Scientific facts need to prevail

At the recent Dutch parliamentary elections, there was some ado about the fact that the parties who eventually turned out to be the big winners of the election hadn’t subjected their plans to the rigorous economic modeling and analysis of the governmental Central Planning Bureau (CPB). Apparently, other countries envy us for this tradition, under the assumption that it prohibits parties from doling out ‘free beer,’ as it was called during the election campaign. The parties who won the elections were accused of just doing that, by those who didn’t win.

To some extent, the value of economic modeling may be questioned, since it doesn’t take the future value of public investment in R&D into account. The latter was highlighted by two former CPB employees in a recent article in FD (link in Dutch). To someone who has spent a lifetime in bringing deep-tech innovation to market, this appears very counter-intuitive. The example of the explosive market valuation of ASML comes to mind. The firm has certainly benefited from government funding getting funneled into the Eindhoven region and the high-tech industry as a whole, including tax breaks for innovation and expats, NWO, STW and EU subsidies and funding for Eindhoven University of Technology. Another example would be the well-known success of the US DARPA program (Arpanet and internet, GPS, stealth technology, autonomous vehicles, robotics).

There are two main reasons given by the CPB not to consider the return of R&D in their modeling. One is the outcome of numerous studies and analyses that the productivity growth in advanced countries has decreased markedly over the past decades, despite a very large increase in total investment in R&D, as measured by the number of researchers. In other words, there’s increasingly less low-hanging fruit. The other reason is that, although very rare breakthrough technologies (steam, electricity, ICT) can lead to growth spurts, they’re inherently uncertain and unpredictable and take decades before leading to significant returns.

If the model is very sensitive to the precise value of input parameters for the return on R&D investment, I can understand that it’s wiser to leave these out when they’re so inherently unknown, despite my intuition. Of course, we may then question whether input parameters for other factors (like tax breaks for women’s labor participation, as the authors of the aforementioned article point out) are better known.

Maybe the whole notion of the Dutch election winners, ie that there’s little value added by the CPB modeling, isn’t so bad after all. It may make more sense to thoroughly discuss the political vision(s) rather than getting stuck in number fetishism. Of course, the sustainability of the national budget including investment capacity over time remains the determining boundary condition.

In such a situation, government investment in R&D would strongly depend on how parliament perceives the benefits of such funding. And that’s a tricky one. As an example, it’s safe to say that there’s broad consensus on the importance of education for future welfare, but apparently, nothing is or can be done about the growing discontent among teachers about their working conditions. With regard to R&D, there isn’t even a consensus that it brings benefits to society. Scientific facts are being replaced by opinions and feelings in the public discourse, the huge benefits of R&D results are taken for granted and many of the ills of society (like COVID or global warming) are somehow linked to them.

All this means is that the science and technology community has to make its case very convincingly, which – it has to be said – it’s not very good at. Scientific facts must come to prevail over opinions and feelings again.