Policies to Stimulate Innovation and Entrepreneurship

Abstract

Most advanced industrialized countries today justify their innovation policies on the basis of the dominant knowledge-based growth paradigm—primarily investment in R&D and education, as we saw in Chap. 2. However, in terms of how growth policy should be formulated, this provides an incomplete guide which even borders on the misleading. The reason behind this misguided thinking is simplistic assumptions, both about how an economy functions and about the innovation process itself. Innovation is still seen as an exogenous force that can be stimulated through government funding targeting R&D, start-ups, and small businesses, which will then fuel higher growth and increased prosperity.

Most advanced industrialized countries today justify their innovation policies on the basis of the dominant knowledge-based growth paradigm—primarily investment in R&D and education, as we saw in Chap. 2. However, in terms of how growth policy should be formulated, this provides an incomplete guide which even borders on the misleading. The reason behind this misguided thinking is simplistic assumptions, both about how an economy functions and about the innovation process itself. Innovation is still seen as an exogenous force that can be stimulated through government funding targeting R&D, start-ups, and small businesses, which will then fuel higher growth and increased prosperity.

We claim that this line of reasoning is also subject to debate in the specific context we address in at least two respects (Frydman et al. 2019). First, given that we seek to explain the causes of economic growth, a deeper causal understanding is required. Second, the adequacy and precision of policy proposals hinge on a good causal understanding of the growth process and its microeconomic foundations. Innovation is an intricate and complex process involving a multitude of actors with complementary competencies. The willingness to handle uncertainty is key, a feature often attributed to the entrepreneur, but it also involves other important agents in an innovation bloc. Thus, to stimulate innovative processes, policies must encourage risk-taking and increase the willingness of agents to act without certainty of success. Since uncertainty is genuine, there will be failures, and these will negatively influence those who have been engaged in innovative endeavors. The policymaker can cushion such failures to some extent but, perhaps more importantly, incentivize innovation activities by, for instance, maintaining taxes at reasonable levels should the attempted venture prove successful.

Given that the group of countries doing cutting-edge innovation of great commercial value is no longer restricted to wealthy industrialized countries, this task becomes even more important for the policymaker. The former planned economies have become market economies, millions of new engineers are being herded through universities in China and India, and considerable fortunes are being made thanks to new business models and rapid technological development. Sophisticated new knowledge is being created all over the world and spread through new channels at a furious pace. Even small firms can, with the help of modern information technology, become internationalized and make use of research expertise on the other side of the globe. Still, the so-called born globals are a rare phenomenon and entail considerable risks (Braunerhjelm and Halldin 2019; Ferguson et al. 2021).

Nevertheless, increased trade (including internal trade within firms), rising investment between countries, and increasingly intensive exchanges between universities are some of the mechanisms for disseminating information and knowledge. Economies are linked via sophisticated Internet-based tools that promote cross-border networks and collaborations—as well as commercialization. The result is intensified global competition, making it increasingly difficult to obtain a high return over a longer period in the form of high relative prices for advanced products where Swedish companies have traditionally held market power. Some of these links have, however, become weaker due to the present tendencies towards deglobalization and increased geopolitical tensions.

For a small country like Sweden, the first task in this new landscape will be to build a globally competitive body of knowledge based on domestic education and research efforts but also on knowledge acquired from abroad. The second task is to create effective mechanisms that disseminate knowledge and transform it into societal benefits, i.e., innovations that lead to new and growing businesses, rising investment, higher value-added and increased employment. Maximum leverage from policy initiatives to boost innovation and entrepreneurship is achieved when such measures are coordinated across different policy areas. At the same time, we are well aware that it is not politically feasible to implement all proposed measures simultaneously; whether to implement a specific policy must be weighed against societal interests other than boosting innovation and entrepreneurship.

Innovation presupposes a long-term perspective, trust, and transparent systems. Existing institutions and regulations strongly influence the opportunities for entrepreneurs to develop and commercialize innovations. In this chapter and the subsequent one, which deals exclusively with taxation, we review a number of policies and measures that can promote innovation. We will describe in some detail the changes in tax codes, regulation, education, and other policy areas that we consider most important. In the concluding Chap. 6, we will then summarize the proposed measures in the form of an innovation policy framework, inspired by the frameworks for monetary and fiscal policy introduced after the 1990s crisis, which created long-term predictability in macroeconomic policy.

4.1 The Need for a Broad Strategy

Our approach is much broader than the one that dominates the daily political debate on innovation and innovation policy, which still largely revolves around support for R&D investment and capital for start-ups. As discussed in Chap. 3, some basic institutional conditions must be in place for other policies to be effective. Most prominent are the enforcement of the rule of law such that contracts are honored, property rights are well-defined and secure, and corruption is minimal. According to Transparency International, Sweden ranked among the top five countries with the lowest level of corruption in 2022.¹ However, the Swedish score has continually declined since 2012, indicating that corruption is on the rise. According to the ranking of the Heritage Foundation Index of Economic Freedom, Sweden is also among the top global performers, ranked at tenth place out of 163 countries in 2022. The organization states in particular that “the transparent and efficient regulatory and legal environment encourages robust entrepreneurial activity.”² Hence, from an international perspective, Sweden seems well-positioned when it comes to the most basic prerequisites for the emergence of an environment conducive to innovation. Nevertheless, these are merely necessary conditions that must be fulfilled to foster innovation. In this chapter, we will present what we deem to be the most important of the remaining and critically important building blocks of an effective innovation policy, excluding tax policy, which will be discussed in the subsequent chapter.

4.1.1 Subsidies to R&D?

“Innovation strategies”—in other words, R&D investment and seed financing—are often presented as the solution to future challenges. The OECD launched its innovation strategy in 2010 (OECD 2010), which was updated in 2015 (OECD 2015), and the so-called Innovation Union was one of the flagship initiatives in the European Union’s Europe 2020 Strategy to strengthen the EU’s global competitiveness. The most recent framework program to bolster the EU’s R&D, innovation, and competitiveness—Horizon Europe—stretches from 2021 to 2027 and has a budget of EUR 95.5 billion. It builds on three pillars—excellent science, global challenges, and industrial competitiveness—and takes a mission-oriented approach.³

These ambitions are commendable, and the bar is set high. However, the arsenal of means to achieve these goals is not described in sufficient detail. Policymakers often run into obstacles when translating general principles into practical proposals. Strategies developed by international organizations seldom become particularly innovative. Member countries’ ministries of industry all too often merely impose their own blueprint on the general measures prescribed, especially if these are kept at a general level, such as “support for R&D”—a proposal that virtually everyone will endorse. But the path to success resides in the specific and often unique competencies and conditions that are best defined on a much more disaggregated level, information which must be translated into concrete proposals covering a wide range of policy areas.

As we have already discussed, there are a number of evaluations of the effects of R&D support. Table 4.1 summarizes the advantages and disadvantages of subsidies and tax relief, respectively.

Table 4.1 Effects of R&D support

In this comparison, tax reliefs are socioeconomically preferable to direct subsidies. Most international studies also identify a positive effect, albeit the magnitude differs and crowding-out effects are present (Hall and Van Reenen 2000; Hovdan et al. 2023). In particular, recent research provides evidence that effectiveness differs sharply across firms of different sizes and between countries, depending on the specific institutional context in which businesses operate. There are also signs that firms act tactically by increasing the level of R&D just enough to achieve maximum state support. When it comes to large firms—particularly those with multinational operations—they tend to relocate R&D activity to locations with greater R&D tax relief, implying R&D tax incentives could raise R&D domestically but at the expense of R&D in other locations. Hence, tax incentives work as a beggar-thy-neighbor instrument, which casts doubt on its ability to increase R&D globally (Knoll et al. 2021). More generally, R&D tax incentives and subsidies seem more efficient for smaller firms (Galindo-Rueda et al. 2020). Further, the OECD (2020) concludes that there are vast differences in returns to companies of different sizes, with most bang-for-the-buck for the smallest firm cohorts. In addition, tax incentives seem better suited for supporting R&D projects closer to the market, while direct government funding—such as grants and R&D procurement—is more efficient when targeting basic research.

The Swedish experience is mixed. In the 1970s, there was a targeted R&D tax relief scheme that allowed firms to deduct more than 100% of their R&D spending. However, the tax relief accrued long after the spending was incurred, which contributed to the inefficiency of the system. Since the subsidy reduced tax revenues without yielding the intended results, it was repealed. Later evaluations have shown that Sweden was among the countries with the highest tax expenditure related to R&D.⁴ In 2014, a new subsidy was introduced that targeted employees working with R&D. Initially, payroll taxes on firms’ R&D staff could be reduced by ten percent of wage costs before taxes with a ceiling of SEK 230,000 per month. This design gave the subsidy a small firm angle, i.e., the relative benefit was largest for smaller firms. Effective from July 1, 2023, the maximum payroll tax reduction for a firm amounts to SEK 1.5 million per month, designed as a reduction of the payroll tax rate by almost two-thirds for R&D workers.⁵ Overall, evaluation of the subsidy finds significant positive effects on the number and share of scientists. Moreover, the treatment effect of the subsidy does not differ across firms by size or debt ratio but seems to be somewhat higher for labor-intensive firms (Stavlöt and Svensson 2022).

The European Union does not allow general state aid within the internal market.⁶ Nevertheless, a plethora of national R&D subsidies exists. The reason is that the EU’s regulatory framework is more permissible towards supporting R&D as a means of increasing productivity and growth. As stated above, this form of competition is difficult to avoid completely—virtually all countries engage in it (as do individual states in the United States). We therefore arrive somewhat reluctantly at the conclusion that limited subsidies for R&D can be an innovation policy tool. But our main message is that innovation policy should be well-balanced across different policy areas, where subsidies might be one component but requires careful evaluations before being implemented, and the strategy as a whole should be much broader.

4.1.2 Not Only Start-Ups or Small Businesses

The emergence of new and growing firms is crucial for the transformation of R&D into production and prosperity. This presupposes favorable conditions for innovative entrepreneurship. Policies to promote it have all too often been equated with measures that favor small and new businesses.⁷ Our analysis in Chaps. 2 and 3 makes clear that such an approach is too narrow; instead, the focus should be on more general measures that facilitate the dynamic transformation of as many firms as possible, and in every sector. These broad-based measures should propel the economy as a whole towards a new competitive landscape by encouraging entrepreneurial experimentation and business growth throughout the economy.

Consequently, policy should encourage renewal and growth, regardless of whether development takes place in start-ups or incumbent firms, large or small. New entrepreneurial initiatives exert competitive pressure on other firms, which requires adjustment and rationalization or, for those entities that fail to cope, closure. When the goal of policy is to facilitate the rapid growth of high-quality firms, the focus shifts from encouraging the establishment of new ones to creating a coherent framework that affects a large number of policy areas and basic institutions. Such a policy should encourage (Hölzl 2010, p. 191):

a heterogeneous pool of entrepreneurs to generate new ideas and ventures, as well as mobilize productive resources by enhancing a society’s educated workforce and capacity to generate new knowledge. … it should also provide an undistorted process of competitive selection, in which small firms face equal opportunities and the market spurs the reallocation of resources from exiting firms with low performance to fast growing firms with high performance.

The main differences between small business policy and entrepreneurship policy are summarized in Fig. 4.1. Entrepreneurship policy facilitates productive entrepreneurship activities. Whether such activities involve a high or low proportion of self-employed individuals or small firms is of minor importance. Instead, the policy focuses on more qualitative aspects, favoring entrepreneurial firms with the potential to become high impact entrepreneurship (HIE).

Fig. 4.1

Small business policy versus entrepreneurship policy

4.1.3 The Importance of High-Growth Firms

When companies are compared, it turns out—unsurprisingly—that there are large differences in terms of age, size, growth ambitions, profitability, and growth rate. The small business researcher David Birch coined the term “gazelles” 30 years ago to denote the small group of firms that create the majority of all new jobs in the economy.

In later studies, a distinction has been made between gazelles and high-growth firms (Petersen and Ahmad 2007). High-growth firms are defined as firms with an average annualized growth greater than 20% over a three-year period, where growth is measured either in terms of employment or turnover. Gazelles are a subset of high-growth firms, which are up to five years old. The definitions also include the qualification that enterprises should have at least ten employees at the start of any observation period.

In their reviews of available studies on high-growth firms, Henrekson and Johansson (2010) and Coad et al. (2014) found that:

• High-growth firms are crucial to net job growth, generating a large share of all net jobs. This is particularly pronounced during recessions, when high-growth firms continue to grow while other firms decline.

• Although small firms are overrepresented among high-growth firms, such firms come in all sizes. A small subgroup of large high-growth firms are major job creators.

• High-growth firms are younger on average.

• Young and small high-growth firms grow organically and make a larger contribution to net employment growth than do larger and older high-growth firms.

• High-growth firms are present in all industries, though they are slightly overrepresented in service industries.

Young firms that survive have higher productivity and grow more rapidly than more mature ones, suggesting an “up or out” dynamic. Firms have higher productivity and grow more rapidly than incumbents or they fail. Thus, the evidence suggests that high-growth firms, especially those that grow rapidly when young and small, are instrumental to economic growth (Haltiwanger et al. 2013; Heyman et al. 2019).⁸

Heyman et al. (2021) have examined how the Swedish economy managed during three major crises since the 1990s: the real estate/bank crisis 1991–1994, the IT-bubble 2001–2003, and the financial market crisis 2008–2009. Their overall conclusion is that the restructuring of the Swedish economy was relatively smooth, particularly during the two latter crises. Some companies shrunk, some went bankrupt, while others were started. Moreover, the reallocation of labor was quite efficient. The authors attribute this ability to cope with crises to policies that have facilitated mergers and acquisitions of firms, good opportunities for laid-off workers to acquire relevant competencies through retraining programs, and improved channeling of savings to new and young firms. Firms of different types and sizes chose different strategies to adjust to a crisis, but by and large the economy jumped back to its pre-crisis path relatively quickly. Heyman et al. argue that it is preferable to strengthen the crisis resilience of businesses through proactive policies that foster and encourage dynamics rather than providing governmental support when a crisis strikes. However, there is room for improvement regarding insolvency institutions, i.e., transfer of ownership is crucial as are the possibilities for new and young firms to exploit market opportunities and absorb dismissed labor.

One important conclusion from this research is that innovation and entrepreneurship policy should focus on general measures. Instead of focusing exclusively on start-ups, the objective is to create an environment facilitating the further development of potentially fast-growing firms irrespective of size, age, and industry. This means a policy that rewards education, knowledge transfer, competition, entrepreneurship, experimentation, and risk-taking. This is what characterized Swedish policy during the “golden” period 1870–1913 when many of Sweden’s large, still extant, multinational firms were founded—companies such as Atlas Copco, Ericsson, Electrolux (a merger of two firms founded in that period), Ericsson, SKF, Asea (now ABB), Scania, and Astra (now AstraZeneca). Moreover, firms grew rapidly and became highly internationalized. These large, innovative firms became important “anchors” and nodes for further industrial development in the form of spin-offs and as a breeding ground for other new and growing firms. However, as described by Bornefalk (2017), after World War II, the Swedish economy gradually became more regulated, interventionist, and sclerotic. As a result, no new large firms emerged and remained domiciled in Sweden after the 1960s. Instead, exceptional entrepreneurs emigrated to countries with more business-friendly institutions (notably IKEA’s Ingvar Kamprad and Tetra Pak’s Ruben Rausing). Economic dynamism stalled and growth was impaired for several decades.

The structural problems of the Swedish economy became increasingly apparent during the 1980s, and these foibles eventually triggered a market-oriented reform process lasting for roughly 25 years, from the mid-1980s until 2010. This set off a wave of entrepreneurial effort, where numerous new firms attained considerable market value, although there are no examples of such firms achieving global prominence rivalling the level attained by several businesses established before World War I. Instead, several of these firms have been acquired by foreign firms and investors and expanded from bases outside of Sweden. Thus, even if Sweden has become something of a start-up nation, it is far from what could be labelled a scale-up nation. This poses a potential threat, not only to Sweden’s future industrial base but also to its future welfare and prosperity. Large growing firms are an essential part of a competitive industrial setting based on heterogenous and complementary competencies. In the remainder of this chapter, we will discuss the most important building blocks of a broad innovation policy (except for tax policy, which we will discuss in the next chapter).

4.2 The Education System

As mentioned, our strategy has two main pillars: first, to build knowledge and ensure that it reaches a critical mass, and second, to disseminate and commercialize knowledge. A natural starting point for policy proposals is the first pillar: knowledge building. Obviously, a precondition for knowledge dissemination is that citizens are endowed with useful knowledge and have the ability to absorb knowledge held by others, i.e., absorption capacity is important (Cohen and Levinthal 1990). As we will see in the next section, the estimated effects of school quality on national economic performance are sizable.

4.2.1 Test Results and National Economic Performance

What explains economic growth and thus social welfare at the country level? Over time, it became increasingly clear that the share of GDP used for capital investment could explain only a small share of per capita growth, perhaps as little as 10–20 percent. In the 1980s, endogenous growth theory brought human capital to the fore. Human capital is the stock of habits, knowledge, and social and personality attributes (including creativity) constituting the ability to perform labor to produce economic value.

A problem arose when policymakers began to equate human capital with quantitative measures of formal schooling. Large cross-country differences in the average scores in internationally comparable tests among students with the same number of school years provide compelling evidence that, in fact, the quality of the educational system matters.⁹

Studies focusing on the effect of the average number of years of formal schooling across countries have not been able to establish any robust effect on economic growth.¹⁰ In contrast, empirical research shows a strong positive relationship between the results of internationally comparable tests and economic growth. Moreover, the estimated effect is large. Hanushek and Woessmann (2015) estimated the effect for 50 countries during the period 1960–2000, where cognitive skill is measured as the average score on all international test 1964 to 2003 (notably TIMSS and PISA) in math and science. They find that an increase of one standard deviation in the average test results in mathematics and science, i.e., 100 points, “is associated with approximately two percentage points higher annual growth in GDP per capita.”¹¹ Thus, the fact that the quality of the educational system is what matters rather than the duration of schooling per se is crucial for economic growth. More precisely, the purpose of a high-quality school system is not simply to impart substantial knowledge to students but also to instill noncognitive skills such as self-discipline, perseverance, reliability, and emotional maturity. International tests also capture such skills, suggesting a double-dividend effect of a high-quality school system: A school system that imparts a high quantity of valuable knowledge to students cannot succeed in doing so without at the same time teaching them noncognitive skills, such as the ability to focus, diligence, and perseverance.

Heller Sahlgren and Jordahl (2023) updated Hanushek and Woessmann’s studies, adding results from the PISA and TIMSS tests in mathematics and science through 2015 to explain average GDP growth per capita in 50 countries for the period 1960–2016. Thus, this analysis includes the years after the financial crisis of 2007–2008. Their results are presented in Fig. 4.2, which shows a strong positive relationship between the test results and economic growth when average years of schooling and initial GDP per capita are controlled for. The estimated effect is substantial: average test results at one standard deviation higher—corresponding to 100 PISA or TIMSS points—are associated with an increase in the GDP growth rate of 1.3 percentage points. In contrast, the corresponding calculation between number of years of schooling (controlling for the effect of initial GDP per capita and test results) shows no effect, which concurs with earlier results.

Fig. 4.2

The relationship between international test results and growth in GDP per capita, 1960–2016. Note: Added Variable Plot showing the relationship between average growth in GDP per capita and average test score after having removed the estimated effect of initial GDP per capita and average years of schooling. The values on the x- and y-axes thus indicate the difference between the actual results and what is projected by the two control variables. Source: Own figure based on data from Heller Sahlgren and Jordahl (2023)

Heller Sahlgren and Jordahl (2023) also find that the share of high-performing students is substantively more important than the share achieving basic skills. While an increase of ten percentage points in the share of students who achieve basic skills is associated with an increase in the GDP growth rate of 0.18 percentage points, an equal increase in the share of high-performing students is associated with an increase of 0.87 percentage points.¹²

There are several reasons why the share of high-performing students is especially important to economic development and social welfare. High scorers are more likely to support growth-friendly policies and institutions and to hold politicians accountable for abuse and malfeasance.¹³ A highly educated population is generally more likely to resolve disputes through negotiations and informed democratic decision-making than through violent conflict and coercion.¹⁴ Other reasons why high scorers are particularly important are that they tend to save more, be more cooperative, be more innovative and more successful at using highly productive team-based technologies, and be more prone to imitate and adopt productive behaviors and solutions used by others.¹⁵

4.2.2 The Quality of Sweden’s Primary and Secondary Education

The quality of the Swedish education system is no longer as high as it once was (especially relative to other countries) in terms of building knowledge among its children and adolescents. Research consistently points to growing problems in Swedish schools—not everywhere, but in too many places to ignore: low quality, low reading comprehension, weak mathematical knowledge, and all too often, poor teachers. Every fourth student fails to achieve passing grades, both in primary school and high school. Among those who did not manage to receive credit in at least one of the core subjects—Swedish, English, and mathematics—failure in mathematics was most common.

In international comparisons such as PISA and TIMSS, Sweden has fallen far behind the leaders. Its weakest students lag the furthest behind, students’ socio-economic background has a strong impact on results, and extremely few students reach the advanced level in mathematics, while 35–50% of students do so in the leading countries. Swedish students also performed below average in the PISA creativity test, which was conducted in 2012. Sweden ranked 20th out of 28 participating countries.¹⁶ Moreover, low status for the teaching profession means that schools do not attract the quality of teachers that they need.¹⁷

If we apply Heller Sahlgren and Jordahl’s estimated effects of test results on growth, we can see that the effect for Sweden is sizable. Based on the average decline of 34.5 points (or 0.345 standard deviation) in mathematics and science in TIMSS from 1995 to 2015, the average growth in GDP per capita can be expected to decline by 0.45 percentage points (0.345 × 1.3). By using the estimated effect of the share of students attaining the advanced level, we can infer that if the Swedish share had been at the same level as Singapore, it would be associated with an increase in the annual growth rate of GDP per capita of an impressive 1.3–1.4 percentage points.

Weak basic education spills over directly into universities. The throughput time is long, and a large share of university students drop out before graduation, approximately 30% on average but with considerable variations across disciplines (among engineering students, the dropout share is about 50%).¹⁸ Interest in science is weak (partly due to poor mathematical skills at earlier stages), and the number of graduating engineers is therefore low.

Poor education also complicates recruitment to the labor market. In tandem with increasing dropout rates and lower quality of student skills, the mismatch on the Swedish labor market has accelerated. As shown by Eklund and Larsson (2023), the ratio between vacancies and unemployment has steadily increased since the 1980s. At present, Swedish firms fail to recruit the right person in three out of ten recruitment processes despite a high level of unemployment. More generally, seven out of ten firms report difficulties in finding the competence they need, in particular when it comes to people with advanced vocational training (Svenskt Näringsliv 2022). Companies say that lack of skills is by far the most important reason why they fail to recruit, and they have a particularly difficult time finding engineers. It is noteworthy that young people, who ought to have the most up-to-date education, have particular difficulty finding jobs. In international comparisons, Sweden has a high level of youth unemployment. As of October 2022, Sweden had the fourth highest unemployment rate among the young (15–24-years-old), only surpassed by Greece, Italy, and Spain.¹⁹

Our conclusion is that the quality of both secondary and higher education needs to increase. Improvement at the basic level presupposes a move away from the current postmodern social constructivist view of knowledge, which determines both how subjects are taught and the role of teachers. Swedish schools need to return to an updated classical view of knowledge characterized by teacher-led instruction, a well-structured teaching environment with clear objectives and systematic progression based on concrete course syllabi and curricula. Experience from other countries shows that such a reform would lead to rapid and sharp improvements.²⁰ Furthermore, apprenticeships to counteract the exclusion of students who are not motivated to enter academic studies and better collaboration between schools and businesses—and above all a real improvement in the education and status of teachers—deserve considerably more attention in the case of Sweden.

4.2.3 The Role of the University Sector

Swedish R&D spending is still high in relation to GDP—3.4% in 2021. Between 2001 and 2015, the trend showed a decline, but it began to rise again in 2016. Almost 72% of R&D investment is private. Over the last 20 years, Sweden has sustained its position as a leading nation in terms of R&D spending compared both to other similar countries and to leading research nations (Vetenskapsrådet 2021). However, government R&D spending is the lowest in the Nordic region at 0.9% of GDP. Moreover, with regard to government-funded research at universities, Sweden has produced several internationally successful graduate programs and research centers, mainly in medicine and technology. But in recent years, these have dropped in international rankings. Sweden no longer has a research institute or university that can be said to be among the world leaders, with the possible exception of the Karolinska Institute in the medical field.²¹

The United States maintains its leading position, with huge investments in terms of both finance and personnel, but quite a number of East Asian countries, notably but not only China, have progressed rapidly and can now be found high on the list of world rankings of universities. In recent years, China’s R&D spending has exploded, attaining a share of 22% of global R&D spending in 2019, second only to the U.S. share of 28%. Since the year 2000, China’s R&D expenditures have increased by almost 1500%, which compares with 250 and 175% in the United States and the European Union, respectively. In absolute numbers, China has now surpassed the European Union although it still lags the United States (NSF 2022). Together with India, China has a greater focus than the traditional research countries on applied and needs-motivated R&D in technology, AI/digitization, and materials science—which is a likely reason for their rapidly increasing competitiveness in fields relevant to those sectors that are exposed to international competition.²²

With such large and fast-growing competitors, a small country like Sweden must invest available resources wisely. But Sweden faces the problem that its resources are too widely dispersed. Instead of concentrating on large research universities where a critical mass can be achieved, government research funding is dispensed across too many academic institutions, many of which are, in effect, regional university colleges. A system based to a greater extent on quality criteria for the distribution of governmental research funding would—together with the opportunity for specialization that follows from greater autonomy—create critical mass in research and innovation.

The organization of university research in Sweden also seems to be more rigid than in the United States, for example. In Sweden, it is still quite common for researchers who have earned their PhD to continue at the same university when they proceed to the postdoc level. In the world’s leading research nation, the United States, this is nowhere near as common. In principle, the top universities always demand that their PhDs leave to pursue their career as assistant professors at another university. It is believed that this stimulates creativity. Rothstein (2009) asserts that a different practice in Sweden is connected with the fact that Swedish researchers tend to continue to work within the same tradition and even at the same institution as their supervisors, while the tradition in the United States is rather that younger researchers challenge their senior colleagues.²³

The leading U.S. universities are also characterized by intense and close contacts—including the creation of research-based spin-offs—with the regional business community, as well as greater opportunities for immigrant researchers and entrepreneurs. More than half the start-ups in Silicon Valley reportedly have at least one immigrant founder.²⁴ Even more strikingly, Andersson (2022) reports that of the 582 “unicorn” start-ups valued at USD 1 billion or more in the USA, 55% had at least one immigrant founder. That number rose to two-thirds when counting companies that were founded or co-founded by immigrants or the children of immigrants.

From our point of view, while both knowledge building and knowledge transfer are important, collaboration with the business community is central. This can take many different forms, such as spin-offs, research villages, and adjunct professorships. A general trend, likely to also characterize Swedish companies, is that less research is conducted jointly or in consultation with private industry.²⁵ Within the so-called knowledge triangle—research, education, and innovation—greater emphasis should be placed on innovation. In 2009, a major reform was implemented with the aim of increasing the autonomy of universities, thus reducing government regulation regarding recruitments and senior appointments, and giving individual universities more room to specialize their research and curricula.

However, this increased autonomy was not extended to their innovation systems. In our opinion, this is a shortcoming; universities should own and develop their own innovation processes. We also believe that they should adopt the intellectual property institutions they believe serve them best, i.e., where faculty own their ideas (professors’ privilege) or the university system does (Bayh–Dole).²⁶ Rules that prevent universities from developing holding companies should be altered. Even though they are referred to as the universities’ holding companies, they are fully government-owned and their activities are constrained in several ways: Any additional capital injections must be decided by the government, dividends to the universities are not allowed, and contractual agreements involving the universities are complicated (Braunerhjelm 2021). Another area of concern is the low mobility between different academic environments and between academia and business. Here there is obvious room for improvement.

Furthermore, we suggest that research resources should be allocated according to quality and innovation criteria, and that the time horizon in research policy should be extended. Today, the latter usually lasts four to five years. But research, not to mention its commercialization, often takes much longer. A declaration of intent concerning research investment along a longer perspective—10 to 12 years—would create greater predictability and confidence.

4.3 Private R&D

R&D also takes place on a large scale within private firms in terms of financial outlay, investment there is more than twice as high compared to universities. Firms also face much greater pressure to commercialize their results than do universities. Sweden has proud traditions, not least in medical-technical R&D. The number of granted patents from Swedish applicants increased between 2005 and 2018 even though this rate has decreased in recent years. Throughout the period, Swedish patents increased by 28% while the corresponding figure for the world was 29.5%. Sweden matches the world average and exceeds both the OECD average (13%) and the United States (19%). China exhibits an outstanding development with an increase of 589% between 2005 and 2018, although its starting level was low, and it is rapidly approaching the U.S. level.²⁷

A large part of Swedish R&D investment takes place within a few large company groups, which often have R&D centers outside Sweden, and where different divisions compete for internal investment. In some cases, research centers have been relocated to other countries, such as AstraZeneca’s large research centers previously in Lund and Södertälje. There are also a few research nodes that have been established in Sweden recently, e.g., Volvo Car’s R&D establishment in Gothenburg (following Chinese Geely’s acquisition of Volvo’s automobile division), and Northvolt’s establishment in Västerås.

There are several explanations for this reshuffling of R&D activities. One is that Swedish firms have not succeeded in asserting themselves sufficiently in international network-based research, where firms must increasingly acquire ideas, inventions, and material assets from other firms or foreign universities. This development is taking place increasingly on the borderline between products, services, and business models.²⁸ However, the most commonly invoked economic reason is the lack of adequately skilled domestic researchers (here we return to the issue of inadequate education) and the combination of low salaries and high income taxes. Low net salaries make it difficult to attract leading researchers from abroad. If foreign researchers are to find Sweden attractive, their pre-tax salaries need to be set very high. New R&D establishments are primarily linked to the so-called re-industrialization of the north of Sweden and related to mining, steel, and battery production. Several of these investments face considerable challenges, not least in terms of energy supply, and the future outcome remains uncertain.

We have already expressed our skepticism about focusing excessively on increased government R&D subsidies. This skepticism is reinforced by the problems described above. We believe that the system itself should be reformed. One of the most important measures concern taxes, and in the next chapter, we present a number of proposals in this area that have direct and indirect effects on business R&D.

4.4 Channeling of Savings

In the welfare state, the household partially relinquishes the need to save—the state and the social insurance sector save for it. As a result, household saving dropped sharply in tandem with the growth of government transfer systems that removed most of the need for precautionary saving and saving for retirement. As shown in Fig. 4.3, household saving even turned strongly negative in the 1980s. Therefore, a typical individual rarely has any major savings available to start a firm or join one as a (part) owner.

Fig. 4.3

Household saving as a share of disposable income, 1970–2022 (%). Source: Statistics Sweden

The deep crisis in the early 1990s, the gradually reduced replacement rates of public transfer systems and the growth of (largely mandatory) supplementary pension schemes has led to a sharp rise in household saving, exceeding ten percent of disposable income by a wide margin in the most recent decade. However, it is also clear from the graph that saving is dominated by pension insurance saving; collectively agreed supplementary pension schemes cover virtually all tenured employees with payments into those systems amounting to approximately ten percent of taxable wage income.²⁹ In 2022, the total assets of private life and pension insurance funds amounted almost exactly to 100% of Swedish GDP and almost 200% of household disposable income.³⁰

As a long-term solution, the best way to ensure the financing of entrepreneurial firms is likely to be the pursuit of policies that encourage private wealth accumulation in forms that do not preclude the assets being used as equity in entrepreneurial ventures.³¹ At the same time, for the foreseeable future, the combination of welfare-state arrangements and the extensive contractual pension insurance saving will stifle direct individual saving. Since such a large share of this money goes into pension funds, there is a growing need for at least part of these assets to be invested in entrepreneurial firms and not almost entirely in real estate, public stocks, and high-rated bonds. Although Swedish pension funds are allowed to invest in venture capital and buyout funds, these investments are still very small. The largest pension fund, Alecta, had total assets of SEK 1155 billion in 2022. Investments in private equity assets constituted a mere 11.1 billion or one percent of total assets.³²

Thus, the assets of many savers are placed in funds which are managed in such a way that they cannot be used as risk capital in the saver’s own or related firms. This institutionalization of saving, combined with the fact that tax legislation until recently, as we have seen above, made it very difficult to develop an efficient VC industry, has further contributed to complicating the supply of capital for entrepreneurial firms. However, beginning in the 2010s, numerous previous entrepreneurs who had achieved successful exits assumed new roles as business angels and investors, which has boosted access to early-stage funding.

With one stroke, rule changes that enable a “deinstitutionalization” of pension savings would make a large share of pension assets available for equity investment in entrepreneurial firms. Of course, limits would need to be set, both in terms of the amount and to prevent tax arbitrage. Here it can be noted that it was such a rule change that made a small part of pension savings available for start-up financing that formed the basis of the North American VC industry (Henrekson et al. 2021).

Changing the foundations of the entire pension system and institutionalized saving is not feasible. But some changes can and, in our view, should be made. We propose that a certain proportion of pension savings be made available for investment in unlisted firms. At present, only one out of the six public pensions funds can invest in unlisted companies (AP6),³³ targeting the segments of buyouts, venture/growth, and secondaries (when investors in a fund would like to divest their holdings before the fund is closed). They do not invest in early stages. The rules that govern which financial instruments may be included in pension savings, in individual investment savings accounts (ISK) introduced in 2012, and in capital pension savings accounts (now only listed securities and funds) should thus be made less restrictive so that some portion of the assets can be invested in unlisted (i.e., private) firms.

4.5 The Functioning of the Labor Market

In a functioning market economy, a massive restructuring of jobs and employees is constantly taking place. Firms hire and fire people, young people enter and retirees leave, people enter and exit because of parental leave, illness, further education, and more. The gross flows in the labor market are many times greater than the unemployment figures suggest.³⁴ In the midst of these flows, fast-growing firms have a dire need for flexibility in terms of contract and space to vary the size and composition of their workforce. Technological breakthroughs and changing markets mean that businesses are continuously faced with demands for adaptation and alterations in work organization and relative wages.

Acemoglu (2002) states that the profitability of a firm strongly affects the connection between new technology and wages. During early industrialization, companies were able to achieve large increases in productivity as well as higher profitability by replacing advanced craftsmanship with machines operated by unskilled labor. The past century, on the other hand, has largely been characterized by the fact that technological development, a well-educated workforce, skills, and professionalism have been complements rather than substitutes. In economic terms, this is usually called skill-biased technical change (Berman et al. 1998).

The main explanation for the increased spread of wages in recent decades is that technical change has increased the need for skills, especially those not learned through formal education and work experience measured in years of service. Highly valued skills include ability to cooperate, conscientiousness, ability to cope with new tasks and perhaps above all the individual’s general capacity to work diligently (Juhn et al. 1993). At the same time, new technology has made it possible to rationalize many labor-intensive production processes while simultaneously paving the way for the gig economy. These divergent patterns also show up in a widening income distribution over the last decades, squeezing middle income earners. Hence, technical change not only affects wage formation, but also the entire organization of the labor market in a way that includes not only institutions and policy but also the organization of production and firms.

Key labor market institutions affect how well the skill structure functions. Regulations that restrict freedom of contract limit the opportunities to find the most efficient mix of the factors of production. Three areas are particularly important for fast-growing firms: labor-market regulations, wage formation, and social insurance.

4.5.1 Labor Market Regulations

The design of labor market regulations varies considerably between different countries (OECD 1994, 2004; Skedinger 2010). Research on structural change and gross flows of firms, jobs, and workers gives us reason to believe that strict employment protection and other regulations that reduce freedom of contract are more troublesome for firms that want to grow quickly than for mature firms and those without growth ambitions. As the employer acquires a clearer picture of the individual employee’s skills—which develop over time—the appropriate tasks for that employee are constantly changing. The opportunities to find new tasks within the firm are usually better in a larger firm, because the large firm has more positions to choose from. In an unregulated labor market, a continuous matching of individuals to optimal work tasks means that individuals change employers. Such changes thus become more common for people who work in smaller, often younger, firms, and their staff turnover tends to be high.

Higher mobility in the labor market turns out to be positively related to increased productivity and the ability to pay higher pay wages. As shown by Kaiser et al. (2015) and Braunerhjelm et al. (2020), labor-market mobility is associated with higher rates of innovation. The explanations are more efficient matching, network effects and efficient knowledge diffusion. Labor is endowed with different types of know-how, abilities, and skills, and when these knowledge bearers enter an environment where their knowledge can best be combined with the knowledge of others, innovation is more likely. Research has also shown that labor market regulations tend to have different effects on small and large firms. Smaller ones have a greater need to vary and change their knowledge base—sometimes in rapid, large steps—while larger ones can benefit from a more rigid labor market where innovative staff can be more easily retained (Braunerhjelm 2011).

Strict regulation of the conditions for employment and redundancy therefore makes it difficult for entrepreneurs to adapt their workforce to fluctuations in demand. This increases the risk for fast-growing firms (Audretsch et al. 2002). In general, the proportion of jobs that are created and disappear decreases as the firm grows larger, older and more capital-intensive. Strict regulation is therefore relatively favorable towards mature firms, but creates disadvantages for young, fast-growing firms. This is illustrated in Fig. 4.4, which shows the relationship between the degree of labor market regulation and the degree of what in the literature is called high-growth expectation early-stage entrepreneurship—the type of entrepreneurial activity associated with fast-growing gazelles. The figure shows a clear negative connection between stricter employment protection and such entrepreneurial activity.

Fig. 4.4

The strictness of employment protection, 2013, and high-growth-expectation early-stage entrepreneurial activity, 2015–18. Note: High-growth-expectation early-stage entrepreneurial activity is averaged over the four years 2015–18. Permanent employment protection legislation (EPL) is given a weight of 2/3 and temporary EPL a weight of 1/3. Source: Global Entrepreneurship Monitor and OECD/IAB Employment Protection Database, 2013 update

The relative value of having permanent employment also decreases if employment protection is weak, which reduces the opportunity cost of being self-employed) (van Stel et al. 2007). Strong employment protection increases the opportunity cost both of changing jobs and becoming an entrepreneur. This reduces the tendency to attempt to start a fast-growing firm and makes it more difficult for such firms to recruit good employees.

Admittedly, firms can increase flexibility by taking advantage of temporary employment. However, there are clear disadvantages to this. Fixed-term employees are less motivated to invest in firm-specific knowledge than permanent employees, which makes it more difficult to attract workers who have or are prepared to develop valuable skills. The greatest obstacle caused by stringent labor market regulations is probably that it makes it more difficult for the individual to improve, advance, and take on new challenges. Temporary employment and staffing companies can to some extent remedy this shortcoming, but rarely for the recruitment needs of the type of highly skilled entrepreneurs that are central to a dynamic economy.

The Swedish labor market has undergone significant regulatory alterations to foster greater flexibility. The primary aspects of these changes, which became effective in 2022, and their potential implications for the labor market’s responsiveness to shifting economic conditions are as follows (Iseskog 2022):

• Emphasis on Permanent Full-Time Employment: The modified regulations stipulate that permanent full-time employment should be considered the norm. Employers’ ability to hire workers on temporary contracts has been reduced from 24 to 12 months. After 12 months, the employee is entitled to a permanent contract. In addition, there are provisions for permanent contracts under certain conditions, notably that temporary employees who have been hired for more than 12 months over a five-year period are automatically granted a permanent employment contract.

• Modified Criteria for Employee Dismissal: The criteria for terminating employees has shifted from “objective foundation” to “objective reason.” This subtle alteration is intended to simplify the process of dismissing employees for misconduct, while maintaining the burden of proof on employers and the requirement to find a suitable replacement before terminating employment.

• Expansion of Employee Exemption Criteria: The regulations have broadened the scope for exempting employees from seniority-based termination orders, permitting greater flexibility for employers. Previously two employees could be exempted provided that the firm did not have more than ten employees, which has been extended to three employees irrespective of firm size. However, this change may not significantly differ from prior negotiated agreements and associated economic compensation.

• Adjustments to Exemption Limits Based on Firm Characteristics: The updated regulations allow for variations in exemption limits based on the number of establishments a firm has in one location and the collective agreements with different unions. These adjustments offer employers the possibility to exempt up to 15% of their labor force, subject to certain restrictions.

Despite the expanded possibilities for employers to exempt workers from the “last in – first out” principle during layoffs, the new regulations also impose stricter obligations on full-time and permanent employment, potentially limiting the use of temporary contracts. Consequently, the overall effect of these changes on labor-market adaptability remains to be seen.

4.5.2 Wage Formation

An important step towards more predictive and business-friendly wage formation was taken in 1997 with the Industry Agreement (collaboration between employers and trade unions on industrial development and wage formation). The Industry Agreement prioritizes negotiations between unions and industries that are highly exposed to international competition. The outcomes of these negotiations serve as a reference point for wage increases and other employment-related conditions across the broader economy.

The Industry Agreement contributed to consistent real-wage growth annually until 2021/2022, demonstrating its effectiveness in fostering economic prosperity for workers. Prior to 1997, a significant portion of wages were determined centrally, with a focus on relative wage development across industries and worker categories. The Industry Agreement has facilitated a shift towards greater emphasis on individual skills development and productivity. The shift towards collaborative negotiation has led to a decrease in labor-market disruptions, such as strikes, which previously resulted in substantial costs for employers.

The major share of the Swedish labor market is covered by the Industrial Agreement, or by other agreements based upon it. Under the Agreement, many arrangements are made without any connection to centrally negotiated wage structures—everything is decided at the local level. However, most of these agreements still include minimum wage-level guarantees. The Industrial Agreement stipulates that trade unions and employer organizations must take greater responsibility for wage formation by concluding special collective agreements on cooperation and negotiation.³⁵

Despite the fact that the Industrial Agreement constitutes a major improvement compared to the previous arrangement, it is still the case that the norms and institutions that govern wage formation have not been fully adapted to the new economic context in which firms operate, which creates recruitment problems for fast-growing firms. Centralized wage negotiations in combination with high minimum wages (compared to the median in the industry) tend to disadvantage smaller and younger firms, especially in the service industries where it is otherwise usually easiest for firms with a good business idea to grow quickly (Henrekson and Johansson 2010). This follows from the fact that the wage level is consistently higher in larger and older firms (Oi and Idson 1999).

This compressed wage structure also means that the private financial return on higher education is relatively low in Sweden, which reduces the incentive for individuals to treat education as an investment in economically valuable knowledge and skills. As stated by Andersson and Thulin (2008), a low private economic return on higher education also tends to lead to a low societal return, probably because individuals pay less attention to demand in the labor market when deciding what to study. Instead, the consumption element in education is given greater weight, and education is postponed and becomes less useful (which can be seen in the fact that graduates work fewer hours after completing their education than they used to).

The further from the individual workplace that the salary is set, and the less the consideration given to the specific circumstances of each individual case, the more difficult it becomes for fast-growing firms to build a workforce with the right skills and competencies. Idiosyncratic intrafirm differences tend to be particularly large in new, small but fast-growing industries and firms (Caballero 2007; Haltiwanger 2011). Thus, these firms tend to be harmed the most by wage-setting institutions that render it difficult to take into account specific circumstances and individual employee characteristics when employees are compensated for their efforts.

4.5.3 The Labor Market and the Social Insurance System

By providing insurance against failure, the public sector can reduce the risk for individuals in entrepreneurial firms characterized by high levels of uncertainty. Here, however, it is important that the social insurance system is generous enough to be relevant even to income earners with above-average salaries, and that security is not linked to length of employment with the current employer. There is reason to believe that a Danish flexicurity-type system combining weaker employment protection with more generous public income insurance makes it easier for firms to grow quickly (Klindt 2010).

Research and policy recommendations in these areas are obviously controversial in a Swedish context. Nevertheless, it is important to clarify the inverse relationship here between security and dynamism. Unfortunately, it is a fact that several of Sweden’s traditional social security systems and norms for wage formation restrict freedom of contract and limit the possible combinations of competencies in the labor market and in production. This makes it more difficult for entrepreneurs to experiment and test new ideas according to the ideal model of their socio-economic and innovative role described above. The entrepreneur therefore creates less value than in an environment with less restrictive rules.

Against this background, we believe that a comprehensive policy for a more innovative society should strive to make the security aspect of social insurance systems completely mobile. Public risk insurance should provide flexicurity. This means a decent and generous unemployment insurance fund that places high demands on mobility, accepting new jobs, and being prepared to retrain and acquire new, more productive knowledge. Sickness and unemployment insurance should be clearly designed as transition insurance, not as an alternative to permanent employment, and pension systems should be fully actuarial.

In a number of countries (including Sweden), the self-employed have poorer social insurance coverage than employees. And for those who succeed in profiting from their entrepreneurship, high social security contributions become an additional marginal tax if the beneficiary reaches the ceiling of the system.³⁶ Moreover, the self-employed, regardless of the legal form chosen (incorporated, sole proprietorship, etc.), do not utilize welfare schemes to the same extent as ordinary employees; they cannot afford to completely ignore their businesses if they are sick or on parental leave, which is a requirement for various transfers. In addition, it is a considerably more complex task to calculate the wage that provides the basis for remuneration from social insurances since these are based on historical wages. As an example, the remuneration for employees is based on an average of wages over the prior 12 months up to a certain ceiling. Wages above the ceiling can be covered through income insurances. For the self-employed entrepreneur, wages are much more volatile and total income is often reported later, i.e., when the business closes the books for the year. Income insurance is also harder to obtain.

The remuneration of entrepreneurs therefore tends to be much more uncertain and volatile than for employees, implying higher risks. Combined with a progressive tax system, this makes the overall environment less conducive to entrepreneurship. This aspect is accentuated by the emergence of the gig economy, where it is often unclear whether workers are employees or self-employed; institutions need to be adjusted to accommodate these new contractual arrangements in the marketplace (SOU 2016:72).

4.6 Product Market Regulations

If product markets are to give participants the impetus to dare to experiment, dominant firms must not abuse their market power. Regulations must be appropriate and provide the right incentives to market actors. This is easier said than done—it is easy to find examples where regulations miss the mark, benefit a certain interest group, or entail insurmountable costs. Technological progress can also make regulations obsolete and thereby hinder growth and adaptation to new conditions.

Recent research points in particular to the risk that regulations may weaken competition by making it more difficult for new entrants to establish themselves in the market. Weak competitive pressure erodes companies’ drive to innovate and adopt new technology. Not least, research shows that adaptation to new IT and communication technologies can thereby be inhibited, which has major negative effects on productivity growth. Whereas the United States used to be considerably less regulated and more flexible than Europe (Poschke 2010), this advantage seems to have at least partially eroded (Philippon 2019). More generally, the trend towards a more lenient regulatory framework between the 1990s and 2010 seems to have levelled out or reversed. This mirrors political concerns regarding financial market stability after the 2008–2009 crisis, the emergence of digital platform firms, and more generally an increasingly insecure and tense geopolitical situation that has prompted governments to define certain strategic areas and to impose trade restrictions.

Regulations that reduce competitive pressures also mean weaker incentives to move capital and labor from firms with lower productivity to those where it is higher. High-productivity growth presupposes that such transfers can take place relatively painlessly, given that productivity differences are often large between firms in a given industry at any given time. Depending on the composition of the industry and the skills of the workforce, the effects can vary. Nor must they be linear—according to Arnold et al. (2011), regulation can cause abrupt shifts in the production function. Between 2020 and 2023, such impediments to efficient allocation of resources have been aggravated by the plethora of support schemes and tax reliefs introduced during and in the aftermath of the pandemic.

In the case of Sweden, major deregulatory steps were taken in the early 1990s, in connection with the crisis. After that, the tempo has slowed down. However, Fig. 4.5a and b shows that Sweden still retains a favorable position compared to other countries in terms of product market regulation

Fig. 4.5

(a) Product market regulations in a number of leading countries (index 0 to 6). (b) Economy-wide product market regulation indicators broken down by major components. Note: Index scale from 0 to 6 from most to least competition-friendly regulations. Source: OCED (2018)

It is critically important that the recent tendencies towards increased regulation are kept at bay, or that altered regulation strives to improve the functioning of markets, such as constraining obvious instances of abuse by dominant firms, or increasing national security. At the moment, there are forces that seek to limit the extent of the market: trade and foreign ownership restrictions, competition and state aid regulations, and the introduction of a laxer fiscal policy framework in the European Union (in 2024).³⁷ These factors, individually and in combination, open up possibilities for curbing competitive forces and partially insulating domestic markets from international competition (Braunerhjelm and Lappi 2022). There is consensus that regulations that restrict market entry and competition also inhibit productivity (Inklaar et al. 2008; Andersson et al. 2012), since such regulations protect old and less efficient firms. This is one likely reason why productivity growth has slowed in the last decade.

Since most sectors have innumerable and detailed regulations, we will refrain from submitting detailed proposals, sector by sector and industry by industry. Instead, we want to expand the mandate of Sweden’s Better Regulation Council (established in 2008 with the aim of simplifying business regulation), and concurrently provide the Council with substantial resources to be able to conduct thorough impact assessments.³⁸ It should examine new legislative proposals and identify any risk that they will have an inhibiting effect on competition and innovation.³⁹ In the event of major risk, the Council should have a right of veto. It should also review existing regulations. In addition, all new proposals should be subject to a “sunset clause”—new regulations must be reconsidered on an ongoing basis; otherwise, they will automatically expire after a few years.

4.7 Tax-Financed Welfare Services

In welfare services, machines cannot replace people in the same way as in industry or transport. A nurse, for example, cannot be automated. Labor productivity—output per working hour—therefore tends to increase more slowly in welfare services than in manufacturing. This in turn means that the relative cost of welfare services will rise. Therefore, it becomes especially important to organize this activity in such a way as to stimulate new thinking and innovation. In particular, competition and transformation pressure are required to help offset higher costs. Without such competition, the cost of schooling, nursing, and social care will rise even further.

As early as the 1980s, certain initiatives were taken that allowed for private alternatives (such as the private nursery school franchise Pysslingen), but the main decisions were made in the early 1990s. Parliament then adopted the position that tax financing would be retained as a basic principle, but that private alternatives, for example in childcare, healthcare, elderly care, and education, should meet the same financial conditions as the ones administered by the local or regional government. Other possibilities such as public basic financing with a right for competing contractors to charge fees for better quality or additional services were never considered.

As always, it takes time before institutional changes take effect in a meaningful way. Firstly, actors must realize the consequences of the changes, which must be perceived as lasting, and secondly, behaviors on both the producer and consumer side must adapt to new conditions. Today, several decades after the opening up of social, medical, and educational services to competition, the growth of private service providers has gained momentum. This is clear from Table 4.2, which examines the private sector production share for major welfare services that are either entirely or primarily tax financed. The private health center share is rapidly approaching 50%, for elderly care the share is around one quarter, almost one third of secondary school students attend a private school, and two-thirds of personal assistants for the disabled are employed in the private sector.

Table 4.2 Private sector production share for major services that are primarily publicly funded, 1996 and 2020/2022 (%)

There are now more than 15,000 private firms in the health and social care sectors. Several large firms have also been built up in a fairly short time, both in education (the largest one is AcadeMedia with 18,800 employees) and in healthcare and social care (notably Attendo, Ambea, Humana, and Capio with roughly 30,000, 26,000, 16,000, and 11,000 employees in the Nordic countries, respectively).

In normal markets with private financing and production, it is hardly controversial that freedom of choice and competition benefit consumers. That the same should apply in markets with private production but public financing is far from obvious. Today’s welfare services are traded in quasi-markets with a wide range of problems: the services in question are complex and difficult to procure; formalized procurement processes benefit large actors and limit competition; information is asymmetric between users, producers, and financiers; quality improvements are not sufficiently rewarded when the producer cannot charge a higher price for improved quality; overcapacity can arise; lack of information makes it difficult for users to make decisions; follow-up and control are skill- and resource-intensive; there are segregating forces; and individual users do not take into account how society as a whole is affected.⁴⁰ Manipulation and waste occur more easily when an anonymous and absent third party (the taxpayer) acts as an intermediary and finances all transactions between producer and consumer. In addition, most of these services are ideologically charged, which also makes technical and administrative details politically controversial.

The difficulties posed by quasi-markets are at their most challenging with respect to the so-called credence goods. These goods are characterized by, on the one hand, the producer knowing more than the consumer about the latter’s needs, and on the other, the fact that the quality of what is bought cannot be observed by the consumer even after purchase. Expert services such as medical procedures, automobile repairs, and dietary supplements are typical examples. The consumer often finds it difficult to assess for herself whether she needs a certain service, for example a certain type of treatment under the care of a specialist. A producer of a credence good can therefore, by virtue of their expertise, exaggerate or understate the consumer’s needs. In order to achieve good quality and efficiency in the production of credence goods, the producer must therefore provide the correct information. If it is not possible to verify the treatment performed (or the care or training provided), or to hold the producer accountable for a bad outcome, the producer is unlikely to deliver the desired quality.⁴¹ If, on the other hand, at least one of these conditions is met, it is possible to draft an agreement that gives the producer the right incentives to provide an appropriate and high-quality product or service. To some extent, consumers can rely on the producer’s reputation, but reliable information can only be disseminated by consumers who can understand and evaluate the service in question.

Empirical studies of contract procurement confirm that credence goods have the worst outcomes (Andersson et al. 2019). This is illustrated by institutional care of young people, where the total cost was found to be twice as high for young people in private homes compared to municipal ones (Lindqvist 2008). How such tendencies are counteracted is decisive for the functioning of quasi-markets.

4.7.1 Public or Private?

Privatization of quasi-markets creates challenges when the profit motive has wider scope. Without adequate regulation, lenient grading becomes a means of competition among schools; reduced staffing in nursing homes is a means of increasing returns; and an unnecessary number of return visits to health centers is a way of increasing patient reimbursements. But we must not forget that large segments of the public sector still suffer from the same kind of problems highlighted in the political battles of the 1980s: vague goals, problems with delegating responsibility, lack of competition and low quality of services provided, absence of “carrots and sticks,” inability to process dispersed and fragmented information, “soft” budget constraints that keep underperforming entities afloat, political considerations during election years, and so forth. Third-party financing cannot simply be designed so that these problems disappear (Andersson et al. 2019). Our point is that there is no evidence to suggest that public actors—despite the imperfections of private alternatives and despite sensationalist media headlines—outperform private ones. Swedish comparisons indicate that differences in quality between public and private contractors are generally small and uncertain (Blix and Jordahl 2021). Moreover, it should be noted that the best suppliers are often private.

Comparisons between public and private service providers must be based on a theoretical assessment of the advantages and disadvantages of different incentive systems and the organization of production of the service in question so that empirical tests are formulated correctly. One starting point is Shleifer (1998). He sets out four criteria under which public production can be superior to private:

• When it is impossible through contracts and penalties to prevent the producer from reducing his costs through lower quality

• When the potential for innovation is relatively small

• When competition is weak and consumer choice is inefficient or inoperative

• When damage to the supplier’s brand carries no consequences, or if the brand is hard to damage

Shleifer himself sees these criteria as highly restrictive for publicly produced services, as it is usually possible to create adequate conditions for private services through regulation that can ascertain whether the above criteria are met. He asserts that the number of activities that must be carried out by the government is very limited, but not zero.

However, if the regulatory framework is not properly designed, it may well be that Shleifer’s criteria are met in practice. In cases where the choice of producer and the framework for production are determined by public procurement, the competitive element (Shleifer’s third criterion) is limited compared to a market in which several producers sell directly to multiple consumers. It is therefore all the more important that competition be maintained throughout the (often highly complex) procurement process. A great deal of development and learning has taken place in this area in recent years. However, there still remains much to be done, not least when it comes to how the rules for auction processes should be designed.

The contract itself (Shleifer’s first criterion) is also crucial. One challenge is to draft contracts that make it impossible to reduce costs by lowering quality. On the other hand, contracts must not be so detailed that they rule out productive cost savings or innovations. There is an extensive literature discussing how such agreements should be designed and how the procurement process should be structured to achieve maximum efficiency and clarity.⁴² The simpler the service to be procured, the easier it is to arrive at the right form of contract or the right procurement process. The more complex the service becomes, the greater the demands on both.

Our conclusion is that the public sector should be exposed to sustained and intense pressure to reform. Wise procurement, good regulatory systems, and well-functioning competitive markets can improve the quality and reduce the cost of welfare services—even if the task is difficult. Procurement by tender is also absolutely necessary for major energy and infrastructure projects where the time horizon is too long for private investors. Stronger institutions and regulations are thus necessary to monitor contracts, ensure compliance, and impose fines and other sanctions when the contractual terms are not met.⁴³

4.7.2 Customer Choice and Consumer Protection

No matter how well one succeeds in the development of contract and procurement processes, the competitive element is limited for certain markets to the actual procurement opportunity. The actor who wins a tender acquires a temporary monopoly in the relevant area. One should therefore strive for providers of welfare services to be determined by customer (or user) choice.

A fundamental problem, however, is asymmetric information. The buyer is consistently at an information disadvantage in relation to the seller. In other areas, this has been solved by means of consumer protection regulations. These grant the consumer a number of rights, such as the right of return, price information, and opportunities to sue for damages. In addition, the consumer receives information, for example in the form of product comparisons and blacklists of sellers who do not comply with the rules. For property sales—a situation in which the problem of asymmetric information is obvious—there is extensive regulation governing the sales process which aims, among other things, to neutralize the seller’s information advantage. If the seller has withheld negative information about the property, the buyer can seek legal remedies.

In the welfare sector, consumer information and protection are limited and at times nearly non-existent. We find this odd, because there is usually significantly more at stake for the consumer regarding the delivery of welfare services than when purchasing other goods and services. Finding the best cataract surgeon should be more important than finding the right hotel on a package holiday. The quality of a child’s education is more important than the quality of one’s domestic appliances. Sweden’s National Board for Consumer Disputes can intervene if a traveler has a bad hotel experience, if their flight is delayed, or if the hotel’s broadband connection is slow. But quality in the welfare sector is seldom “verifiable” in the sense that it is possible to prove to a third party (i.e., in court) that the quality is unacceptable. For example, in elderly care (public as well as private), although staff and clients may know when quality is substandard, they still cannot prove this claim “beyond all reasonable doubt.”

Unfortunately, the weak regime of consumer protection and advice regarding public sector welfare services has been transferred to emerging private sector services. Private welfare firms are thus probably the only consumer firms in Sweden that operate in markets that are largely devoid of consumer advice and protection. This contributes to inefficient consumer choice. The media debates caused by the negligence and quality problems involving a number of service providers show—despite exaggerations and perhaps even false alarms—that media scrutiny is a corrective to abuse. However, such scrutiny is probably insufficient to prevent such abuse and neglect, regardless of whether care is provided privately or publicly. An important element of any strategy in conveying the benefits of entrepreneurship to the welfare sector is effective rules for consumer advice and protection in these areas as well.

4.7.3 The Role of Profit in the Welfare Sector

Shleifer notes in his analysis that profit-seeking and dividend opportunities are necessary for the creation of the dynamics that make a private, market-based system superior to a planned economy. Opportunities for profit are an important impetus for innovation (Shleifer’s second criterion), which grows in importance in pace with the growing importance of the welfare sector. Opportunities for profit also accelerate the dissemination of innovation and good examples to other establishments—companies are quick to codify and standardize elements that have a positive effect on the bottom line. In addition, if entry is allowed, competition is likely to bring back profits to more normal levels.

Evolutionary development of contracts, procurement procedures, and quality evaluations within the framework of a transparent regulatory framework is the most effective value-creating way to proceed (Elert and Henrekson 2023). When markets are functioning well, nurturing one’s own brand serves as the strongest corrective against abuse. This in itself creates competitive advantages for larger players, where users know that damage to a brand in a single branch can jeopardize the reputation of the entire firm, thus leading to large losses for the owners. However, this mechanism is much weaker for smaller enterprises such as private families caring for young people with serious social problems.

Non-profit foundations can also play a role as producers of welfare services, and they can raise the bar and intensify competition. They also contribute to diversity and broader options for consumers. However, it is hard to imagine that non-profit providers, whose market share is falling, can satisfy the increased demand resulting from rising real incomes. Some non-profit providers want to remain exclusive and do not wish to grow (some prestigious schools, for example), while others aim for niches where supply would otherwise be lacking (such as parochial schools). Therefore, it is hard to see how anyone other than for-profit firms can accommodate the mass market and thus be able to respond to an increase in demand.

This makes it necessary to improve the management of the profit motive. Although there are no simple solutions to this issue, further measures can be taken. External evaluation should become more significant. In educational institutions, grading and examinations should be conducted externally; this appears to be a reform that can increase the consistency of grades for all students and over time.⁴⁴ To manage the risk of rising costs, increased elements of personal financing should be considered in areas where the distributional effects are small. For example, increased patient fees in primary care would raise cost awareness without seriously ill people suffering major financial damage. Another example relates to nursing homes. In this area, it would be much more acceptable to allow the purchase of additional services (topping up). This would remedy the epistemic problems generated by fixed prices and create a sort of experimental workshop including a diagnostic tool to identify users’ true needs.

Requirements for transparency and documentation should be equal for public and private contractors. The fines levied against for-profit companies should be significantly higher, and for international groups in particular where no local owner has a reputation and personal finances at stake. The fines should also be levied against the owners and not the establishment.

Several of these proposals require new forms of oversight. To use the terminology of McCubbins and Schwartz (1984), some argue that the so-called “fire alarms” can be a more effective surveillance strategy than “policing.” As the labels suggest, policing refers to centralized, active, and direct surveillance, while alarms are based on regulatory systems that give individual citizens and interest groups the opportunity to be noticed and to demand accountability. Both aspects are needed. But while policing is more objective, it risks becoming rigid and formalistic. Fire alarms are more subjective, for better or worse—they are triggered when an interested party believes the situation has gone too far and flags this to the media or authorities.

4.8 The Housing Market and the Benefits of Agglomeration

In the previous chapter, the point was made that conditions for growth and job creation are generally more favorable in metropolitan areas than in smaller towns.⁴⁵ Urban areas also create fertile ground for innovation, particularly in the service sectors, and are thereby essential for overall growth in an economy (Andersson and Larsson 2022). A large city is attractive not only because salaries tend to be higher there and cultural offerings more ample; one’s income has a higher value because there are more goods and services available to spend it on. In other words, the benefits of earning more are therefore greater in most cases.

4.8.1 The Importance of Price Efficiency

In larger and more densely populated cities, land is a scarce resource. This means that the market price for both housing and commercial premises is higher. Under such circumstances, a well-functioning housing market is crucial. To this end, price formation to arrive at a structure that correctly reflects how users value different locations and types of premises is central. This is because it sends the right financial signals to city planners, construction companies, and landowners regarding how and where new development will attain the greatest value. Healthy price formation is also necessary because requirements for living space and the form of housing change during the span of a resident’s lifetime.

Given these factors, it is essential to determine how the housing market can work better. A look at Sweden’s recent history of rent control can be enlightening in this respect. At the beginning of World War II, Sweden began to regulate rents, a situation which later became permanent despite extensive criticism.⁴⁶ Above all, the value-in-use principle has meant that rents have only marginally reflected location. Apartments in attractive locations have therefore been far less expensive than had they been offered on the free market. This in turn has created significant lock-in effects. However, some measures have been taken to loosen the value-in-use principle.⁴⁷ In 2011, the role of municipal housing corporations in setting rents was abolished. According to these new rules, the standard, quality, and location of each individual apartment must to a greater extent determine its rental value. However, rents are still negotiated collectively, and these negotiations are normative for all tenancies. Municipal corporations are required to operate commercially, i.e., construction costs no longer form the basis for rent negotiations.⁴⁸

This new practice is a step towards rents that more accurately reflect the overall economic picture, but the system is still complicated and will continue to give rise to large and unjustified differences while falling short of creating a well-functioning market. In attractive areas, the lock-in effects on the rental market will to a large extent remain.

At the same time, pricing is determined more freely for another type of apartment, namely condominiums (and from 2009 owner-occupied apartments as well). This has given rise to a strong tendency for property owners to convert their rentals into condominium/owner-occupier arrangements.⁴⁹ The housing market needs both a large tenancy sector and a large ownership sector in order to be healthy; however, many households are unable or unwilling to take the risks that ownership entails. It is also unfortunate that many are forced to exhaust all available credit on a home, when they for example may have preferred to use some of it to finance a business, either their own or that of an acquaintance. Similarly, private residential housing causes a loss of the economies of scale provided by property management, maintenance, and so forth.

It is also those new to the property market who must bear the costs of the regulated system. They are pushed into risk-taking through condominium purchases, subletting, or transactions on the black market. Poor adaptation of housing consumption to consumer preferences also has side effects on the labor market and production. If it is difficult for newcomers to find housing, this affects a company’s ability to recruit labor. All in all, one can consider today’s housing market—in terms of construction, renting, and mortgages—an unnecessary obstacle to entrepreneurial development.

4.8.2 Driving Forces for Relocation

Relocation is an important factor in a healthy property market, but tax regulation in Sweden causes impediments to it. We believe that continuous taxation of housing would facilitate growth in those densely populated environments where the benefits of agglomeration are greatest. Given that housing needs change during one’s life span at the same time as there is a shortage of living space in densely populated environments, it is important that there are strong incentives for relocation, such as selling a detached house and moving to an apartment or a smaller home when the children move out. This is facilitated if housing is taxed on an ongoing basis rather than when the occupants move out, as is now the case. The change in property taxation from continuous taxation to payment of a high capital gains tax has therefore damaged the market. In metropolitan areas in particular, where the need for relocation is greatest, moving from one property to another may trigger a significant capital gains tax which reduces the inclination to move to a new residence when the family situation or preferences change. This effect has been mitigated since 2021, when it became possible to defer the capital gains tax if the seller bought another home.⁵⁰

If a tenant does not wish to move, more relaxed rules for subletting could increase the supply of apartments, especially for first-time residents in large urban areas. To increase the number of residents living in sublet accommodations,⁵¹ these rules would need to include freedom of contract and a radical alteration of the right to tax-free subletting.⁵² Increased opportunities for subletting and tax incentives that do not lock households into oversized housing would also make it easier to introduce market rents.

In the current situation, market prices for housing are being gradually introduced in practice, but in a very cumbersome way: by converting rental apartments into condominiums. Unfortunately, this also forces many to take an undesirably large financial risk in order to find housing. Without freedom of contract for rentals, these condominiums often remain unused during times when the actual owner is not utilizing the apartment.

4.8.3 Other Measures to Increase the Benefits of Agglomeration

Facilitating the growth of metropolitan areas also presupposes improved infrastructure and expanded public transport. Investment in railways and trams, both locally and regionally, is particularly important.⁵³ In contrast to long-distance, high-speed trains, requiring large investments,⁵⁴ these are often profitable. Often, their development requires several local and regional centers; in other words, high population density must be combined with multiple urban “cores,” which further underlines the need for investment in local infrastructure.

Another challenge to be addressed is the system of municipal tax levelling, in which a large part of richer municipalities’ tax revenues is siphoned off to poorer municipalities. An amount corresponding to roughly three percent of GDP is redistributed through the system.⁵⁵ The purpose of this redistribution is legitimate: to guarantee equal service regardless of the strength of the local tax base. However, the cost of many public services usually rises in line with municipal income per capita. This is partly due to the fact that local costs are higher in denser environments. Another reason is that staff employed in tax-financed activities must have wages in line with those in the local business sector so that the city can afford to recruit and retain competent personnel. A desired increase in urbanization, and thus higher innovation and growth, requires a well-balanced redistribution of municipal taxes across urban and more rural communities.

4.9 Attitudes and Cultural Perceptions

The behavior of entrepreneurs and other actors is obviously not governed solely by economic incentives. Cultural and psychological factors also play an important role. An entrepreneur can, for example, be driven by a desire to realize a project or business idea for its own sake, or by the dream of proving to herself and others that she is capable of putting an idea into practice and achieving success. A society that rewards and encourages such dreams becomes more creative and more entrepreneurial than one that rewards conformity.⁵⁶

Hence, the entrepreneur does not need to be rewarded primarily in pecuniary terms. Social standing, media attention, awards of various kinds—the driving forces can be many and varied.⁵⁷ But even if financial gain does not need to be an end in itself, it still serves a function as an indicator of ability and success. In addition, successful entrepreneurs serve as role models who encourage others to enter this field and test their own entrepreneurial abilities. Profit, actual or anticipated, is also a necessary condition for obtaining the resources for innovation and growth. Although the pursuit of profit may not be the final goal for the individual entrepreneur, profit in the current economic system is a means for those who want to realize their entrepreneurial vision in the form of a successful business.

A measure of the change in entrepreneurial intention is presented in Fig. 4.6. As can be seen, the share of individuals who report plans to start a business in Sweden has increased since 2016–2017, but this trend was broken in 2022. The Netherlands also regressed somewhat in 2022 after a remarkable increase in entrepreneurial intentions in the last two decades. The United States and Switzerland experienced a rise after the financial crisis, which was likewise reversed in 2022. Israel showed the most marked increase until 2018/2019, which was followed by a similarly sharp decrease. Then there was a sharp upturn in 2022, making Israel the country with the highest entrepreneurial intentions of the included countries. At the bottom, we find the large EU countries together with Norway. Overall, most countries saw a decline in 2022, likely to have been influenced by an increase in pandemic-related bankruptcies in several countries and faltering growth prospects.

Fig. 4.6

Entrepreneurial intention: Share of 18–64-year-olds who intend to start a business within three years, 2002–2022 (%). Note: Large EU countries consist of France, Germany, and Spain. Source: Thulin (2023)

In a broader sense, attitudes in society significantly influence the opportunities and career paths an individual considers or even notices. In this book, we have explained the importance of providing room for curiosity and experimentation and of rewarding success. But concrete policy measures in these areas are unlikely to occur unless the broader culture and its values provide an impetus in that direction. Positive attitudes towards entrepreneurship and business ownership are thus a prerequisite for maintaining a high level of entrepreneurial activity or stimulating increased entrepreneurship. Politicians play an important role in signaling the value that a society attaches to entrepreneurial endeavors.

Many of the attitudes and beliefs we carry with us originate in the home, in school, and in our closest circles of acquaintances, often when we are quite young. Much can be done to promote entrepreneurship during the formative years, for example through training and encouraging creativity and entrepreneurship in children. Not infrequently, entrepreneurs have family members, for example a parent, who have also chosen to be entrepreneurs.⁵⁸

At the same time, the expressed attitudes regarding firm ownership and entrepreneurship are likely to be a reflection of regulations and reward structures. For many individuals, pursuing a career as an entrepreneur does not seem sufficiently attractive. The expected compensation for successful entrepreneurship is not perceived to be in reasonable proportion to the risks and uncertainty it entails.

Negative attitudes towards entrepreneurship can thus be based on more fundamental factors and deep-seated ways of thinking. To a large extent, reward structures in a society are a codification of attitudes and norms. The previous tax regime—so unfavorable for entrepreneurship and business ownership in Sweden—was thus, as we have argued above, an expression of an underlying attitude that the optimal state of society was well-functioning capitalism—but without individually successful capitalists.

The relationship between norms and attitudes, on the one hand, and the institutional framework on the other is complex and difficult to change. If institutions facilitate and encourage value-creating activities that lead to increased welfare for the majority, it is more likely that people will favor institutions that lead to increased predictability and legal certainty, stronger protection of private property rights, and a high return on productive entrepreneurship.⁵⁹ But changing legislation, tax codes, and societal reward structures often presupposes changing attitudes. Policies more favorable to an entrepreneurial society therefore include both technical and practical interventions in tax rates and regulatory systems as well as opinion building in the long term to reshape attitudes towards entrepreneurship and business ownership.

4.10 Conclusions

In this chapter, we advocate a much broader policy approach than the one that dominates the daily political debate on innovation and entrepreneurship policy. Focus should be on general measures to lay the foundation for the emergence of an environment facilitating the further development of potentially innovative and fast-growing firms irrespective of size, age, and industry. We have reviewed the following policy areas that we deem crucial to promote innovation and entrepreneurship: the quality of the education system, the role of the university sector, incentives to private R&D, the channeling of saving, the functioning of the labor market, the social insurance system, product market regulations, tax-financed welfare services, the housing market, and attitudes and cultural perceptions.

We have devoted a separate chapter to what is arguably the most potent measure in the government’s policy arsenal, namely taxation. The overall conclusions we draw from our analyses of the different areas, including taxation, will be presented in the concluding section of the next chapter.