THE IMPACT OF PUBLIC R&D EXPENDITURE ON BUSINESS R&DΦ
Dominique GuellecΛand Bruno van Pottelsbergheα
Submitted October 2000, Revised version November 2001
This paper attempts to quantify the aggregate net effect of government funding on business R&D in 17
OECD Member countries over the past two decades. Grants, procurement, tax incentives and direct
performance of research (in public laboratories or universities) are the major policy tools in the field. The
major results of the study are the following: Direct government funding of R&D performed by firms has a
positive effect on business financed R&D (except if the funding is targeted towards defence activities).
Tax incentives have an immediate and positive effect on business-financed R&D; Direct funding as well as
tax incentives are more effective when they are stable over time: firms do not invest in additional R&D if
they are uncertain of the durability of the government support; Direct government funding and R&D tax
incentives are substitutes: increased intensity of one reduces the effect of the other on business R&D; The
stimulating effect of government funding varies with respect to its generosity: it increases up to a certain
threshold (about 10% of business R&D) and then decreases beyond; Defence research performed in public
laboratories and universities crowds out private R&D; Civilian public research is neutral for business
JEL: E22, O31, O57.
Keywords: Technology policy, tax credit, R&D, panel data.
ΦWe thank the participants to various seminars, including the OECD Committee for Scientific and Technology
Policy and the NBER 2000 Summer Institute on Productivity for helpful comments and suggestions. All opinions
expressed in this article are those of the authors and do not reflect necessarily the views of the OECD or Université
Libre de Bruxelles.
ΛOECD, Directorate for Science, Technology and Industry, 2 rue André Pascal, 75775 Paris cedex 16; email:
αCorresponding author; Université Libre de Bruxelles, Solvay Business School, Associate Professor and Research
Associate of the CEPR project on Product Markets, Financial Markets and the Pace of Innovation in Europe; Centre
Emile Bernheim and DULBEA, 21 av. F.D. Roosevelt, CP 145-01, B-1050, Brussels; Tel/Fax : ++ 32 2 650 48 99;
OECD governments spent around USD 150 billion in research and development (R&D) activities in
1998, almost one-third of total R&D expenditure in the concerned countries. Beside fulfilling public needs
(such as defence), the economic rationale for government involvement in this area is the existence of
market failures associated with R&D. These market failures are typically twofold. First, imperfect
appropriability, or the diffusion of knowledge beyond control of the inventor, implies that the private rate
of return to R&D is lower than its social return. In addition, high risk for research implies extremely high
hurdles, discouraging firms from engaging in such activities. This is especially detrimental to small firms
for which access to funding is more difficult. For both reasons, the amount invested by firms in research
activities in a competitive framework is likely to be below the socially optimal level (Arrow, 1962).
The wedge between private and social returns is likely to be highest in basic research, requiring a
stronger involvement of government in this area. But government may also want to stimulate R&D
performed by business, either to reduce the private cost of R&D (e.g. grants) or to help firms in
understanding the technological opportunities that are available, thus reducing both the cost and
uncertainty of research. If these policies are effective, public and private funding may be complementary
and increasing the former will enhances the latter. The effectiveness of policies aimed at stimulating
private R&D outlays can be challenged on three main grounds, however.
First, government spending may crowd out private spending, by increasing the demand of R&D and
hence its price. Goolsbee (1998) and David and Hall (2000) argue that the major effect of government
funding is to raise the wage of researchers. When faced with higher research costs, firms will shift their
funding to alternative investments. This implies that, even if the total amount of R&D is higher due to
government funding, the real amount of R&D (adjusted for the higher cost of research) will be lower.
A second argument is that public money may directly displace private funding, as firms may simply
substitute public support for their own, while undertaking the same amount of research as originally
planned. In this case, the government supports R&D that would have been performed anyway, there is no
“additionnality” coming from government funding. It is also possible that a firm starting a project thanks to
government funding has the effect of deterring other firms to start a similar project although they were
previously considering to do so. In such a case there is “aggregate level non additionality”, due to pre-
emption of a research project thanks to government funds. It is a direct form of crowding out, or
displacement, which does not work through the price mechanism.
Third, governments are less likely to allocate resources efficiently than market forces, which may
generate distortions in the allocation of resources between fields of research. It may also distort
competition between firms by supporting some at the expense of others.
The purpose of this paper is to assess the effect of government spending on R&D that is funded and
performed by business. It addresses the following questions. Do public performed research, direct funding
and fiscal incentives stimulate business-funded R&D? Does the stimulation effect dominate the crowding-
out effect? How do the policy instruments interact with each other -are they complements or substitute?
The analysis covers 17 OECD countries over the period 1981-1996. It is an integrated and cross-country
approach at the macroeconomic level, covering three policy instruments, making it distinct from previous
work. The next section identifies the various channels taken by money flows: where and how public money
is spent. Section 3 presents the econometric model and the data. The empirical results are interpreted in
section 4 and section 5 draws the conclusions and policy implications.
2. Public policies to support private R&D
The effect of public spending may differ according to the policy instrument. Three main policy
instruments are typically used : public (government or university) performed research, government funding
of business-performed R&D and fiscal incentives. Public research is carried out in public laboratories or
universities, and is funded essentially by government. Examples are the national laboratories in the United
States or the CNRS (Centre National de la Recherche Scientifique) in France. A key goal of these bodies is
to satisfy public needs and to generate basic knowledge, some of which may eventually be used by firms in
their own, applied, research. Government laboratories are primarily concerned with meeting public needs,
while universities and similar institutions are more concerned with the generation of basic knowledge.
Universities typically also have a more independent research agenda than government laboratories, making
them less responsive to policy. However, the government controls much of the research budget of these
institutions (through grants, contracts or fellowships), and university research is therefore a relevant
instrument for policy makers. These two tools primarily provide indirect support to business R&D,
however. It has been argued that the kind of science produced by public research facilities is irrelevant to
the business sector (Kealey, 1996), with the idea that if it were useful, business would do it itself.
However, weak appropriability of basic knowledge makes it difficult for firms to reap its rewards: as a
contribution to this ongoing discussion, the effect of public spending in this area is tested in this document.
The second policy instrument is direct support of research performed by the business sector.
According to the Frascati Manual (OECD, 1993), two categories of government funding can be identified.
First, funding aimed at the procurement of R&D where the results of the R&D belong to a recipient that is
not necessarily the performer. Second, funding for R&D performers in the form of grants or subsidies,
where the results belong to the R&D performer. In both cases, subsidies are targeted to specific goals
chosen by the funder. The government may fund technological projects that have a potentially high social
return (e.g. “generic technologies” or “pre-competitive research”) or that are useful to the government’s
own objectives (e.g. health, defence). Such government funding supports the recipient (the technology or
the firm) even if the recipient may be initially inferior to competitors, which has led to the criticism that
governments, rather than markets, are “picking winners”. Grants often include specific conditions, e.g. the
firm may be required to establish research alliances with other firms, or to collaborate with universities.
Thirdly, government can help firms through tax breaks. Most OECD countries allow for a full write-
off of current R&D expenditures, which implies that depreciation allowances are deducted from taxable
income. Among the 17 countries included in the present study, about one-third also provided R&D tax
credits in the mid-1990’s (see Annex Table A1). These are deducted directly from the corporate income tax
and are based either on the level of R&D expenditures – flat rate – or on the increase in these expenditures
with respect to a given base – incremental rate. In addition, some countries allow for an accelerated
depreciation of investment in machinery, equipment and buildings devoted to R&D activities. Some
countries also provide special R&D tax breaks for small firms. The drawbacks of tax credits are that they
provide a windfall gain to firms, that they are unlikely to change a firm’s R&D strategy, and that they
primarily compensate for past effort.
Tax breaks do not discriminate much, implying that firms can use public money for any goal,
whatever its social rate of return. Non-discrimination may be regarded as an advantage, since it does not
distort the research agenda as shaped by market forces. However, tax incentives also have some
discriminatory features, as they are not accessible to firms that are not taxable, e.g. new firms with
investment higher than sales. Such companies may, however, be among the most innovative and may also
be the most in need of liquidity. In some countries, special provisions in the tax law allow cash refunds for
certain categories non-taxable firms.
Of the three policy instruments, only direct funding and tax breaks have thus far been the subject to
quantitative analysis. This is unfortunate, since the three policy instruments have partly similar, partly
complementary objectives. Potential interactions between these tools make it difficult to analyse the
effectiveness of one instrument independent of the others. For instance, public research, whether
performed in government labs or universities, provides basic knowledge that is especially helpful for firms
in the most advanced technology areas (close to basic research). Grants help firms in the applied research
stage and encourage co-operation, as another way to internalising externalities. R&D tax breaks, since they
are not or weakly discriminatory, help all R&D performing firms, especially those that do not have access
to grants (often small companies) or those that perform research that is not sufficiently “basic” to benefit
from other policy instruments. There are also interactions between the instruments. Those affecting applied
research, such as R&D tax credits, may enhance the efficiency of instruments oriented towards basic
research, as they may strengthen the absorptive capacity of recipient firms. The different tools thus
constitute a system, and their efficiency can be best captured by analysing the system as a whole.
3. The model and data
Previous studies attempting to evaluate the effectiveness of government support to business R&D
have focused either on the relationship between R&D subsidies and business-funded R&D, or on the effect
of fiscal incentives.
1A comparison of these studies is difficult due to the heterogeneity of the empirical
models used -e.g. regarding time periods, data sources, aggregation levels and econometric techniques. On
average, however, most studies find a positive effect of government funding and tax incentives on privately
financed R&D. This is even more clearly the case for studies at the aggregate, macroeconomic level;
among seven such studies referenced by David et al. (1999), six find that public and private R&D
expenditure are complementary, while the seventh finds no significant relationship. Nevertheless, the
existing literature has disregarded two important dimensions. There has been no attempt thus far to test the
effectiveness of all instruments simultaneously. And second, there are only few macroeconomic studies [of
the 33 studies referenced by David et al. (2000), seven are macroeconomic], most empirical analyses being
at the firm or industry level.
As compared to the firm level approach that is more common in the field, the macroeconomic
approach allows indirect effects of policies– negative as well as positive spillovers- to be captured. These
effects may be quite important. A firm benefiting from subsidies is likely to boost its own R&D activity,
but the R&D activity of competing firms might decline, for instance because the financial advantage given
to the recipient might reduce the rate of return of competing firms. Negative externalities can also occur
between industries, as shown by Nadiri and Mamuneas (1996). Conversely, the recipient firm’s research
may generate knowledge spillovers that will also be beneficial to its competitors. The potential presence of
these effects makes the case for empirical studies at an aggregate level, which implicitly take them into
account (be they positive or negative).
1. David et al. (2000) and Capron and van Pottelsberghe (1997) survey studies on the impact of R&D subsidies;
Mohnen (1997) and Hall and Van Reenen (2000) survey studies on the role of fiscal incentives. Guellec and van
Pottelsberghe (1999) measure the simultaneous effect of direct government funding to business R&D and tax
incentives on privately funded and performed R&D. This paper improves on these results by taking into account other
types of public R&D and by performing new econometric specifications.
A second advantage of working at the macroeconomic level is that overall government funding of
R&D can be considered as exogenous with respect to privately funded R&D. At the firm level, the
relevance of the assumption of exogeneity is questionable because public authorities do not provide R&D
subsidies to randomly selected companies. Or, in the words of Lichtenberg ; “Federal contracts do not
descend upon firms like manna from heaven” (Lichtenberg, 1984, p. 74). Public authorities are more
inclined to support firms that already perform R&D and that have good innovative records. This view is
supported by recent empirical evidence (Czarnitzki and Fier (2001) and Wallsten (2000)) at the firm level.
A positive and significant relationship between a firm’s R&D and the government funds it received cannot
be taken as an evidence of the efficiency of government support. This argument may also apply, though to
a lesser degree, to cross-industry studies since R&D subsidies are mainly directed towards R&D intensive
industries. At the macro level, the exogeneity assumption is more acceptable.
A problem at the macroeconomic level may be that both business and government expenditure could
be influenced by common factors, which would bias the estimated relationship. Two factors are likely to be
important. First, changes in the business cycle affect the financial constraints of government and business.
To account for this problem, this study takes GDP growth as an explanatory variable for business funded
R&D. Second, changes in the cost of R&D may affect both sectors. For instance, the price of specialised
inputs or the wages of researchers may increase when government expands its spending, leading to a
growth in business spending that is only nominal in character. This factor will be examined by accounting
for the reaction of R&D prices to demand, as estimated by Goolsbee (1998).
The different policy instruments raise specific measurement issues. Public research is broken down
into two components, government research and university research, for which standard data is available
from OECD. Government funding of business R&D is composed of procurement and grants or subsidies.
Altough the explicit goal of procurement is not to trigger a rise in business-funded R&D, such an effect is
often called upon to justify government spending (the “leverage effect”). Due to data availability
constraints, these two components of direct government funding are combined herein. Government funded
R&D performed by firms primarily consists of procurement and regular grants, although there are also
other forms of support, such as loan guarantees, conditional loans, and convertible loans (Young 1998).
Fiscal incentives may take various forms, making international comparisons problematic. The
so-called “B-index”, designed by Warda (1996), gives a synthetic view of tax generosity (Annex 1
provides a complete description of the B-index). It is a composite index computed as the present value of
before-tax income necessary to cover the initial cost of R&D investment and the corporate income tax, thus
indicating the level at which it becomes profitable to perform research activities. It is a kind of average
effective rate of taxation of R&D. The underlying methodology is highly flexible and enables various types
of tax treatment to be modelled in a comparable manner.
The empirical analysis relies on a simple R&D investment model that considers business-funded
R&D as a function of output, the policy instruments (government funding of R&D performed by business,
tax incentives, government intramural expenditure on R&D, research performed by universities), time
dummies, and country-specific fixed effects.
3Since research activities are subject to high adjustment costs,
a dynamic specification that distinguishes short-run from long-run elasticities is required.
The model allows
for this dynamic mechanism by introducing the lagged dependent variable. This type of specification is not
common in the existing literature on the effectiveness of government support to R&D.
4On a priori
grounds, however, lagged private R&D may be seen as an important determinant of present R&D
investment. Mansfield (1964, p. 32) notices that “First it takes time to hire people and build laboratories.
Second, there are often substantial costs in expanding too rapidly because it is difficult to assimilate large
percentage increases in R&D staff. (...) Third, the firm may be uncertain as to how long expenditures of
2. The B-index is similar to the marginal effective tax rate (METR) computed for eight OECD countries by Bloom et
al. (2001). However, the latter is composed of a tax component and an “economic component” which is the sum of
the firm’s discount rate (actually, the interest rate) and R&D depreciation rate, less the rate of inflation. The empirical
results of Bloom et al. show that the tax component significantly affects business-funded R&D expenditure, while the
economic component has no significant impact.
3. These should take account of stable country characteristics that may influence the private decision to invest in
R&D, especially in the long run, such as culture, tax policies, and institutional differences.
4. Only two of the 18 studies surveyed in Annex Table A2 adopt a partial adjustment mechanism for the R&D
(desired) R&D levels can be maintained. It does not want to begin projects that will soon have to be
interrupted.” The behaviour of private investors can therefore be best described in terms of a dynamic
mechanism that allows for a long-term adjustment path. The model is written as follows:
? ?? ???? ?? ??
This equation is a first-difference auto-regressive model. RP, VA, RG, B, GOV, and HE are
respectively business-funded and -performed R&D, business sector value added, government funding of
R&D implemented in business, the B-index (which reflects the fiscal generosity for R&D, see Annex 1),
government intramural R&D expenditure (i.e. public labs), and higher education R&D outlays
(i.e. university research). The 17 OECD countries are indexed by i (=1, ..., 17), and the years 1983 to 1996
by t (= 1, ..., 14). ∆ is the first (logarithmic) difference operator and τ characterises time dummies.
model, the short- and long-term effects of the exogenous variables are [β] and [β/(1-λ)], respectively. The
signs of the parameters associated with the four policy tools can be either positive or negative, depending
on whether the stimulating and spillover effects outweigh the crowding-out, substitution and displacement
The data on value added is taken from OECD (1999a). Privately funded R&D, direct R&D funding to
business, and R&D outlays by public labs and universities are taken from OECD (1999b). All the variables
except the B-index are expressed in constant USD PPP and deflated with the business sector’s GDP price
index (base year 1990). The B-index has been computed by the OECD from national sources (see Annex
5. Country dummies, which would control for the fixed effects generated by “level” variables, are not included due to
the first difference specification. In addition, in a dynamic context, adding country dummies would yield inconsistent
estimates because the lagged endogenous variable is among the right-hand side variables. Indeed, Nickell (1981) and
Keane and Runkle (1992) show that the within transformation introduces a correlation between the lagged
endogenous variable and the error term. However, had they been introduced into the regression equation, unreported
results have shown that they would have been similar. Time dummies are included to take into account technology
Before estimating the dynamic model (1) and its various extensions, it is helpful to investigate the
influence of the policy instruments on business R&D in a simpler, non-dynamic framework, in order to
show the basic relationships and their time pattern. Results reported in Table 1 show that the effect of value
added on business R&D is essentially contemporaneous, with an elasticity of about 1.20. All policy
instruments have a significant impact on business-funded R&D, although with different signs and time
patterns. Government-funded R&D has a positive and significant effect, but only with a one and two-year
lags. Fiscal incentives have both direct and lagged positive impacts (a lower B-index reflects higher tax
breaks, leading to a negative sign).
6The estimates also suggest that the effect of tax breaks is quicker than
the effect of government funding, as business spending reacts immediately to a change in taxes. This
finding also emerges from previous studies (Guellec and van Pottelsberghe, 1999; David et al., 2000). It
seems linked to the fact that tax concessions are not conditional on the type of R&D performed by the
recipient. Instead of having to launch new projects conforming to government requirements, the firm will
just spend more on on-going projects, hence accelerating their completion or improving the quality of the
outcome. In contrast, Government subsidies and contracts apply to projects that are selected by the
government or meet certain conditions imposed by the government. In many cases, the research is of a
long-term, if not basic, nature, creating new opportunities that induce firms, at a later stage, to start further
research projects with their own funds. This leverage effect of government funding will take some time
before becoming visible in the data.
*** Insert Table 1 around here ***
shocks common to all countries that are not controlled for by the exogenous variables, such as the increasing use of
6These results are similar to the ones of Bloom et al. (2001), who estimate the impact of the user cost of R&D on
total R&D investment as follows: ln (R&D) = •1* ln (user cost t) + •2* ln (user cost t-1). The authors estimate (see their
Table II, column 2) an elasticity of total R&D with respect to the user cost lagged one year equal to -0.347 (non
significant with the contemporaneous user cost). With our sample of 17 countries, for the same specification
(including country and time dummies), we obtain a significant elasticity of –0.765. If we use the same sample than
Bloom et al. (2001) (8 countries) the estimated parameter becomes –0.761. These higher estimates might be due to
the fact that (i) Bloom et al. use total R&D investment as dependent variable, whereas we use business funded R&D;
(ii) their sample ranges from 1979 to 1994 whereas our sample ranges from 1982 to 1996; (iii) the way we compute
the user cost is different.
Government research has a negative and significant impact on business funded R&D. Moreover, this
negative impact is spread over several years (although there is no contemporaneous impact). The
crowding-out effect - which is due either to an induced increase in the cost of R&D or to direct
displacement – appears to dominate the stimulating effect. Public laboratories are supposed to meet public
goals, however, not those of business; spillovers may occur but they are not instantaneous and are not the
primary goal. The zero impact of university research on business funded R&D may point to the difficulties
in transferring basic knowledge to firms.
Table 2 reports the panel data estimates of equation (1), correcting for the potential contemporaneous
correlation of the error term across countries within a three-stage least squares (3SLS) method.
Breush-Pagan test indicates that the error term of the OLS estimates is subject to significant
contemporaneous correlation across countries.
9The estimates presented in column 1 show that the short-
term (long-term) private R&D elasticities are 1.38 (1.50) for value added, 0.07 (0.08) for government
funding, -0.28 (-0.31) for tax incentives, -0.06 (-0.07) for government research and there is no impact of
10 In order to detect potential outliers, and to test the robustness of our results with
respect to the sample of countries, the model was estimated on 17 sub-samples of 16 countries. That is, all
7. It should be kept in mind, however, that a four-year lag might be too short to capture the longer-term effects of
basic research. The effects of basic research can take several decades before reaching the application stage (Adams,
1990). Moreover, it is not clear whether positive externalities should translate into increased private R&D
8. The first two stages, which are deemed to take into account the presence of the endogenous variable among the
right-hand side variables, correspond to an instrumental variable procedure. The last stage is used to correct for the
contemporaneous correlation of the residuals. Stage 1: fit ∆RPt with ∆RPt-2 and all the other exogenous variables.
Stage 2: regress equation (1) with the fit of ∆RPt-1 . Stage 3: correct for the contemporaneous correlation of the
9. This test has to be interpreted with caution. If it globally rejects the hypothesis of cross-sectional correlation for
each pair of countries, there may still be a strong correlation between some pairs of countries. In this case, the
correction for contemporaneous correlation has to be made, even if the null hypothesis is not rejected. With the
present estimates, the test always rejects the hypothesis of no contemporaneous correlation of the error terms. The
pairs of countries that are associated with the highest values of correlation between their error terms are often
characterised by a cultural and geographical proximity, or size similarity.
10. These estimated long-term effects are similar to those obtained by summing up the significant parameters in the
non-dynamic model that includes several lags (see Annex Table A1) : 1.57 for value added, 0.16 for government
funding, –0.50 for fiscal incentives, -0.17 for government intramural expenditure, and 0 for university research.
countries were removed from the sample one at a time. This test of sensitivity further support the
robustness of our results.
*** Insert Table 2 around here ***
To examine how these estimated elasticities translate in dollar terms and analyse the impact of
government policies on the amount of R&D spent by firms, it is helpful to translate the elasticities in
marginal rates of return. These are reported in Table 3. The marginal rate of return is calculated as the
product of the elasticity and the ratio of the impacted variable (business R&D) to the impacting one. If two
policy instruments have the same elasticity, the one with the largest relative size will have the lowest rate
of return. The results indicate that one dollar of direct government funding to business generates a 0.70
dollar marginal increase in business funded R&D, i.e. 1.70 dollar in total R&D. One dollar from
government leads to a 0.44 dollar marginal reduction when spent on government research, and has no
impact on business R&D when spent on university research. This reduction is less than the initial, one
dollar, government expenditure implying that total R&D (public plus business) will rise after government
has increased its spending. In other words, the crowding-out effect of the research performed in public
laboratories is only partial. Finally, assuming that the average R&D intensity in the OECD is about 2%, a
one dollar increase in value added induces an additional 0.03 dollar of private R&D.
Government spending may not only affect the amount spent on R&D by business, but also the price of
R&D: increased demand for the scarce resources used for R&D, e.g. researchers, should increase its price
up. Goolsbee (1998) estimates the long term elasticity of the R&D worker wage with respect to
government spending to be 0.09. As wages account for nearly half of R&D expenditure, we divide this
figure by two for having an estimate of the elasticity of the total cost of R&D. Subtracting this price effect
from the coefficients estimated in Table 2 leads to an elasticity of 0.035 for direct funding in the long term
(0.08 minus 0.045). This coefficient is lower than the one estimated above, but it still aknowledges a real
impact of government funding. In addition, Goolsbee’s estimate is based on data for the United States over
the years 1968-94. The share of government R&D in the US was very high in the first part of this period
(between 50 and 60% until 1980, compared with 33% in 1996). This is also substantially larger than the
share of defence in R&D spending in other OECD countries included in this panel (the average share
across OECD was 32% in 1996) and might therefore overestimate the situation in other countries. It
remains true, however, that part of the effect identified above is due to an increase in price, not in the real
amount of resources allocated to research.
Table 2 also reports a range of alternative specifications of equation 1, examining some of the features
of the basic results in more detail. A first result is reported in column 2 of Table 2, where the private R&D
elasticity of government R&D is allowed to vary across four groups of countries. The groups are based on
the average subsidisation rate of each country over the whole period (see Table A1): countries with
subsidisation rates over 19% (high), those with rates from 11-9% (medium-high), those from 4-11%
(medium-low), and those below 4% (low). The largest elasticities are found for countries belonging to the
two ‘medium’ groups, while countries with the highest and the lowest funding rates have non-significant
elasticities. This suggests that the effectiveness of government funding increases up to a particular
threshold and decreases after that. Unreported estimates with a more detailed country breakdown confirm
these lower elasticities for countries with the highest and the lowest levels of funding.
To test directly for
this inverted U-curve that seems to characterise the relationship between government support and privately
financed R&D, the estimated private R&D elasticity of government funding is combined with the rate of
direct support, in a quadratic specification :
The results of this specification, in which α1and α2are the parameters of interest, are reported in the
third column of Table 2. They suggest that the elasticity of private R&D with respect to government
support increases with the subsidisation rate up to a threshold (estimated to be around 10%), then decreases
with the subsidisation rate, and becomes negative over a threshold of about 20%.
The variation across countries of the elasticity of private R&D with respect to government funding
could simply reflect a constant marginal rate of return to R&D funding across countries. Indeed, a constant
elasticity implies that an additional dollar of private R&D for each additional public dollar spent decreases
with the rate of funding. An elasticity that varies across countries could thus translate into constant
11 As reported above, the product of the estimated elasticities (columns 1 and 2 in Table 2)
and the ratio of private R&D to government funding shows that one dollar of government money induces
an average increase of 70 cents in business-funded R&D. It varies across countries, from no significant
marginal effects amongst the countries with high and low government funding rates, to about 50 cents and
1.01 dollar for countries with “medium-high” and “medium-low” rates, respectively.
A second aspect that could affect the impact of different policy tools is their stability over time. This
is investigated by combining the direct government funding and the B-index with indicators of their
respective stability over time.
13 The two variables that reflect the stability of the schemes for each country
are GT-instability and B-instability, which are respectively the standard deviation of the funding rate (GT)
and of the B-index over the period 1983-96. For both policy tools, the estimates presented in column 4 of
Table 2 show that the more volatile a policy, the less effective it is. R&D investment typically involves a
long-term commitment and leads to considerable sunk costs. Such investment is therefore likely to be
sensitive to uncertainty, including uncertainty that arises from fiscal or government funding. Unstable
policies in the past are often taken by firms as a signal that future change is likely to take place. These
results confirm a finding from Hall (1992), that the impact of R&D tax incentives on US firms increased
11. With a constant elasticity, γ = [(∂RP/∂RG) * (RG/RP)] , the marginal effect ρ = (∂RP/∂RG) = γ * (RP/RG)
decreases as the rate of subsidisation increases.
12. Additional econometric results reported in Guellec and van Pottelsberghe (1999), were used to estimate directly
marginal effects, by replacing the first (logarithmic) difference of government R&D by the ratio of the increment of
government R&D to the level of private R&D. Results are similar to those reported here.
13. There is less of a case for the stability of government or university research affecting their impact on business-
funded R&D as their effect is more spread over time.
over time, once it was clear that the scheme would be maintained in the future. Similar evidence
concerning R&D subsidies is available at the industry level (Capron and van Pottelsberghe, 1997). They
find for the G7 countries that R&D is more likely to be stimulated in industries where government funding
is more stable.
The interaction between the various policy tools is also an important aspect, as their effects rely on
different incentive mechanisms that may conflict with each other. Are the instruments complementary or
substitutes in stimulating business-funded R&D, i.e. are they mutually reinforcing or do they partly cancel
out? Estimates reported in column 5 of Table 2 show that government funding of business R&D is
substitute to fiscal incentives. Increasing the direct funding (tax incentives) of business research reduces
the stimulating effect of tax incentives (direct government funding). The strong interaction between the
two ‘direct’ policy tools indicates that an integrated approach to R&D policy is needed; a loss of
effectiveness is to be expected when the instruments are used separately.
A final alternative specification of equation 1 investigates the impact of defence-oriented R&D policy
on business funded R&D. Defence technologies are less likely to be characterised by spillovers, as they are
often specific, with little emphasis on cost but primarily on extreme performance in extreme conditions.
Secrecy constraints may also imply that the results will only diffuse slowly to civilian applications.
Furthermore, because defence contracting is attractive- it generates high rewards at low risk - firms might
allocate resources that would otherwise have been used for civilian research. Hence, even if defence R&D
had a positive impact on business funded R&D, the effect would be expected to be lower than the effect of
a same amount of funding that would flow into projects with a civilian purpose.
The share of defence in government R&D budgets in OECD countries is around 30% on average
(OECD, 1999). There are huge differences across countries, however, with three countries having a high
share (around 60% in the United States, and 30% in France and the United Kingdom) and the rest having
14. Defence-related funding of business R&D typically crystallize into procurements : the results do not necessarily
belong to the R&D performer, or might be constrained towards government market. Lichtenberg (1987) shows that