to the empirical facts.