This paper attempts to establish the effects of transfers to household on the rate of growth of labour productivity. In particular, it focuses on the effects of social security expenditure on the rate of growth of GDP per labour units in 19 sectors and 13 OECD countries in the period 1976-2000. The main result is that transfers, both social security spending and health expenditure, have positive and significant effects on labour productivity in sectors that require low-skilled workers and pay lower wages, such as manufacturing of non-durable goods, energy supply, construction and services. These results could be due to a risk insuring mechanism: employees with low wages (on average low skilled and high labour intensive jobs are less paid than high tech ones) find in higher government spending a guarantee of safety and wellbeing, otherwise difficult to achieve with their own resources. Moreover the increased security allows them to divert resources towards higher saving and investment in education. The results are consistent with the assumption that fiscal variables affect growth by means of total factor productivity and robust to the test of a possible spurious correlation between public transfers and growth, due to openness to trade. Robustness has been tested by changing the time span and by performing the Engle and Hendry (1993) superexogeneity test, as well.