Figure 5a - EU15 vs. USA tax pressure Figure 5b - EU15 vs USA marginal tax rates202530354045501990 1995 200020253035404550- quartile 2. quartile 3. quartile 4. quartile USA
Tax revenue as % of GDP Tax revenue as % of GDP010203040506070BELDEUFI NAUTITAFRANLDDKSW EESPLUXGRCPRTUKIREUSA010203040506070Single, no children Married, 2 childrenMarginal taxes for APW, 2002- quartile 2. quartile 3. quartile 4. quartile
Notes: Quartiles and weights based on population shares.
Source: OECD Revenue Statistics 1965-2002, OECD Taxing Wages 2003, Eurostat and own calculations.
There are potentially three good reasons for public spending on R&D. First the incentives to
produce economically viable research may be too weak, since firms that engage in such
activities cannot capture all the benefits through for example patents or employment contracts
that prevents researchers from walking away with the knowledge paid for by their employers.
And even when rent appropriation is feasible it may not be cost-effective for society, because
the marginal social cost of using the information (after research expenses have been incurred)
may well be small, but the benefits large. So patents, which impose a price on the use of existing
knowledge, may not always be appropriate. However, the effective assignment of property rights
to private firms and the associated costs should be carefully weighed against the problems
associated with providing sufficient incentives for innovation in public (or subsidised private)
institutions, where competition and cost-effectiveness may be much more difficult to sustain.
A second, traditional, but in general not entirely convincing, argument is a failure of financial
markets. The argument is that the innovator/firm has strong reason to believe that a project is
viable, but the bank etc, due to lack of trust and understanding of the information supplied,
refuses financing. It is inherent in such situations that the proponent of a specific project has
more knowledge about the specific project, though not necessarily the marketing potential, than
outsiders. However, it is not clear how this “asymmetry of information” can be overcome.
Third, the case for government support for R&D rests, also, on the internalization of (positive
net) R&D spillovers and other imperfections that may cause the private return on R&D to
deviate from the social return.
A fourth potential argument is that individual projects may be very risky and therefore not
undertaken by firms. However, that is a weak argument as a whole. If a firm will not fund a
specific risky project, but potentially highly profitable project, it should go to the financial
markets, to share risk. They may have problems understanding the project and giving it due
credit for its value, but that is a problem facing any government controlled financing mechanism
as well as argued above.
But both of the above arguments do suggest that government regulation in the field of financial
markets should be careful not to put in place unwarranted barriers that raises the cost of, or
reduces the amount of, capital that would otherwise flow to in particular small innovative firms
that will have to rely on external capital.
However, in order for government spending on R&D to be desirable, the total net benefits must
exceed total cost, including the distortionary cost.