Appendix III
An Optimal Policy Mix?
If one accepts that a government cannot introduce policies that will solve the core problem at source, what are the types of options available to a government? Keuschnigg (2003) suggests that governments can be active in four dimensions:
- Spending on basic research, which has a spillover into providing a pool of potential entrepreneurs to come up with innovative ideas.
- They can internalize the positive R&D spillovers of private start-up firms by means of a subsidy.
- They can strengthen private incentives for joint effort which "tend to be too low in a situation of double moral hazard in the relationship of venture capitalists and entrepreneurs".
- They can, at least in principle, intervene to address imbalances in bargaining power between potential entrepreneurs and venture capitalists.
Keuschnigg, with use of a sophisticated theoretical model, derives an optimal policy that he argues is able to decentralise a first best allocation in the venture capital industry. The policy mix is reasonably complicated involving a number of inter-related and inter-dependent strands. Keuschnigg's optimal policy mix involves:
- output subsidies to successfully established firms,
- revenue subsidies to entrepreneurs and venture capitalists,
- a tax on start-up investment spending,
- specific taxes or subsidies on seed investments by entrepreneurs and acquisition activity of venture capitalists, and
- government spending on basic research.
The aim of a mix of policies like this is to meet the main policy challenge of not only creating more start-ups to internalise the gains from product innovation but also to promote start-up firms of higher quality, where quality relates to the survival prospects. Keuschnigg emphasises that policy needs to be active on several fronts at the same time, as the utilisation of policy initiatives in isolation (i.e. individually not as a package) results in outcomes that either cannot be unambiguously demonstrated as being superior to the no-policy outcome or are clearly inferior.
Keuschnigg's arguments behind each of these components are as follows:
Output subsidy - producers of new products are likely to charge a price in excess of marginal costs and thereby restrict demand in order to boost profits. These profits are necessary to reward entrepreneurs and financiers for the risk, effort and other start-up costs of a new venture. An output subsidy aims at inducing producers to charge the efficient demand price (i.e. equal to marginal costs) but nevertheless allow the same profit margin (i.e. retain incentives to innovate).
Revenue subsidy - in private equilibrium, with double moral hazard and simultaneous effort choice, there is a natural underinvestment of entrepreneurial effort and managerial support since agents must bear the full private cost of effort but share the marginal return. The government can, in principle, remedy the problem by giving a revenue subsidy (negative taxes) to decentralise the conditions for socially optimal effort levels.
Tax on start-up - if the distribution of bargaining power between entrepreneurs and capitalists is even so that it induces efficient search activity then a tax on start-up is required to undo the stimulating effects of the revenue subsidy. That is, the revenue subsidy is aimed at promoting entrepreneurial and managerial effort once a relationship is established, but this subsidy will in itself potentially induce excessive start-up activity.
Taxes or subsidies on seed investment - if the distribution of bargaining power is uneven a tax on, or subsidy to, entrepreneurs is required to balance the returns between entrepreneurs and venture capitalists, depending on the relative bargaining strength of the two sides.
Basic R&D - public expenditure should be expanded until the marginal budget cost is balanced by the reduction in aggregate private research costs of entrepreneurs. Publicly funded R&D leads to a larger frequency of start-up entrepreneurship by reducing private search and research efforts of potential entrepreneurs. The argument being that publicly funded R&D not only provides a spillover in terms of supply of potentially commercial ideas, but also in increasing the supply of potential entrepreneurs.
The main logic of the proposed optimal policy is to reward success. This should strengthen incentives for joint effort and thereby raise the success rate. Output and revenue subsidies may, however, boost entry by more than is desirable. For this reason, a start-up investment tax is introduced to prevent excessive entry. In the end, Keuschnigg suggests, net costs to public revenues might not be that large.
It is not clear to what extent Keuschnigg's suggested suite of policies actually represents an optimal mix. However, a key contribution of his paper is to highlight the pitfalls associated with piecemeal policies supporting the venture capital market. Keuschnigg models the theoretical economic consequences of the following actual or potential strategies to encourage start-up entrepreneurship, venture capital finance and innovation:
- Tax advantages for venture capital start-ups
- Subsidies for start-up capital costs by means of interest subsidies and indirect tax credits
- Funding of basic R&D in applied areas of with "promising commercial applications".
- Research grants
- Output subsidies
None of these initiatives yield unambiguous welfare improvements and sometimes lead to perverse effects. For example, the subsidies for start-up capital costs by means of interest subsidies and indirect tax credits are initially likely to raise venture capital profits. But this attracts in more venture capitalists, and does not necessarily change the supply of successful ideas. This implies a reduction in the average return from venture capital investments. Thus the joint surplus falls and thereby ultimately discourages effort and lowers the average success rates. In other words the result is perverse compared with the policy aims.
Keuschnigg could also not find any unambiguous evidence for boosting public support for research or offering tax advantages based on venture capital arguments. Keuschnigg concludes that policy should be active in several fronts at the same time, but that if only a single measure was to be pursued that this should be an output subsidy. This is because the subsidy encourages joint effort between entrepreneurs and venture capitalists, encourages larger scale production and increases the variety of new firms. However, the benefits of an output subsidy are less clear over the long run and it is not clear that the short run benefits identified by Keuschnigg would be robust to an open economy analysis.
Critical Assessment
While the analysis by Keuschnigg represents a commendable addition to our understanding of the interactions between policy and the venture capital market, it provides more insights about the complexities of the issues involved and what not to do rather than providing a useable policy template. We identify six key outstanding issues:
- Keuschnigg's optimal policy still requires central authorities to have a level of information that exceeds what they will have in reality.
- Even if the government did have sufficient information about the precise nature of the distortions, it remains difficult to implement an administratively simple system of subsidies and taxes.
- The complexity of the optimal policy mix, combined with practical complications is likely to encourage governments to undertake a piecemeal mix of expedient policies that result in outcomes that could well be inferior to the private (non-government policy) equilibrium.
- Some of Keuschnigg's policy recommendations are based on double moral hazard arguments. Moral hazard is a different issue from information asymmetry. While information asymmetry is essentially about difficulties in obtaining the right match between entrepreneurs and venture capitalists, moral hazard problems begin once the entrepreneur and venture capitalist have found each other and essentially relate to the complexity in design and enforcement of contract arrangements. What's good for the entrepreneur might not be good for the venture capitalist (and vice versa) and, more importantly, it may be difficult observe and monitor the other's actions.
However this issue may not be as serious in practice as suggested by Keuschnigg. First, best practice contractual arrangements have developed over time (based on country specific social and institutional settings) that both mitigate the underlying risk and the cost of designing appropriate contracts. Second, the importance of moral hazard problems is related to attitudes to risk, with moral hazard problems becoming more prominent in the presence of risk aversion. Mas-Colell et al demonstrate that "in the principal-agent model (between owner and manager) with unobservable managerial effort and a risk-neutral manager, an optimal contract generates the same effort choice and expected utilities for the manager and owner as when effort is observable" (p482). We would contend that this is more likely to be the case in a venture capitalist-entrepreneur arrangement than in most manager-owner relationships because equity sharing arrangements makes the relationship more akin to a partnership than manager-owner. - It is not clear whether Keuschnigg's analysis adequately takes dynamic incentive effects into account. For example, part of the high price set by entrepreneurs reflects a return on risk. Provisions of subsidies as suggested by Keuschnigg might potentially sub-optimally reduce this entrepreneurial risk and so lead to sub-optimal allocations (e.g. excessive risky investment activity).
Keuschnigg's analysis is based on a closed economy - moving to an open small economy model might alter the optimal policy prescription, e.g. the arguments for output subsidies are less convincing in an open economy environment where demand can be regarded as unlimited and the consumer surplus benefits are exported.
Back to Top