Technology-specific Production Functions
|Title||Technology-specific Production Functions|
|Publication Type||Working Paper|
|Year of Publication||2017|
|Authors||M. Battisti, F. Belloc, M. Del Gatto|
|ISBN Number||978 88 9386 047 5|
|Keywords||mixture models, production function estimation, technology adoption, tfp|
We rely on mixture models to estimate technology-specific production functions avoiding any type of ex-ante assumption on the degree of technological sharing across firms and leaving the number of available technologies unconstrained. Internationally comparable firm-level data are used, to potentially capture all possible technologies available worldwide. Differently from conventional TFP estimates, where the terms “TFP", “productivity" and “technology" are often used interchangeably, our approach enables us to isolate the contribution to labour productivity stemming from technology (i.e. between-technology TFP) from the contribution associated to idiosyncratic productivity shocks not related to technology (i.e. within-technology TFP). While we find the former to be much larger than the latter in most sectors, the relative role of these two dimensions varies considerably across firms, being often reversed. We also find that the firm-level gaps are non-linearly correlated with the international flows of technology, as measured by the OECD country-sector technology payments and receipts. In particular, we show higher incoming (outcoming) flows of technology to be associated to higher (lower) average and dispersion of the between-technology TFP gaps. This stresses the growing importance of the availability of internationally comparable data in dealing with the technological dimension of firm-level productivity.