The current discussion about "Novas oportunidades", a multi-million-euro public program introduced in 2006 to certify skills and improve schooling levels, highlights the dearth of evidence about its effects, especially in the labour market.
So let me present some preliminary results from my ongoing research on this issue, based on data from "Quadros de Pessoal". The sample used is based on over one million workers with between four and 12 years of schooling that are working in the same firms in both 2006 and 2009. I then focused on those who increased their schooling levels between those two years and compared their wage growth over that period with the wage growth of other workers whose schooling levels did not change.
The estimates are, of course, just a first approximation and are subject to a number of caveats, which will bias the results in different directions. For instance, some of the workers whose schooling levels increased may have obtained their higher degree from a standard school, not NO. Some of the changes in schooling may be driven by measurement error, not NO. And the wage growth of people whose schooling levels do not change may not provide a good counterfactual. However, as far as I know, there are no studies about the labour-market effects of NO and very little data available. In this context, these estimates will hopefully be of some interest.
What do the results indicate? First, that average nominal wage growth over the period for the entire sample was about 12%; and that about 5% of those workers increased their schooling levels, mostly from 6 to 9 or from 9 to 12 years of schooling.
Second, and more importantly, the wage growth of those workers that increased their schooling degrees (presumably in many cases because of NO) was, on average, about 0.6 percentage points higher than those that had the same schooling in 2009 and in 2006. Dividing by the number of additional years of schooling (i.e. on a per year of schooling basis), the effect of NO is around 0.2 percentage points. This compares with a 0.4 percentage increase in wage growth for each extra year of schooling across the full sample.
My take on these preliminary results is that the labour-market effects of NO appear to be weak. Perhaps this is not surprising - if the program is essentially about certifying skills that people already have, then presumably those skills are already rewarded by the labour market. In other words, it matters much more what people can do than the diplomas that they have. If the program is to be adjusted in the future, it would probably be good to strengthen the skills dimension to the detriment of the certification element. Releasing data for researchers would also be nice.