How can providing businesses with financial support to undertake R&D increase local economic growth?

Blog co-authored by Victoria Sutherland

In Rewiring the State, the UK government set out a new model of government, committing to more decisions – including on growth – to be made locally. This briefing is part of a series aiming to help local policymakers think through the economic effects of different local growth policies and how to maximise the benefits of those policies. Each briefing introduces the policy, sets out the outcomes it affects, the mechanisms by which it does this, and issues to consider in developing local policy.

What is financial support to undertake R&D?

Research and development (R&D) is a key stage in innovation.

Innovation is the process of applying knowledge to:

• Improve, or develop new, goods or services – known as product innovation.
• Improve, or develop new, ways of producing goods or services (such as assembly lines, just-in-time manufacturing or mobile ordering) – known as process innovation.
• Change how businesses operate (such as the move from one-off purchases to subscriptions) – known as business model innovation.

Innovation is a key driver of productivity.

R&D is inherently uncertain. Businesses cannot know whether R&D will result in innovation they can use and the payoffs to successful innovation are uncertain. In addition, knowledge spillovers mean that other businesses and organisations may benefit from innovation driving a wedge between social and private returns. Finally, businesses may face financing constraints which mean they cannot undertake innovation that could be growth enhancing. In combination, these factors mean that businesses often underinvest in R&D.

One way of overcoming this underinvestment is for government to provide financial support, for example grants, loans or tax credits to businesses undertaking R&D.

What are the mechanisms by which providing businesses with financial support to undertake R&D increases local economic growth?

Financial support reduces the cost of R&D or helps remove financing constraints and aims to increase R&D expenditure and activity in supported businesses. If these intermediate outcomes do not occur, it is unlikely that productivity, employment or other longer-term outcomes will be affected. Many evaluations focus on these intermediate outcomes.

If R&D leads to product innovation and the product finds a market, it can increase business turnover. It may also increase profits, productivity, employment or wages, depending on the innovation, business, and market. For example, a transformational product that is in much demand will be more likely to affect these outcomes than a product entering a highly competitive market and offering only a minor improvement on existing products. Outcomes from product innovations can take a long time to materialise. For example, businesses may not increase employment until they generate significant and sustainable sales from a new product.

If R&D leads to process innovation, this should improve productivity by increasing outputs for a given set of inputs (or equivalently reducing the inputs required for a given output). Effects on other outcomes will vary. For example, some process innovations might reduce employment (machinery replacing workers) while others might increase it (if reduced costs result in increased sales).

If R&D leads to business model innovation, this could affect turnover, profits, productivity, employment or wages depending on the particular innovation, business, and market.

What do you need to consider?

Spillovers may limit the impact on local economy

One way in which innovation benefits the economy is through spillovers. Spillovers can happen through the movement of staff, collaboration or imitation. For example, when staff change jobs, they can implement process innovations from their previous workplace. Some interventions, such as peer mentoring, are designed to encourage such diffusion. Joint ventures play a similar role for collaboration. Patents work in both directions – encouraging knowledge disclosure while partially limiting imitation. In contrast, trademarks and secrecy try to limit imitation and reduce knowledge disclosure relative to patents.

Spillovers mean that those undertaking R&D only secure some of the benefits from investment. A key consideration for local policy is the risk that spillovers occur outside the local area. In this case, the local area bears the costs of the support but receives only some of the benefits. Local policymakers should carefully consider the likelihood and spatial pattern of spillovers from any potential investment. Spillovers are more likely in the same or connected industries or between businesses that use similar technologies. And proximity often facilitates spillovers. For this reason, supporting businesses in a tightly defined local cluster may have greater local benefits.

The overall benefits from spillovers tend to be smaller the closer the innovation is to market. At the same time, the proportion of total benefits that stay local tends to be higher (not least because a lot of the benefits go to the supported business). This is why national support will continue to be important for early-stage R&D, while local support may be better targeted at incremental, new-to-business, and close-to-market R&D. It is also why national governments tend to support a higher percentage of costs for early-stage R&D (because the spillovers may be bigger) and why local governments may be more willing than national to fund a higher percentage of late-stage R&D that is close to market (because a higher proportion of the spillovers stay local). Finally, the fact that local policy makers tend to ignore benefits to other areas is one of the reasons why a lot of R&D funding will remain at national level. Collaborative agreements with other areas may help mitigate these effects (for example, The Cambridge x Manchester Partnership).

Information asymmetries will mean building local capacity is essential

Any R&D support programme will need to determine which sectors, businesses, and projects to support and how much support to provide. This can involve decisions at the intervention design phrase (for example, on eligibility criteria) and assessment phase (choosing between applicants) and requires specialist sectoral and technical knowledge.

Unless the programme is large, it may not be economical to recruit for these skills (either as staff members or external contractors). There may also be issues with conflicts of interest, especially in specialised areas (for example, because of employment or financial links with applicants). Employing sectoral and technical expertise is easier to justify at the national than the local level, as the costs are shared over more activities.

Even with these skills, there will often be ‘information asymmetries’ (a type of market failure), with applicants having more information about themselves, their projects and their financial constraints than the funder. Reducing information asymmetries (for example, through well-designed application processes) is an important step to identifying projects to fund. There is also a risk that where a business or sector is seen as important to a local area it is prioritised for R&D support, even if they do not generate the most appropriate projects (known as ‘capture’). Clear and transparent application processes help partially address these concerns.

Understanding the local business base, including supply chains and competitors (locally, nationally and internationally) can also help overcome asymmetries. This requires investment in research and broader business support functions. Other options include adopting a narrow focus so skills only need to be developed in a small number of areas (although this does not overcome challenges in identifying which areas to focus on) or working collaboratively with other local areas (for example, MSAs could pool resources to establish a joint technical team to assess bids). Collaboration may also help address an inherent advantage of national schemes that get to see more bids each of which provides information that may be relevant to other bids.

The design of the intervention can also help reduce the impact of information asymmetries. For example, requiring business to cover a high percentage of R&D costs or using loans rather than grants could encourage businesses to bring forward projects with a greater likelihood of success. However, these projects are more likely to have happened anyway, leading to more potential deadweight than supporting more speculative projects.

It is often suggested that local areas have better information about what interventions the local economy needs and this is one of the central arguments for devolution. These arguments carry particular weight when costs and benefits fall in the same area, and information asymmetries are limited or easily addressed – local transport schemes or housing projects provide good examples. As the discussion here makes clear, neither of these conditions hold for innovation which is why local schemes will need to be designed carefully and should not simply replicate national schemes with local decision making.

Not all R&D projects will be successful

The results of R&D are uncertain. Projects may fail or generate innovations that are not implementable, cost-effective or competitive. R&D failure rates are high, particularly for product innovation where there is both technical and market uncertainty. This uncertainty is challenging for policymakers as there is an expectation that public monies are used appropriately and can demonstrate value-for-money.

Uncertainty also creates challenges for ‘revolving funds’ where, for example, repaid loans are used to fund future loans. If projects are unsuccessful, or if benefits take a long time to come through or accrue to others (as in case of spillovers), the scope for full repayment is lower, limiting ability to recycle the funding and invest in future projects. Indeed, the fact that the benefits of the project do not all flow to the innovator is one of the arguments for providing public assistance in the first place.

Policymakers should be realistic about the likelihood of success and carefully consider likely repayment rates. Similar considerations also apply to schemes that take equity stakes in supported businesses. Consideration should also be given to how programme design might increase the likelihood of success. For example, there is some evaluation evidence that grants and loans programmes that involve collaboration with a university or other research institute perform better.

Other resources

• What Works Growth has published summaries of the evaluation evidence on R&D grants, loans and subsidies, R&D tax credits, and access to finance.
• The UK government has published an assessment of the value of public R&D.
• The Royal Society’s has published a policy briefing on science and the economy as part of its Science 2040 programme. Chapter 1 provides a good overview of the ‘paths to economic impact’.
• Providing businesses with financial support to undertake R&D is only one possible intervention that might encourage innovation. This blog looks at an associated policy – the role of universities in supporting local innovation – and includes useful data on the main barriers businesses face in innovating.