There is growing concern that dynamism in the business sectors of western economies is slowing down. Decker et al. (2014) document a decline in the activity of young firms in the US, as displayed in Figure 1. The figure shows a steady decrease in the share of young firms from almost 45% in 1992 to approximately 35% in 2011. It also displays a decline in the share of employment in young firms, from 16% in 1992 to below 11% in 2011. Finally, it shows that the share of job creation in young firms dropped from 37% to below 29% during this period.

Figure 1 Declining share of activity among young firms in the US (firms aged 5 years or less)

Source: Decker et al. (2014).
Image: Vox EU

Using detailed Swedish employer-employee data over the 1990–2013 period, we observe no decline or a smaller and later decline in Sweden for different measures of entrepreneurial activity (Heyman et al. 2019).

Figure 2 shows the development of the Swedish business sector during the same period as in Figure 1. Young Swedish firms (aged five years or less) accounted for more than half of all firms, averaging approximately 55% per annum, in sharp contrast to the case for the US depicted above in Figure 1. There is a similar pattern for the job creation share, i.e. higher levels for young Swedish firms and more stable over time, which is different from the pattern in corresponding US firms.

The evolution of the employment share of young firms in Sweden, however, does share some similarities with the development in the US. While young firms account for a larger share of employment in the business sector in Sweden, there is a tendency for a downward trend in the share of employment for young firms (aged five years or less). This decline is not present if we define young firms as start-ups, i.e. completely new firms. For these firms, the employment share has been stable during the studied period.

Figure 2 Share of activity among young firms in Sweden (firms aged 5 years or less)

Source: Heyman et al. (2019).
Image: Vox EU

When studying firm and job dynamics more in detail, we find that the start-up rate has increased in the Swedish business sector over this period. However, job destruction rates have also increased for firms, especially for those in their first year after entry. Since the mid-2000s, this increase in job destruction for the very youngest firms has led to a small decline in the employment share for young firms that have been in business for less than five years.

Looking across sectors, most of the job creation by young firms in Sweden has occurred in the expanding service sector, in contrast to the evidence for the US, where job creation by young firms has decreased in the service sector. Other dynamic patterns appear to be driven mainly by the service sector in Sweden. For instance, job creation from start-ups stems from the service sector, and job destruction has become increasingly important in this sector also.

Start-ups and young firms play an important role in US job creation. Evidence from the US suggests that there is a distinct negative relationship between firm age and employment growth within each age group. Mature firms are associated with lower rates of job creation than younger firms. While we do find similar trends in Sweden, the relationship does not seem as strong as in the US. Net job creation among Swedish firms seems to be rather stable for firms aged one to seven years old, unlike for the US, where the rate drops at a higher pace.

A micro-based reform package in the 1990s increased entrepreneurial activity in the Swedish business sector

Then, what could explain why the Swedish business sector has been more entrepreneurial in recent decades in comparison with the US? Why do the signs of lower entrepreneurial activity appear later in Sweden?

In two previous papers (Heyman et al. 2015, 2018), we argue that the implementation of an extensive business reform package conducive to entrepreneurial activity after a severe economic crisis in Sweden during the 1980s and early 1990s is a likely explanation. It is often suggested that it was the macroeconomic reforms, such as moving from a fixed to a flexible exchange rate or adopting an inflation target with an independent central bank, that enabled Sweden to recover from the crises. In contrast, we stress the micro-based reforms, which removed barriers to entry and growth for new firms in several product and services markets and made corporate ownership and labour markets more flexible.

In Heyman et al. (2015) we study productivity development in the Swedish business sector during the 1996-2013 period. As Figure 3 shows, a large share of productivity creation during this period stems from the entry of efficient firms and the exit of inefficient firms. New entry does not create productivity growth in the short term because firms entering the market are often less productive than average. New entrants, however, create major productivity gains in the longer term. More than half of the overall increase in average productivity in the Swedish business sector stems from firms that were founded in 1996 or later, reflecting a selection process in which only the most productive new entrants survive.

Figure 3 Labour productivity growth decomposition, 1996–2009

Source: Heyman et al. (2015).
Image: Vox EU

This development of the business climate in Sweden contrasts with the development of the business climate in the US, as described by Zingales (2012), among others. He argues that the US has evolved into an economy that favours incumbents over entrants, becoming pro-business instead of pro-market.

Reforms or technological developments?

In Heyman et al. (2019), we also discuss other potential explanations. We highlight that a complementary explanation is that information and communication technology (ICT)-driven structural change began later in Sweden, which could explain the stronger job creation process in the service sector during this period in Sweden compared to the US.

Further analysis of the observed differences between Sweden and the US would be of interest. If the different pace of adjustment in the two countries is due to differences in regulations and institutions, it suggests that micro-based efficiency-oriented reforms have potentially large positive consequences for long-run growth in jobs and productivity. On the other hand, if the driving force is the different timing of the implementation of ICT, then the welfare consequences of such micro-based efficiency-oriented reforms will be different. Research along these lines seems warranted.

Current need for reforms

While Sweden's reform period led to improved economic performance in the decades after the deep crises in the early 1990s, there are also signs that the dynamics of the Swedish business sector have deteriorated in the last decade. Productivity in the country’s business sector increased only by an average of 0.2% per year during the period 2007-2014. Moreover, the number of employees in start-ups has decreased continuously since 2010. In Sweden, as in many other countries, there is also an increased worry that new technologies, ‘smart’ robots and increased globalisation will lead to job losses, high unemployment, and increased income inequality.

Our analysis suggests that reforms increasing incentives for firms to find new business opportunities, utilise new technologies, experiment, and grow – locally as well as globally – can be one way to handle these problems.