Summary of findings

Overview

The collated literature on the impact of research and innovation is varied in both analytical approach and the research questions it seeks to answer, driven by the fact that identifying a holistic view of cause and effect of research and innovation investment is recognised to be challenging.

This literature review focused initially on a shortlist of 32 documents, identified from an initial long-list of 156 documents. Approximately half of these documents were published journal articles, one third published papers, and the remainder reports, chapters in books, thought pieces or business insights. All documents were published between 1998 and 2023, with 70% published in the last eight years. While only 17% of these papers related specifically to the UK, it is considered 77% relate to a socio-economic context that has ‘strong’ or ‘general’ similarities to the Scottish economy.

The shortlisted papers were chosen for providing a focus on how to measure the economic and social returns from research and innovation investment, and innovative activities. Rather than taking particular interest in sector-specific effects, the review explored how investment can stimulate economic and social activity, and whether a relationship exists between research and innovation investment and social returns.

Insights from shortlisted papers

As set out within Chapter 2, a primary assessment of the 32 shortlisted papers was conducted and was used to populate a structured review template. In particular this was utilised to identify the range of inputs, outputs, outcomes and impacts identified across the papers (reported below within Section 3.2).

From the primary assessment, the papers were categorised into four classes, based on the specific level of insight provided. While some of the papers ultimately transpired to offer relatively limited insight, 18 were considered to provide valuable evidence, with a further six giving insight on specific issues. Table 3.1 shows the relative classification of papers.

Table 3.1: Classification of relative insight of papers (Source: Mott MacDonald )

Classification of insight	Number of papers	Percentage
No substantial insights	8	25%
Some limited insight on specific issues	6	19%
Key insight on specific issues	6	19%
Good insight from across the paper	12	38%
Total	32	100%

The primary review identified that nearly all papers report statistically significant evidence that investment into research and innovation is an effective use of public and private resources. According to the papers reviewed, research and innovation activities deliver economic and social impacts through a range of mechanisms including but not limited to externalities and spillovers, productivity gains, macroeconomic returns, and diffusion of knowledge between sectors (including between public and private).

While most papers reviewed report that research and innovation investment has a positive and statistically significant impact on social and/or economic metrics, the variability of the magnitude of benefits is high. The literature in general recognises this variability and discusses the reasons for this as well as the casual chain of impacts.

Topic areas of papers

The literature reviewed offered a range of insights across many topic areas. These topic areas align with the general or specific research questions the shortlisted papers look to explore. From attempting to identify the returns to investment at the macroeconomic level, to setting out the ex-ante assessment criteria for innovation in a specific sector, the shortlist offered ranging and comprehensive evidence.

Assessment Criteria

Evidence for investment

Rate of return

Returns to investment

Case for expenditure

Key indicators

Impact of innovation

Drivers of innovation

These topic areas aim to show the overarching commonalties and differences between the papers, and research questions considered. For example, it is evident that mapping input (i.e., research and innovation investment) straight to impact (e.g., macroeconomic indicators) is common in econometric analyses, with the causal chain of outputs and outcomes left open for interpretation. In a few cases sector-specific approaches (e.g., transport decarbonisation and science) offer more insight into the causal R&D and innovation chain.

Literature evaluation techniques

The literature review highlights that there are difficulties in capturing the economic and social impacts of research and innovation investment. These are partially founded in the time delay, otherwise known as lag, between the investment and realisation of impacts, as well as the widespread impact specific interventions can have (i.e., spillovers and externalities). To fully understand the insights provided by the literature, it is necessary to outline some of the techniques employed to make causal inferences about the effect of research and innovation. Common themes and techniques from the shortlisted papers are identified and discussed below.

Data

Broadly, the metric used by the literature in measuring the inputs to research and innovation is the expenditure, or investment, often as a proportion of GDP in a particular geography. The analyses generally map these data against impact metrics such as Total Factor Productivity and GDP per capita (discussed in more detail in Section 0).

The data requirement of the papers are usually panel or time series datasets to see effects over time (accounting for lags). Outputs and outcomes are not as widely considered in econometric analyses as inputs or impacts (for more information, refer to Sections 3.2.1 and 0 respectively).

Lags

In the context of research and innovation investment, lags refer to the time delay between conducting research and innovation activities and realising the benefits from the output of these activities. The literature uses lags in econometric analysis, assuming the level to which impacts are lagged behind inputs and outputs of R&D. Assuming the duration of the lag is necessary for econometric analyses as it is not certain how an innovation and its effects entirely transpire, even ex post.

Timescales and lags are discussed in more detail in Section 3.5.1.

Level of aggregation

The level of aggregation of data refers broadly to the level of analysis conducted, most commonly changing by geographic and sectoral levels. Most of the papers reviewed use macroeconomic indicators when estimating the impacts of research and innovation, meaning the analysis is economy wide. This approach offers benefits in identifying direct and indirect impacts of research and innovation investment and avoiding bias of considering only successful research and innovation projects.

The constraint on this analysis, however, is that sometimes the causal chain considers only input (research and innovation investment) and impact (e.g., GDP growth), leaving outputs and outcomes open for interpretation.

Econometric approach

The econometric approach employed by several of the papers is regression analysis, ranging from meta-regression models to multiple regression models, and simplest of all linear regression models. Several studies are meta-analyses bringing together insights from a range of sources, developing an idea of a consensus in the literature regarding the rates of return to R&D investment. A small number of studies take a more novel approach such as the following:

The evaluation techniques differ (at least marginally) across all papers reviewed as part of this study. This section intends to show some of the common themes, and key differences in quantification that have been deemed important in the development, and therefore interpretation, of econometric results.

Summary of outputs, outcomes and impacts listed across literature

Research and innovation investment is targeted to enact positive change and progression from the status quo. The causal chain through which this works is key to understanding how the initial investment can become the positive change, and what is necessary to realise it. This section outlines key findings across the literature for each aspect in this casual chain from input (R&D investment) through to impact (e.g., economic and social impact) as shown below.

Outputs

The investment into research and innovation is intended to directly deliver outputs. Outputs in this context refer to what research and innovation activities materially deliver in the specific context the investment is targeted.

Outputs, depending on their quality and relevance, can create positive change in outcomes and, therefore, impacts.

The completion of research, consequent publication of academic literature, obtaining patents, and upskilling workers are all considered to be outputs following investment into innovation. The following example details some the key outputs arising from investment in science and innovation.

In some instances, outputs can be monitored and measured, for example using academic citations as a metric to review the return to funding academic institutions and research. In other instances, outputs of research and innovation investment are hard to identify in their entirety, for example the broad development of knowledge infrastructure.

Output metrics identified by the literature are shown in Section 173.3.

Outcomes

In the context of research and innovation investment outcomes refer to the benefits or changes that occur as a result of the outputs. Outcomes are often more specific and immediate than impacts, which materialise over time through the cumulation of outcomes.

The outcomes that can be realised by outputs of research and innovation investment are also varied depending on the type of activities undertaken. Knowledge spillovers and diffusion, cost reductions, profitability and employment creation / safeguarding are all outcomes commonly associated with innovation.

The literature review validates these associations, identifying these outcomes as some of the core determinants in realising positive change and progress following research and innovation activities.

The following examples find evidence on key positive outcomes of investment into research and innovation.

Much like outputs, the ease in capturing outcomes varies due to the availability of appropriate metrics. For example, at the micro level, data regarding cost savings and profitability are relatively straightforward to identify; however in aggregation across an industry, this may be more difficult.

Outcome metrics identified by the literature are shown in Section 173.3.

Wider impacts

In the context of research and innovation investment impacts refer to the long-term effects, or value, created by the outcomes. They are often broader, and more significant, than the outcomes that support the realisation of impacts.

The wider impact areas that could result from the outcomes of research and innovation activities. The literature offers evidence to suggest that the private sector is stimulated by public sector research and development, corroborating the significance of knowledge transfer and diffusion.

Additionally, some of the impacts detailed in the literature, such as environmental outcomes and standard of living, align with the Scottish Government’s commitment to a ‘Just Transition – A Fairer, Greener Scotland’. A Just Transition refers to the following.

Regarding the analysis of the literature, the types and magnitude of impacts of research and innovation investment are varied with the scale of returns discussed in more detail in Section 3.4.

The next section identifies the metrics that the literature has identified for the capture of outputs, outcomes, and impacts.

Summary of metrics listed across the literature

The way in which the literature reports how inputs affect outputs, outcomes and impacts has been discussed; however, it is necessary to consider how, in practise, the magnitude of effects of research and innovation investment may be captured across the causal chain.

The particular research question that is under consideration will affect the usefulness of the metrics. For some types of analyses specific metrics, such as patents granted, market share changes, and FTEs created, may be the most appropriate for the identification of causal effects. In other types of analyses, more general indicators, such as GDP and Total Factor Productivity, may be more appropriate.

As a general observation, the literature often opts for more general indicators (including some of those listed above), to capture the wide-ranging impacts of research and innovation activities. Specific metrics may suffer from underestimation of impacts, through the omission of spillover effects in benefit estimation. Reported scale of returns from research and innovation is discussed in more detail within Section 3.4 below.

Scale of Returns

Core findings

Following the discussion about the way in which research and innovation investment can lead to positive change, this section outlines the magnitude of returns to research and innovation investment documented by the literature.

The scale of returns is an important consideration given economic decisions are founded in channelling resources into their most efficient and productive use.

Economic and social rates of return

Channelling economic resource towards research and innovation activities becomes more justifiable, and hence feasible, where there is a consensus of a material economic or social return to the investment.

A rate of return refers to the monetary equivalent yielded in impact from input. For example, a 30% rate of return from £100 research investment (input) would yield £130 gross economic return (£30 net) in the impact measure.

Across the literature, multiple studies report a rate of return as the key metric. Many of these are economic rates of return, but some attempt to capture a broader value in social rates of return. The majority of the literature examines returns at a national level, often using a lagged measure of GDP (e.g., GDP one or more years post the recorded value of national research investment) as a core measure. Much of the literature suggests that there tends to be a decline after initial returns on investment are realised.

As the majority of empirical studies examine the impact of research and innovation with a lag below three years, primarily constrained by data availability and confounding factors, the estimates reported will only capture the relatively high initial rate of return before material depreciation in impact occurs. Determining the rate of depreciation is extremely difficult so most studies assume a 15% depreciation rate for returns to R&D investment following Griliches (1998), whereas some other studies assumed 10% or 20% annual depreciation rate. These assumptions are reasonable given that international and UK guidelines estimating lifecycles of R&D assets in the order of 10 years before obsolescence.

Table 3.2: Rate of return to R&D estimates

Source	Type of analysis	Rate of return
S-2 Value of Research - Policy Paper by the Research, Innovation, and Science Policy Experts (Georghiou, 2015)	Meta-analysis of multiple studies	20-50% 300-800%
S-3 Rates of Return to Investment in science and innovation (Frontier Economics, 2014)	Econometric / empirical analysis	12-20% (UK private range) 16% (UK private median)
Social 14-155% Social 85% (median) Social 44% (mean)
S-4 Rate of return to investment in R&D: A report for the Department of Science, Innovation, and Technology (Frontier Economics, 2023)	Meta-analysis of multiple studies	14%
S-5 Why fund research? A guide to why EU-funded research and innovation matters (Hines, 2017)	Guide detailing the importance of research and innovation	20%
S-6 The Economic Significance of the UK Science Base (Haskel, Hughes and Bascavusoglu-Moreau, 2014)	Econometric / empirical analysis	41-82%
S-8 R&D and productivity in OECD firms and industries: A hierarchical meta-regression analysis (Uger et al., 2016)	Meta econometric / empirical analysis	14%
S-13 The rate of return to investment in R&D: The case of research 01infrastructures (Del Bo, 2016)	Meta-analysis of multiple studies	20-67%
Social 28%
S-22 Measuring the Social Return to R&D (Jones and Williams, 1998)	Theoretical framework / estimation of social value	Social 30-100%
S-24 A calculation of the social returns to innovation (Jones and Summers, 2020)	Framework for social value estimation	Social 100%

The findings of the literature review show that the estimated impact of R&D investment is reported as extremely varied overall, with significant extremes of either zero or up to 800%. However, through synthesis of the evidence, there appears to be a general consensus of a core lower bound of around 10% rate of return, and a core upper bound of 85% rate of return.

Furthermore, clustering of estimates within this range enables the identification of a central estimate for the return to research and innovation investment. This central estimate has been identified as a range of 20% to 40% social rate of return to research and innovation investment.

What is apparent from the literature is that there are a wide range of determining factors that will impact upon the potential rate of return. These factors are discussed further within Section 3.5.

Alternative analytical metrics

Beyond rates of return, several other analytical metrics have identified the impact of research and innovation investment. Those detailed within the literature worthy of note include:

• One percentage point increase in R&D expenditure as a share of GDP contributes to a 0.2 percentage point increase in annual growth rate of GDP (Rodríguez-Pose and Crescenzi, 2008).
• One percentage point increase in R&D expenditure leads to increase in GDP per capita by around 10% (Surani, Gendron and Maredia, 2017).
• Average social returns to innovation investment over $20 per $1 spent (Jones and Summers, 2020).

The literature also identifies a relatively broad consensus around an output elasticity from research and innovation investment of 0.07 to 0.08. In practice, this means that by increasing research and innovation investment / activities by 100%, a 7% to 8% rise in outputs could be expected, albeit this is likely to be subject to decay over time (as discussed further within Section 3.5.1). The literature is clear that not all research and innovation activities are successful, and this measure may be useful in managing expectations relating to research and innovation investment.

Contextual impacts

Generally, the literature and associated empirical analyses have contextual factors affecting the analyses, such as geography and sector (e.g., public / private). At in some instances these factors are explored as to their impact on realising positive effects of research and innovation, either when considering the robustness of analysis, or when determining the main research question.

For example, Surani, Gendron and Maredia (2017) propose that country specific policy can affect the rate to which there is economic return to R&D investment. The paper finds the UK as a nation has an insignificant effect on returns to public sector R&D investment, whereas Ireland has the strongest positive effect of the 22 OECD countries considered. In contrast, Aitken et al. (2021) find evidence to suggest that productivity boost from innovation is significant in Scotland.

The literature claims that the private sector is likely to underinvest in research and innovation for reasons such as credit constraints due to imperfect capital markets, risk and uncertainty, and short-termism. It is suggested that public subsidies can mitigate these frictions, spurring R&D activities (Martin and Verhoeven, 2022). Indeed, Jones and Williams (1998) suggest in exploration of the social return of R&D, that R&D in the USA should have been four times larger than the amount observed at the time of writing, supporting the notion that the private sector underinvests in R&D.

There are many factors that can affect not only the performance of research and innovation, but also the magnitude of benefits realised from innovative actives. Section 3.5 provides more detail on the key factors affecting benefits realisation.

Summary

To summarise the findings of the literature, the estimates of the rates of return to research and innovation investment vary in magnitude; however, a clustering of key estimates, and consensus in the literature, has enabled the synthesis of estimates of social rates of return on investment:

• Core lower bound: 10%
• Central estimate range: 20% to 40% (most likely return)
• Core upper bound: 85%
• Out of range estimates: Some estimates are considerably higher than the upper bound (i.e., 100% or more)

While the robustness of estimates discussed throughout this section and used to develop these boundaries of return is generally high, they may not fully capture the wider social impacts of research and innovation investment, particularly in relation to health and wellbeing. At present, the literature remains limited in capturing these benefits, which in turn means estimates largely don’t include them in the magnitude of benefits. It should be stated that this is not to say that these benefits are not realised following research and innovation investment, but rather the academic and research community is constrained in quantifying these impacts.

This section has outlined the benefits that could be expected from R&D investment; however it has also alluded to the fact the realisation of these benefits depends upon range of key factors. Section 3.5 discusses some of the key factors affecting benefits realisation from R&D investment.

Key factors to benefits realisation

As highlighted within Section 3.4, the literature identifies a wide range of factors that are likely to affect the scale of rates of return from research and innovation investment. These factors can affect either the likelihood of the success of research and innovation activities, as well as the magnitude of the positive effects that may be realised.

Key factors for the realisation of benefits

Timescales

Type of R&D

Type of Innovation output

Replication

Obsolescence

Duplication and substitution

Crowding in / out

Research Linkages

Talen

Status of Markey / Industry

Additional Innovation costs

Absorption

Supply Chains

Policy

Financial and legal institutions

This section details the way in which many of these key factors can influence the process of how research and innovation inputs lead to outcomes and impacts, as well as providing insight into some of the unintended consequences of innovation, both positive and negative.

Timescales

The realisation of the benefits associated with research and innovation investment are subject to timescales. Timescales in this context refers to any duration of time that affects the realisation of positive impacts of research and innovation. This includes lags between input and impact, duration and longevity of impact, and growth in benefits over time.

As referenced in Section 3.1.3 the benefits of research and innovation investment are often not immediately realised, with impacts lagged behind inputs. Some of the reasons for this lag include:

• There is time required to conduct research and innovation activities that the investment enables, which intends to produce output(s).
• Once research and innovation activities have been completed successfully, it can take the market time to adopt advancements, such as new technology.
• Once an innovation has been adopted and taken to market, it takes time to realise the benefits in their entirety as the market response may not be immediate, or the benefits accumulate over time.

In the estimation of benefits to research and innovation investment, the literature often must make an assumption regarding the likely duration of lags. Private research and innovation is considered to deliver commercialisation and economic return more quickly (one to three years) than public research and innovation (three plus years) given the latter may tend to be more general, without explicit commercial applications (Frontier Economics, 2014). Jones and Summers (2020) suggest that on average there is a 6.5-year delay until benefits are realised from R&D, while 10-to-20-year delays are considered lengthy lags.

As well as lags, the duration and longevity of benefits is a key consideration. It is recognised that benefits can only accrue for as long as the specific innovation remains relevant, or it becomes obsolete by a superior innovation. There is no clear evidence provided by the literature on the duration of benefits, with project specific outcomes generally not quantified. Nevertheless, it is important to consider that benefits will accrue over a finite period at least partially because of obsolescence and superior innovation, which is discussed in greater detail in Section 3.5.3.

Regarding growth of benefits, the literature generally gives reference to benefits expanding to others over time, in the form of spillovers, which is a positive impact on wider industry and society. However, there is a recognition that, in a competitive landscape, it could be expected the private sector would protect innovation gains that generate returns to the individual firm. In such circumstances the benefits to wider society and industry may not be fully realised. There is no discussion to this point in the literature reviewed, but consideration for the competitive nature of industry is important when considering the likelihood and possibility of spillovers following private research and innovation.

Types of R&D and innovation outputs

The likelihood of success of research and innovation investment and activities, as well as what can be considered a success, can vary depending on the type of research and innovation performed, whether this be applied or general.

In the performance of research and innovation, both applied and general, there is a range of output types that can be delivered by investment.

The key types of output of innovation include ideas, patents and intellectual property rights, as well as new-to-market services. Ideas enhance knowledge creating spillovers, but are insufficient alone to create tangible change, instead requiring further capital investment to implement the idea.

Patents and intellectual property rights can take time to secure (generating lags), and generally look to increase profitability by either reducing cost or increasing revenue.

New to market services can be aimed at inducing increased revenue and profitability, however, can also be via spin-offs which can require capital investment in development and rollout.

Overall, the types of research and innovation, and the types of outputs of innovation, can both be varied. This can affect the chain of events required to realise the benefits of research and innovation.

Interactions and unintended consequences of public sector intervention

With public sector intervention in research and innovation there exist interactions and unintended consequences with the private sector and, more generally, the operation of markets and research.

Positive consequences / interactions of research and innovation include replication and ‘crowding-in’; negative consequences / interactions include obsolescence, duplication and substitution, and ‘crowding-out’.

With regard to positive consequences / interactions, replication can enable widespread adoption of the outputs of research and innovation which can derive wider positive social impacts, while ‘crowding-in’ catalyses further research and innovation investment. The countering negative consequence / interaction to the latter would be ‘crowding-out’, which would see public research and innovation replace private activities; positively, the literature finds greater evidence that public research and innovation ‘crowds-in’ further private sector R&D activities.

Obsolescence refers to new products replacing or superseding old ones, which reduces the net social impact of research and innovation, and its outputs, as the benefits of previous research and innovation are curtailed.

Duplication and substitution refers to the risk that new research and innovation activities are replicating research that is already on-going, or that other research activities are curtailed because of the new activity. This is not only unproductive, but also acts as an inefficient use of public funds if the private sector has intended to perform such research and innovation.

Interactions and unintended consequences can be considered the byproducts of innovation in one sense but also a key factor affecting the realisation of benefits and their magnitude. The literature indicates that they should be a key consideration of what constitutes the efficient use of public resource to perform research and innovation.

Influence of existing research networks and available talent

The performance of research and innovation builds on an existing stock of knowledge and research networks. These research linkages can increase the productivity and likelihood of success of research and innovation activities, also helping to maximise the realised impacts.

Research linkages through networks and between institutes and industry, can bring about benefits such as increasing the intensity, synergy, and efficiency of R&D, as well as increasing the collaboration between industry and research institutes. There is evidence that more intensive research and innovation networks deliver higher cumulative benefits and spillovers, increasing the returns to research and innovation.

Additionally, linkages between research institutes and industry fosters strong collaboration to deliver larger benefits; however, it is suggested that geographic proximity of the firm to research institutes affects returns (Rodríguez-Pose and Crescenzi (2008) and Frontier Economics (2014)).

Additional to research linkages, availability of talent is identified as a key factor affecting research and innovation activities. The stock of human capital, and therefore the supply of skilled labour, affects the ability to perform research and innovation, also affecting the ability to successfully realise the benefits from it. The literature most notably identifies this as a restriction on regions with limitations in skills base.

Market / industry factors

The extent to which research and innovation is performed, and is successful, can be dependent on the type of industry performing it, and the historical success of innovation upon that research and innovation activities build upon.

The historical performance of research and innovation by an industry can determine the number of firms that are actively involved in, and can benefit from research and innovation and, therefore, the availability of capital funding. The greater the competition is for funding, the greater the quality of research and innovation, given the increasing need to evidence superior innovation to successfully access funding.

The literature finds evidence to suggest high-tech industries with a strong R&D and technology base to build on perform strongly in research and innovation activities. On the contrary, there is limited evidence to support differing scales of impact by industry, which restricts understanding in practise of the importance of access to market, and the ease of disrupting markets with new innovations.

These market and industry factors raise questions about how the absorbative capacity of markets for developing new concepts can affect the success of research and innovation, as well whether such markets support research and innovation through robust supply chains.

Entrepreneurial costs, the informal costs of innovating (e.g., product development), and capital investment, the investment required to take an innovation to market (including the cost of obtaining capital), are evidenced to reduce the returns of research and innovation.

Additionally, regarding absorbative capacity of markets, the literature finds some evidence to suggest some industries don’t have capacity to absorb new innovations, and that innovations can stall due to limited supply chain capacity.

Evidence in the literature of these additional costs and constraints on research and innovation shows that the nature and maturity (in relation to R&D) of the industry / market can be important in realising the benefits of research and innovation.

Policy and structural systems

A key factor influencing the ability for research and innovation to deliver successful outcomes can include overarching policy and structural systems of a country or region. This includes macroeconomic policy, as well as specific policy pertaining to research and innovation activities and educational policy and systems that could affect the availability of skills (as referenced in Section 3.5.4).

Relating to macroeconomic policy and innovation policy, which could both be expected to affect the ability to perform successful research and innovation and affect the operation of markets, there is no specific evidence presented on how key policy considerations affect outcomes. Surani, Gendron and Maredia (2017) suggest that country specific policy can affect the rate to which there is economic return to research and innovation investment, showing differences in the return to R&D across 22 OECD countries (as discussed in Section 3.4.1.3).

Educational policy and training systems influence, and are key determinants of, the development of talent and skills base within industries, regions and nations. As has been discussed, some regions fail to maximise the benefits of research and innovation because the amount, or quality of research and innovation, that can be performed is constrained by a limited skills base.

Financial institutions affect the ability to raise finance which can constrain or incentivise research and innovation activities, depending on whether institutions are set up to support such capital investment.

Legal institutions affect the ability to protect innovations, which in turn can affect the propensity of investment into research and innovation in industry.

Overall, policy and institutions hold weight in fostering an environment, whether this be in specific industries, regions or countries, that encourages successful research and innovation. Observed disparities at the national level between developed nations (OECD) is postulated to be founded in policy environments (Surani, Gendron and Maredia (2017)).

Research in developed or developing nations

Research can be influenced by the country in which it is performed, more specifically by whether the country is considered developed or developing. How developed a country’s economy is a reflection of a wide range of aspects, encompassing a number of factors already discussed within this section, including financial and legal structures, educational attainment, and technology base, amongst many others. The literature reviewed demonstrates that the overall maturity of a country’s economy affects the ability for benefits to materialise from research and innovation.

The status of a country’s economic development can affect the returns to research and innovation; however, this is intrinsically linked to many other factors such as policy, talent, industry, and technological base. There are differences in outcomes dependent on a country’s classification of developed or developing, with modelling indicating that the more developed a country is, the more it can benefit from R&D (Surani, Gendron and Maredia (2017)).

A question worth consideration is whether limited returns (relative to developed countries) in turn affects propensity to perform research and innovation in developing countries, which then affects the magnitude of returns that can be realised.

International spillovers

Some benefits of research and innovation may be derived within international markets that are not directly captured by the country where the public sector investment originated. This effect, as had been alluded to previously, is known spillover, which generally sees the wider adoption or application of the outputs of research and innovation.

Trade links, the stock of human capital, and network participation are all considered key in enabling international spillovers.

Trade links and participation in networks enable the diffusion of knowledge via the process of trade and collaboration, fostering stronger outcomes across existing trade links and networks.

Human capital stock is a key determinant of international spillovers, as it is necessary that the secondary nation (benefiting from spillover) can apply insights from original research and innovation.

The literature exhibits some mathematical evidence of positive returns from international spillovers, suggesting the benefits of research and innovation are realised beyond the primary nation within which the research and innovation investment is made (Frontier Economics, (2014) and Jones and Summers (2020).

Summary

Overall, this section has discussed the factors that the literature has identified as being key in realising the benefits of research and innovation investment. From a holistic perspective, the factors identified have been suggested to both positive and negative effects on the likelihood of realising benefits. This has provided balanced insight into what may help or hinder the causal chain of research and innovation from input to impact.

It should be stated that the factors affecting benefits realisation identified in this section are not exhaustive, but intend rather to show the common themes identified in the literature identified as being the most useful for this study.