Abstract

In Italy two censuses were taken in 1911: the usual demographic census, that contains labor-force data, and the first industrial census, that contains employment data. The two yield aggregate figures that are very far apart. The literature directly concerned with estimating industrial employment considers the industrial-census figures essentially exhaustive. In point of fact, the industrial census was self-admittedly, grossly incomplete, and its coverage of small-scale manufacturing, and construction, is particularly poor; the extant estimates badly underestimate total industrial employment, and badly distort its allocation by sector. Far better estimates of employment are obtained from the labor-force data, allowing for sector-specific unemployment and other distortions.