IFN Working Paper No.

1147, 2016

Has the Swedish Business Sector Become

More Entrepreneurial than the U.S. Business
Sector?

Fredrik Heyman, Pehr-Johan Norbäck, Lars Persson

and Fredrik Andersson

Research Institute of Industrial Economics

P.O. Box 55665
SE-102 15 Stockholm, Sweden
[email protected]
www.ifn.se
 Has the Swedish Business Sector Become More
Entrepreneurial than the US Business Sector?
Fredrik Heyman*, Pehr-Johan Norbäck*, Lars Persson* and Fredrik
Andersson*

April 12, 2019

Abstract: Recent studies document a 30-year decline in various measures of entrepreneurship

in the U.S. Using detailed Swedish employer-employee data over the period from 1990 to
2013, we find young firms to be more prominent in the Swedish business sector than in the
U.S. business sector. Young Swedish firms, aged five years or less, account for more than half
of all firms during this period. We also observe an increase in Swedish entrepreneurial
activity for start-ups. However, since the mid-2000s, job destruction rates for young firms
have been increasing, which implies a declining employment share for younger firms.
Moreover, most of the job creation by young firms occurs in the expanding service sector. We
discuss different explanations for why Sweden appears not to have the same strong decline in
entrepreneurial activity as the U.S. has had during the last two decades. We argue that one
important explanation is the economic reforms that were implemented in Sweden in the 1990s
that mitigated several hurdles to entrepreneurship.

JEL: J23, K23, L26, L51

Keywords: entrepreneurship, job dynamics, matched employer-employee data, industrial
structure and structural change

* Research Institute of Industrial Economics (IFN), P.O. Box 55665, SE-102 15 Stockholm, Sweden,
[email protected], [email protected], and [email protected].

Acknowledgements: We thank Magnus Henrekson, Ed Lazear, Assar Lindbeck, Roger

Svensson, Joacim Tåg, the Editor and two anonymous referees for their helpful comments.
Charlotta Olofsson and Alexandra Allard provided excellent research assistance. Fredrik
Heyman gratefully acknowledges financial support from the Swedish Research Council for
Health, Working Life and Welfare (FORTE), the Torsten Söderbergs Stiftelse, the Johan och
Jakob Söderbergs Stiftelse and the Jan Wallander and Tom Hedelius stiftelse. Pehr-Johan
Norbäck and Lars Persson gratefully acknowledge financial support from the Jan Wallander and
Tom Hedelius stiftelse and Vinnova.

1
1. Introduction

Recent studies document a 30-year decline in various measures of entrepreneurship in the

United States (e.g., Decker et al., 2014, Hathaway and Litan, 2014). Decker et al. (2014)
argue that a likely reason for the decline in job dynamism in the U.S. has been the decline in
the share of young firms. In this paper, we argue that this has not been the case in Sweden.
Young firms have been more prominent in the Swedish business sector than in the U.S.
business sector over recent decades. Young Swedish firms (aged five years or less) accounted
for more than half of all firms, averaging approximately 55 percent per annum. During the
same period, the share of young firms in the U.S. fell from approximately 45 percent to less
than 40 percent. We also observe higher employment and job creation shares for young
Swedish firms than for young U.S. firms.
When studying firm and job dynamics more in detail, we find that the share of young
firms (aged five years or less) and their share of job creation in the business sector have been
rather stable over the period we study. In particular, we find that it is the start-up rate that has
increased in the Swedish business sector over this period. We also find that job destruction
rates have 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, which contrasts with the evidence for the U.S., where job creation
by young firms has decreased in the service sector.
What could then explain why the Swedish business sector has been more
entrepreneurial over recent decades in comparison with the U.S. business sector? And, why
do the signs of less entrepreneurial activities that we find, appear later in Sweden? In this
paper, 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. These reforms removed barriers to entry and growth for new
firms in several product and services markets and made the corporate ownership and labor
markets more flexible. This development of the business climate in Sweden contrasts the
development of the business climate in the U.S., as described by Zingales (2012), among
others. He argues that the U.S. has evolved into an economy that favors incumbents over
entrants, becoming pro-business instead of pro-market. We also discuss other potential

2
explanations and highlight that a complementary explanation is that the Information and
Communication Technology (ICT) driven structural change started later in Sweden, which
could explain the stronger job creation process in the service sector in Sweden compared to
the US.
The remainder of this paper is structured as follows: section 2 reviews the related
literature; section 3 presents the theoretical background; section 4 describes the data and
methods used for the empirical analysis; and sections 5 and 6 present the main results. In
section 7, we discuss different explanations for why Sweden appears not to have the same
decline in entrepreneurial activity in the business sector as the US has had during the last two
decades, emphasizing the economic reforms that occurred in Sweden in the 1990s that
mitigated several hurdles to entrepreneurship. Section 8 presents the study’s conclusions.

2. Related Literature

Our paper is related to the literature on entrepreneurship and public policy. Entrepreneurship
activity and its productivity crucially depend on a country’s laws and policies, as argued, for
instance, by Henrekson and Johansson (2009), Minnitti (2008) and Parker (2007). Shane
(2009) notes that most start-ups fail and argues that policymakers should focus on reducing
the incentives to create these low-probability companies. Mason and Brown (2013) use data
from Scotland to suggest appropriate policies that would result in high-growth firms (HGFs).
Román et al. (2013) note that several micro-institutional factors, such as social capital, will
interact with macro-institutional factors, such as labor market regulations, to determine the
outcome of entrepreneurship policies. Autio and Rannikko (2016) examine the growth impact
of a high-growth entrepreneurship policy initiative in Finland over a six-year period. Using an
eight-year panel that started two years before the initiative was launched, they find that the
initiative more than doubled the growth rates of the treated firms. Mthanti and Ojah (2017)
use cross-country data to compute a measure of entrepreneurial orientation (EO) using risk-
taking, innovativeness, and proactiveness as proxies for entrepreneurship. They find that their
measure of entrepreneurial orientation is positively correlated with economic growth. We add
to this literature by arguing that large-scope business reforms that are conducive to

3
entrepreneurial activity can have substantial, long-lasting effects on the entrepreneurial
activity in a country’s business sector.1
This paper further relates to the literature on entrepreneurship and job and firm
dynamics. A large literature on job and firm dynamics has shown that minor net employment
changes are often a result of substantial gross job flows (Davis et al. (1997)). Haltiwanger et
al. (2013) analyze the relationship between net job creation and firm size and age. They find
an inverse relationship between firm size and net job growth. This result is consistent with,
e.g., Neumark et al. (2011), who present evidence that small firms contribute
disproportionately to net employment growth. However, when controlling for firm age,
Haltiwanger et al. (2013) find that the relationship between firm size and job growth
disappears.2 A similar result is found in Heyman, Norbäck and Persson (2018). Age, rather
than size, is the important characteristic for job dynamics since start-ups and young
businesses create the large bulk of new jobs. Using microdata for OECD countries, Criscuolo
et al. (2014) show that young firms play an important role in overall net job creation in the
OECD regions but that there is also a great heterogeneity across countries in the extent to
which young firms are job creators. Recent studies document a 30-year decline in various
measures of entrepreneurship in the U.S. (see, e.g., Decker et al., 2014; Hathaway and Litan,
2014).
European business dynamism has been less studied. Bravo-Biosca et al. (2016)
examine cross-country differences in firm growth dynamics and how they are related to
different regulations and institutional settings. They show that less stringent employment
protection and well-functioning financial systems are linked to a more dynamic firm growth
distribution. Calvino et al. (2015) examine cross-country differences in start-up dynamics.
Based on data from the DynEmp v.2 database, they find that the large majority of start-up
firms do not grow but that a very small share of successful start-ups create a disproportionate
percentage of the new jobs. Bijnens and Konings (2018) use Belgian data to show that
Belgium’s business sector has experienced a decline in dynamism that is similar to that of the
US business sector.
We add to the above literature by showing that, in contrast to the US development, we
observe no decline or―in the case of employment shares―a smaller and later decline in

1
Henrekson and Johansson (1999) examine the evolution of the size distribution of firms over time in Sweden
for a period spanning from the late 1960s to the early 1990s. They find poor development for intermediate-sized
firms (10–199 employees), which suggests that institutional barriers to firm growth in the Swedish business
sector were high during that period.
2
See also Anyadike-Danes et al. (2015), who use cross-country data and find that adding controls for age causes
the relationship between firm size and job growth to disappear.

4
Sweden using different measures of entrepreneurial activity. We also examine the firm and
employment dynamics for specific age categories, which provides us with new knowledge on
business dynamism for young firms. In particular, we find an increase in start-up activity but
also an increase in job destruction immediately after entry. We also find that most of the job
creation by young firms occurs in the expanding service sector.
Finally, our paper also relates to the literature on regulation, institutions and
entrepreneurship. Djankov (2009) surveys the literature on the relationship between entry
regulations, on the one hand, and entrepreneurship and productivity, on the other hand. The
presented evidence indicates economically significant effects with regard to entry rates and
productivity growth. Based on a comprehensive database of European firms, Klapper et al.
(2006) find that costly regulations hamper the creation of new firms, especially in industries
that should naturally have high barriers to entry. Branstetter et al. (2014) evaluate the
consequences of the regulatory reform in Portugal, which reduced the cost of firm entry. They
report increased firm formation and employment, but mostly among ‘marginal firms’. Van
Stel et al. (2007) present cross-country evidence on the relationship between regulation and
entrepreneurship. They find that a minimum capital requirement to start a business lowers the
entrepreneurship rates, as do labor market regulations. Ciccone and Papaioannou (2007) find
that countries where it takes less time to register new businesses have higher entry in
industries that have experienced expansionary global demand and technology shifts. Finally,
Prantl (2012) exploits a natural experiment in firm entry regulations. After German
reunification, East and West Germany faced different economic conditions but were subject to
the same law that imposes a substantial mandatory standard on entrepreneurs. Their main
finding is that entry regulations suppressed both long-lived and short-lived entrants.3
We add to this literature by providing evidence that, since the implementation of
large-scope business reforms that are conducive to entrepreneurial activity in Sweden, the
Swedish business sector has become relatively more entrepreneurial in comparison with the
U.S. business sector.

3
See also Da Rin et al. (2011), Djankov et al. (2002) and Kneller and McGowan (2012) for evidence on the
effects of taxation on industry entry and exit rates. Carroll et al. (2000; 2001) examine the effects of US tax
reforms in the 1980s on the investment and hiring decisions of small businesses, finding significant effects.

5
3. Theoretical Background

Why do firm start-up dynamics and entrepreneurial activity differ over time and between
countries? The focus of this paper is that the cross-country variations in start-up dynamics and
entrepreneurial activity reflect changes in the rules and regulations for entry and firm
dynamics in different countries over time. Why then might countries regulate entry and firm
dynamics? The industrial economics literature takes as its starting point the notion that
unregulated markets may experience strong market failures, such as monopoly power and
externalities. For instance, governments may regulate entrants to make sure that consumers
buy high-quality products from “desirable” sellers; see Viscusi et al. (2005) for an overview
and Arrunada (2007) and Färnstrand Damsgaard et al. (2017) for specific contributions. In
this stream of literature, the differences and changes in regulations originate from changes in
technology and structural factors.4
The public choice theory takes as its starting point the notion that incumbents are able
to obtain regulations that create rents for themselves because they typically face lower
information and organizational costs than the dispersed consumers (Stigler, 1971). Acemoglu
(2008) presents a model where political power is in the hands of major producers who erect
significant entry barriers against new entrepreneurs. Perotti and Volpin (2005) present a
model in which incumbents seek a low level of effective investor protection to prevent
potential entrants from raising capital. The incumbents succeed because they can promise
larger contributions to the ruling politicians than the entrants due to the higher rents earned
with less competition. Regulations may also be pursued for the benefit of politicians and
bureaucrats. Politicians use regulations both to create rents and to extract them through
campaign contributions, votes, and bribes (Djankov et al., 2002). In this literature, differences
and changes in political power and outside forces that can change political power could then
explain why entry regulations change over time and between countries.
In section 7, we will argue that, during the 1990s, Sweden undertook a large set of
business sector reforms. These reforms were partly driven by a less substantial need for
regulations due to technological development ― but also by an increased understanding of the
economic cost of extensive regulation of the business sector. We will argue that these reforms
were highly influential in promoting and reigniting entrepreneurial activity in the Swedish

4
See, for instance, Acs and Audretsch (2005) and Santarelli and Vivarelli (2007) for overviews of firm dynamics
and market structure.

6
business sector and could be an explanation for why business dynamism has not decreased in
Sweden as it has in the US. In the next section, will we carefully document the firm dynamics
and job creation and job destruction processes in Sweden in detail during the period 1990–
2013, with a special focus on firm age.

4. Data Description

The empirical part of this paper is based on a comprehensive database from Statistics Sweden
covering the years 1990–2013. The database includes firm, establishment, and individual data
that are linked via unique identifiers. Detailed information on firms, establishments, and
employees enables us to thoroughly analyze questions regarding firm dynamics.
An important issue when studying firm and job dynamics is how to credibly follow
firms over time. Using organization numbers can be problematic because such numbers can
change for various reasons. For instance, internal firm reorganizations sometimes result in
changes in legal registration numbers, even though the employment composition remains the
same. This issue complicates the identification of entrants, expanding firms, contracting firms,
and shutdowns. To more reliably follow firms over time, we use additional data from
Statistics Sweden (FAD by the Swedish acronym). In each year, a firm is identified by both
its legal registration number as well as a survival code. Based on the employment composition
in t compared to that in t-1, firms are categorized as either remaining, new, or shutdown. The
criteria for each survival type are described in detail by Andersson and Arvidson (2001).
Typically, if the majority of workers remain, the firm is categorized as remaining. Each firm
is assigned a unique FAD identifier. If a firm is assigned a new legal registration number but
is considered remaining according to the FAD methodology, its FAD identifier does not
change.
Thus, in this paper, we follow firms based on their FAD identifier, not their legal
registration number. These data make it possible to identify new firm entries and firm exits,
which means that we can analyze employment changes in (i) completely new units, (ii)
continuing units and (iii) exiting units.5 In our analysis, we make use of the codes in the FAD
data that indicate if firms are completely new or are permanently exiting. By means of these
codes, employment changes should not reflect M&A activity since these activities should give
rise to changes that are not in accordance with (i) – (iii) above.

5
See Andersson and Arvidsson (2011) for details on the FAD data.

7
 The FAD database also contains information on the registration year. Each firm is then
divided into a firm age category corresponding to its FAD age, not its legal age. Firms
registered in 1986 or before all have 1986 as their registration year, meaning that we are
compelled to restrict our sample by the number of firm age categories included. For instance,
it is not until the year 2000 and onward that we can safely say that a firm is actually 14 years
old or older. To ensure that the results as reliable as possible, we also limit our analysis to
firms with at least three employees in a given year.6

5. Business dynamism in Sweden: A comparison with the US

Business dynamism and the rate of business start-ups have been declining in recent decades in
the U.S., as shown by Decker et al. (2014). In this section, we first highlight the differences
and similarities between Sweden and the U.S. with respect to business dynamism. We then
undertake a more in-depth analysis of how business dynamism has evolved in Sweden over
the past two decades.

Share of Activity by Young Firms

Decker et al. (2014) provide evidence for a declining share of activity by young firms
(defined as five years old or less) in the U.S. Figure 1, which is constructed from their data,
shows that the share of young firms and their respective employment and job creation
statistics have declined over the last two decades (for which we have comparable Swedish
microdata). Decker et al. (2014) also show that this decline is secular going back to the early
1980s.
Historically, the start-up rate has been declining in Sweden as well. Figure 2, taken
from Braunerhjelm and Carlsson (1993), shows that the number of new manufacturing firms,
as a share of total firms, decreased by half from approximately 1950 to 1990. Going back to
as far as the 1920s, the decline is even more apparent. Our microdata do not go back further
than 1990 and are not entirely comparable with the data in Braunerhjelm and Carlsson (1993).
However, as shown below, our data, which, as previously mentioned, cover the last two
decades, show fewer signs of a declining share of entrants. This evolution thus seems to have
been halted in the early 1990s.

6
The results are robust to using all firms instead of only firms with at least three employees. The only exception
is results on the skewness of the firm growth distribution. This is discussed below in the sub-section on post-
entry dynamics.

8
 -- Figures 1 and 2 about here --

Figure 3 shows the development in Sweden over the period from 1990 to 2013 for the
same measures as in Figure 1 (the share of young firms and their respective employment and
job creation) but for different definitions of a young firm, which range from young firms
being defined as operating for five years or less as in Decker et al. (2014) to defining young
firms as start-ups (completely new firms).7
Generally, young firms appear to be more prominent in the Swedish business sector
than in the business sector in the U.S. We can see that in recent decades, young Swedish firms
aged five years or less accounted for more than half of all firms, steadily averaging
approximately 55 percent per annum (top left panel). In the same period, the share of young
firms in the U.S. fell from approximately 45 percent to less than 40 percent. There is a similar
pattern for the job creation share, i.e., higher levels for young Swedish firms and more
consistent over time, which is unlike the pattern in the U.S., where the share of job creation in
young firms has continuously fallen since the 1980s. 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 downward trend in the share of employment for young firms, with the
exception of that for start-ups (bottom-right panel).

-- Figure 3 about here --

Examining the contribution of different age groups among young firms in more detail
reveals some interesting patterns. In Figure 4, we again graph the share of firms and the share
of employment pertaining to start-ups (age zero), but then break up the age groups in Figure 3
into firms that have been in the market for one, two, three, four or five years. While the levels
between the different age groups vary, each group’s share of firms is rather stable over time.
Employment shares show more heterogeneity; while the employment share for start-ups is
again very stable over time, firms who are in their first year after entry have seen their
employment shares decline significantly after the new millennium. The remaining age groups
show a similar but less pronounced pattern.

7
Note that we are using the same concept of job growth as in the methodology developed by Davis et al. (1997).

9
 -- Figure 4 about here --

Why do start-ups hold a stable share of employment while very young firms attract a
somewhat smaller share of employment over time? Figure 5 depicts job creation and job
destruction shares for firms of age 0, 1 and 5. The share of job creation in start-ups (age 0)
increases over time, except for the beginning of the period marking the aftermath of the
Swedish crises at the beginning of the 1990s. For firms aged one year, we note that while job
creation slightly declines over time, there is a lasting increase in job destruction, where the job
destruction in each year exceeds job creation. As the firm age increases, the churning in terms
of job creation and job destruction tapers off, becoming more similar across age groups. This
is clear in the figure for firms that have been in the market for five years.

-- Figure 5 about here --

What is then the source of the observed increase in job destruction among very young
firms after entry? Figure 6 shows the annual and employment-weighted firm entry and firm
exit rates for firms of age 0, 1 and 5. Figure 6 shows no clear trends in the annual exit rates
across age groups. Figure 6 also explores and depicts the loss of employment or job
destruction from firm exits (employment-weighed). Job losses due to exits in the first year
after entry appear to be very important for explaining the decline in employment shares for
young firms shown in Figure 3. In particular, firms with a firm age of one year are associated
with a strong increase in employment losses from exits. When firms grow older, this pattern
becomes less apparent. This can be seen in the figure for firms that have been in the market
for five years. Finally, Figure 7 shows employment losses (job destruction) in continuing
firms. After an initial peak generated by the financial crises in Sweden that occurred in the
early 1990s, job losses in continuing firms show no clear trends.
The results in Figure 6 and 7 indicate that early exits appear to be the main driver of
job destruction for very young firms. The fact that the overall propensity to exit appears not to
have changed, while job destruction in very young firms seems to have increased over time,
suggests that early exits appear to be biased towards larger entrants. This pattern is consistent
with the theory proposed in Färnstrand Damsgaard et al. (2017), where more risky entry
strategies (large scale entry) are associated with a higher failure rate.

-- Figure 6 and 7 about here –

10
In summary, a major feature of the dynamism in the Swedish business sector stems from an
increasing number of start-ups and an increasing importance of early exit (biased towards
larger units).8 We now proceed to examine dynamism in broader sectors.

Sector analysis

In what sectors do young firms create new jobs? To this end, we can extend the
analysis of Swedish firms by separating the effects by sector (manufacturing, service, or
other).9 Due to space limitations, we now focus on firms aged five years or less, start-ups
(new firms, i.e., firms at age 0) and very young firms (age 1).
Decker et al. (2014) report falling activity for young firms in the service sector. The
pattern is the opposite in Sweden, at least with respect to employment and job creation. As
shown in Figure 8, the activity of young firms is the largest in the service sector. The shares
of employment, firms, job creation and job destruction of young firms are notably much
higher in services than in manufacturing or other industries.
Our earlier findings on the dynamics in the entire business sector also appear at the
sector level. Employment shares for young firms decline after the millennium in both the
service and manufacturing sectors. Other dynamics patterns appear to be driven mainly by the
service sector. For instance, job creation from start-ups stem from the service sector, and job
destruction becomes increasingly important in the service sector, which is consistent with the
decline in the share of employment of young firms in the service sector.

-- Figure 8 about here --

At the aggregate level, we noted that the job destruction associated with exits appeared
to be the main reason for job losses after entry. We have also explored if this is the case at the
sectoral level also (graphs omitted due to space constraints). Our results indicate that exits one
year after entry increase over time in all sectors. While the steepness in the increase tends to
be largest in the manufacturing sector, exits after one year increase from a much higher level

8
Note that from the construction of our data with so-called FAD codes (which tracks the individuals working in
a firm or a work place) we would argue that M&As are likely not the reason for the pattern of increased early
exits of larger entrants. In our analysis, we make use of codes in the FAD data that indicate if firms are
completely new or are permanently exiting. Therefore, due to the use of these codes, employment changes
should not reflect M&A activity.
9
In order to understand what type of jobs young firms have been created we have also analyzed job dynamics in
terms of workers’ educational background. Results are available upon request.

11
in the service sector. In 2013, which is the last year in our data, approximately every third
firm in the service sector exited after only one year in business. At older ages, exit rates
become more stable in the service sector while they appear to increase over time in the two
other sectors.

Post-Entry Job Dynamism

Start-ups and young firms play an important role in U.S. job creation. Evidence from
the U.S. suggests that there is a distinct negative relationship between firm age and
employment growth within each age group. Looking at survivors, 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 U.S. Net job creation among
Swedish firms seems to be rather stable for firms aged one to seven years old, unlike the U.S.,
where the rate drops at a higher pace. As Figure 9 illustrates, it is not until Swedish firms
grow older than eight years of age that we observe significantly lower rates of net job
creation. Apart from the less pronounced trend, we also observe much lower levels of job
turnover in Sweden among the very youngest firms, except for the first year after entry

-- Figure 9 about here --

As for exit dynamics, the trends are similar, but the levels are again lower in Sweden. Both in
the U.S. and Sweden, younger firms have a substantially higher exit rate than mature firms,
and job destruction by exiting firms is higher in the U.S. than in Sweden—again with the
exception of the first year after entry, where job losses from exits in Sweden are remarkably
high. Thus, we find evidence that young Swedish firms also exhibit a relatively strong “up or
out” dynamic.
We continue the analysis by examining the arithmetic means of employment growth
at the firm level and for each age category. This exercise reveals large heterogeneity among
firms. In Figure 10, we illustrate the 90th, 50th, and 10th percentiles of job growth for
surviving firms by firm age for all employees. As is clear from the figure, the median growth
is constant over the firm age distribution. Comparing the outliers, i.e., employment growth in
the 90th and 10th percentiles, we see that the Swedish outliers are not as extreme and do not
vary quite as much. In unreported results, we compare the median to the mean. As in the U.S.,
they are both decreasing with firm age. However, in Sweden, the median growth is higher

12
than the mean growth—this is in contrast to the U.S., where the mean is consistently larger
than the median.

-- Figure 10 about here –

To shed more light on the post-entry dynamics of young firms, we follow Decker et
al. (2016) and investigate the firm growth rate distribution. Decker et al. (2016) find that the
skewness in the firm growth distribution has declined in the US. In Figure 11, we present
different measures of the skewness of the employment growth distribution in Sweden. The top
panel depicts the 90–10 differential. This is the difference between the 90th and the 10th
percentile of the employment-weighted distribution of firm employment growth rates. The top
panel depicts this difference for young firms (less than or equal to five years) and mature
firms (older than five years). While the skewness declined slightly in the 1990s, this trend
stopped after the millennium shift. The middle panel depicts the 90‒50 differential and the
50‒10 differential. Since we find that the growth rate at the medium is very close to zero, the
90‒50 differential will pick up the development for the 90th percentile, while that 50‒10
differentials will pick up the development for the 10th percentile. We note that the 50‒10
differential is falling before the millennium shift for mature firms. A potential explanation for
this is the impact of the reforms in Sweden in the early 1990s, where many mature laggard
firms were forced out of the market.10 Otherwise, we find a rather stable pattern both for
young and mature firms. Overall, we find lower growth differentials for young firms in
Sweden compared to the US. For mature firms, the levels are quite similar. However, when
including firms with one or two employees also, the pattern becomes more like the pattern
found in the US, that is, decreased skewness. Finally, in the bottom panel, we show that the
growth rate for high growth firms (90th percentile of the employment-weighted distribution) is
also stable over time.

-- Figure 11 about here --

10
One interpretation for the pattern in the middle panel in Figure 14 is that mature firms with low growth (10th
percentile) were forced out of the market, or were contracting during the restructuring process, in the years after
the reforms in the early 1990s. Hence, the 50-10 differential should decrease during this period. At the same
time, high growth mature firms likely benefitted from the reforms and, thus, the 90-50 ratio should increase. The
combination of these movements implies a stable 90-10 differential (top panel), since the median growth in the
50th percentile has near zero growth.

13
6. Is it Age Rather than Size that Matters for Dynamism?

In our comparison of business dynamism in the U.S. and Sweden, we have focused on firm
age. However, the strong correlation between how long a firm has been in business and firm
size makes it difficult to distinguish between a firm size and a firm age effect regarding
employment growth. Very large firms have, on average, been in business for quite a long
time. Hence, observing a positive firm age and job creation relationship might in fact be a
positive firm size–job creation relationship, or vice versa. Therefore, is it age or size that
matters for Swedish dynamism? 11 In this section, we thus investigate how firm age (measured
by the number of years since start up) and firm size (measured by the number of employees)
are related to each other in terms of their role in business dynamism and employment growth
at the firm level.
As an illustration, Table 1 depicts net job creation as a function of firm age and size in
terms of the number of employees between the years 2000 and 2013. As expected, there is
positive net job creation in almost all age-size groups. Nevertheless, most of the new
employment is created in small, young firms. While small start-ups create most of the
employment, small firms that are 4 to 9 years of age experience the largest decline in
employment, which is a sign of learning and experimenting in the “up or out” process.
Looking across the diagonal, we observe significant employment creation in middle-aged,
middle-sized firms. Looking across sectors (not shown), most of this net employment creation
occurs in the service sector.
Finally, old-large firms experience a decline in employment, which is consistent with
rapid technological development and globalization, with automation and international and
domestic outsourcing shifting employment away from these firms. In summary, the overriding
message of Table 1 is the fact that small start-ups create most of the new jobs in Sweden.
Next, we will document this result in more detail using regression analysis.

-- Table 1 about here --

11
Haltiwanger et al. (2013) examine job dynamics in the U.S. business sector and find that once they control for
firm age, there is no systematic relationship between firm size and growth.

14
Effects of Firm Size and Firm Age

To further investigate how firm size and firm age correlate with employment growth at the
firm level, we estimate the following regression equation:

𝑦𝑓𝑖𝑗𝑡 = 𝛼 + ∑𝑖=1,2 𝛽𝑖 × 𝐴𝑔𝑒𝑐𝑙𝑎𝑠𝑠𝑖 + ∑𝑗=1,2,3 𝛾𝑗 × 𝑆𝑖𝑧𝑒𝑐𝑙𝑎𝑠𝑠𝑗 + 𝜏𝑡 + 𝜀𝑓𝑖𝑗𝑡 (1)

where f indexes firms, i indexes age classes, j size classes, and t years. The dependent variable
y represents different measures of gross job flows measured in rates (net job creation; gross
job creation; gross job destruction). For simplicity, we include three size categories. Firms are
defined as either small (less than 50 employees), medium (50 to 249 employees), or large
(250 employees or more). For age, we use two groups. Firms aged five years or less are called
young (in accordance with the classification in Decker et al., 2014), while firms older than
five years are called old. We exclude young-large firms, as they might be affected by
measurement error. Our estimates cover the years 2000–2013 and are based on approximately
1.4 million observations.12
Columns 1–3 in Table 2 report estimates for when the age characteristics are ignored.
We observe a clear inverse relationship between firm size and job creation. Small firms are
associated with a 21.3 percentage point higher gross job creation rate than large firms. The
effect for medium firms is weaker, although it is still approximately 3.4 percentage points
higher than that for large firms. For net job creation, we observe a similar pattern—small
firms are associated with 2.5 percentage points of higher growth than large firms, and the
estimate for medium firms is again lower, by 1.5 percentage points.

-- Table 2 about here --

Controlling for age characteristics, however, the estimates change dramatically, as

shown in columns 4–6. The estimates for small- and medium-sized firms are now negative,
which suggests that they are associated with a few percentage points lower (4.2 and 1.7,
respectively) net job creation rates than large firms. The “young” dummy is, however,
positive. Compared to the reference group, i.e., old firms, young firms are associated with
13.1 percentage points higher job growth.

12
See also OECD (2014) for a cross-country analysis of job dynamics.

15
 The results from equation (1) include separate controls for the age and size
characteristics of firms. We continue the analysis by running similar regressions; however, we
now use interaction terms between the different age-size groups. This approach allows us to
assess the role of firm age within size classes as well as the role of firm size within age
groups. The specification is as follows:

𝑦𝑓𝑖𝑗𝑡 = 𝛼 + ∑𝑖𝑗 𝛿𝑖𝑗 𝐴𝑔𝑒𝑐𝑙𝑎𝑠𝑠𝑖 × 𝑆𝑖𝑧𝑒𝑐𝑙𝑎𝑠𝑠𝑗 + 𝜏𝑡 + 𝜐𝑓𝑖𝑗𝑡 (2)

Table 3 reports the estimates for equation (2), where old-large firms are the omitted
category. All groups have higher gross job creation rates than old-large firms, as shown in
column 1. Young-small and young-medium firms exhibit the largest differential. Column 2,
however, shows that old-large firms destroy relatively fewer jobs than all of the other groups
of firms. The most interesting results, i.e., those concerning net job creation, are reported in
column 3. We note that the group consisting of young-small firms accounts for the largest
growth rate, making this group the most dynamic. Young-small firms are associated with
approximately 9.1 percentage points higher net growth rates than old-large firms. Young-
medium firms and old-medium firms also have higher net growth rates than old-large firms,
but, on the contrary, old-small firms exhibit significantly lower net growth rates.

-- Table 3 about here --

Note again that we follow the methodology of Davis et al. (1997), meaning that job
creation originates with either start-ups or expanding firms. Job destruction, on the other
hand, originates with either shutdowns or contracting firms. A possible explanation for the
greater dynamism among young firms, as demonstrated above, is the positive contribution of
entry. Hence, we repeat equations (1) and (2) but exclude entrants from the sample. As shown
in Tables 4 and 5, the estimates drop drastically for young firms, suggesting that the estimates
in Tables 2 and 3 indeed are driven by entrants. In fact, young firms now appear to have
approximately 30 percentage points lower net growth than old firms. We also observe a
change in sign for the young-small and young-medium firms. With entrants excluded, old-
large firms do, in fact, exhibit a greater growth rate than firms in the young-small and young-
medium groups.

-- Tables 4 and 5 about here --

16
 From the regression exercise above, we conclude that the main determinant in
business dynamism is firm age, not firm size. Entrants and young firms seem to play a major
role in job creation. Similar results can also be obtained when we divide the firms into five
age categories (aged 0–1 years, 2–3 years, 4–9 years, 10–13 years, or 14 years and older) and
when analyzing differences across all firm age categories (results available upon request).
Thus, the firm size distribution in Sweden is not an important factor in explaining why
Sweden has avoided the same decline in entrepreneurship as the US in recent decades. It is
firm age and not firm size that matters for Swedish dynamism.

7. Decline in Entrepreneurship: Why in the US but not in Sweden?

Why does entrepreneurial activity differ over time and between countries? There are several
possible sources for such differences. In this section, we will discuss some possible
explanations for the observed over-time differences in entrepreneurial activity between
Sweden and the US.

Differences in reform patterns?

A major difference between Sweden and the US is that a large reform package was
launched in Sweden during the early 1990s. During the 1970s and 1980s, Sweden was lagging
behind in productivity growth while experiencing high inflation and large budget deficits, and
in the early 1990s, Sweden experienced a severe economic crisis. The 1991–1994 period was
characterized by a substantial decline in GDP and increasing unemployment. Previously,
during the 1980s, discussions about how to reform the Swedish welfare state had become
increasingly intensive. The center-right government that came into power in 1991 was
seemingly quick to implement an economic policy based on extensive deregulation in
response to the 1990 crisis. Moreover, in 1993, the so-called Lindbeck Commission launched
a large number of proposals to improve the efficiency of the Swedish economy (SOU
1993:16). Some of these proposals also concerned the business sector.13
Next, we describe crucial reforms that were implemented in the 1990s. We rely on
detailed descriptions of the Swedish business sector and the policy reforms that affected firms

13
See Lindbeck et al. (1994).

17
in Sweden, as described by, e.g., Heyman, Norbäck and Persson (2015, 2018), and the
references therein.

Corporate Taxation Reforms in Sweden

Importantly, until 1991, the Swedish tax system favored large firms and institutional
ownership (e.g., pension funds and insurance companies). The 1991 tax reform and
subsequent minor reforms considerably levelled the playing field for different combinations
of owners and sources of financing (Davis and Henrekson, 2000 and Edquist and Henrekson,
2013). The reforms in the 1990s generated a tax system that was far more positive for
individuals who wished to start, develop, and act as controlling firm owners compared with
the situation in the 1970s and 1980s.
However, only examining the (marginal) tax rates to judge a tax system’s effect on
firm performance is insufficient. Evaluating a tax system’s effect on corporate capital
investment is a complicated task. Many aspects, such as the project type and the financing
form, need to be considered. Here, the devil is in the details. A generally accepted method of
evaluating a capital tax system is to calculate the marginal effective tax rate (METR) on
capital based on the method originally presented by King and Fullerton (1984).
Devereux et al. (2002) use a simplified version of this method to compare the marginal
effective corporate taxes for a number of EU countries, Japan and the US in 1982 and 2001.
They show that Sweden had the second highest rate in 1982, i.e., at approximately 53 percent,
but the second lowest rate in 2001, i.e., at approximately 22 percent. Hence, corporate taxes
were reduced from very high levels to low levels during the reforms of the 1990s.
Using the King and Fullerton (1984) approach, Stenkula et al. (2015) examine the
METR on capital for a longer period for Sweden. This study provides further evidence of
Sweden’s high METR for external capital (not retained earnings or institutional capital) in the
1980s (approximately 100 percent) and its significant reduction in the 1991 tax reform. These
corporate and capital tax reforms created opportunities for firm development, particularly for
the growth of new small firms and firm formation.

The Deregulation of Product Markets in Sweden

The center-right government that came to power in 1991 was seemingly intent on
implementing an economic policy based on extensive deregulation in response to the
country’s economic crisis in the 1990s. However, the possibility of making such reforms, e.g.,

18
the deregulation of the air traffic system, the electricity market, and the postal service, had
already been thoroughly investigated, and government policies had previously been outlined
in government white papers prepared by the Social Democratic government in the late 1980s
and early 1990s. Moreover, in 1993 (SOU 1993:16), the so-called Lindbeck Commission
presented a number of proposals to improve the efficiency and functioning of markets in
Sweden (Lindbeck et al., 1994). Overall, the intensity of competition increased substantially
in many Swedish product markets during the 1990s.
The OECD has long calculated an index of the “knock-on” cost that regulations in the
service and utility industries impose on manufacturing industries. In Heyman, Norbäck and
Persson (2015), it is reported that regulations on utilities and services imposed high additional
costs on manufacturing in the 1970s and 1980s, but these costs sharply decreased as Sweden
began to deregulate in the late 1980s and early 1990s. Thus, beginning in the mid-1990s, the
regulation costs in the services and utilities sectors were substantially lower in Sweden than
the average of the EU 15, and they were even lower than such costs in the US and the UK.
The OECD has also constructed indicators to measure developments in product market
regulation (PMR) across the OECD countries (Wölfl et al., 2009). Interestingly, the “barriers
to entrepreneurship” category has improved the most for Sweden. Between 1998 and 2008,
Swedish authorities made considerable improvements regarding licensing and permit systems
and communications. Moreover, simplifications of rules and procedures were made; legal
barriers were removed; and barriers to competition in network sectors and services were
reduced.
Importantly, these product market reforms made it easier for new firms to enter
industries and made it more difficult for old inefficient firms to remain in the product market.

Labor market reforms

The so-called Allard index measures the strictness of employment protections (Allard,
2005). This measure of the strictness of employment protections continued to increase in
Sweden until the beginning of the 1990s (see Heyman, Norbäck and Persson (2015)). The
average employment protection in the EU 15 countries also increased over this period, though
not to the same extent as in Sweden. The US labor market continued to maintain a low level
of employment protection.
Thus, the Swedish employment protection became substantially stricter in 1974. The
reason was that a new employment protection law (LAS) was implemented. The law

19
mandated that employees could not be fired without reasonable cause, such as abuse or a lack
of work opportunities. Insiders were also favored with respect to firing and hiring procedures
through the so-called “last in, first out” rule. Temporary contracts also became limited.
However, centralized bargaining for private-sector, blue-collar workers gradually broke down
in the 1980s, and it was replaced by uncoordinated industry-level bargaining. Intermediate
industry-level bargaining is a form of collective bargaining that should be more conducive to
wage inflation. These actions made employment protection less strict. State-owned firms
joined the employers’ organization in 1994, which weakened the political influence in wage
setting (Nycander, 2008) and made employment protection even less strict. In 1997, the
Industry Agreement was concluded. The agreement included a system that continued industry-
level bargaining but with strong informal coordination based on pattern bargaining with the
manufacturing sector to conclude initial wage agreements in a bargaining round. This system
established a norm for wage increases for others to follow. The reformed wage-bargaining
system allowed for greater individual wage flexibility (Calmfors, 2012).
A major employment protection reform that permitted staffing agencies was
implemented in 1992 (Skedinger, 2010), and the regulations concerning temporary work were
relaxed. This development created what is referred to as the dual Swedish labor market, with
strong employment protections for regular workers and weak employment protections for
temporary workers. This reform was also evident in the Allard index of employment
protection, which declined significantly. Another measure of the strictness of employment
protections from the OECD demonstrates that the strictness of employment protections
concerning temporary contracts was significantly reduced in Sweden – from a very high level
in 1985 to a very low level in 2010. However, the strictness of the employment protections
concerning regular contracts remained at a relatively high level over the same period.
In summary, these reforms removed barriers to entry and growth for new firms and
made the labor market more flexible. The reforms of the tax system created a system that
made it far more favorable to start and develop firms compared to the situation in the 1970s
and 1980s. Product market reforms were undertaken that not only made it easier for new firms
to enter the industries but also made it more difficult for inefficient firms to remain in the
product market. Labor market reforms were also undertaken, which increased opportunities
for firms to better invest and reward human capital and for new, growing, productive firms to
hire skilled labor.
This development of a new business climate in Sweden since the early 1990s seems to
contrast with the development of the business climate in the U.S. as described by, among

20
others, Zingales (2012). First, the US business sector did not become as regulated as the
Swedish business sector in the 1970s and 1980s. Moreover, Zingales (2012) argues that, in the
last decades, the U.S. has evolved into an economy that favors incumbents over entrants,
becoming “pro-business” instead of “pro-market”. Moreover, De Loecker and Eeckhout
(2017), who examines the evolution of U.S. mark-ups from 1950, find, among other things,
that the decline in job flows is driven by the rise of market power.

Differences in technological development?

An alternative explanation for why Sweden has not experienced the same slowdown
of entrepreneurial activity is that the technological shift over recent decades has stimulated
entrepreneurship in Sweden but not in the U.S. In particular, one could argue that
technological development typically first occurs in the U.S. and then in Sweden, which would
imply that the Swedish business sector, as a result, will be less entrepreneurial in the future.
This explanation seems consistent with the high entrepreneurial activity we document in the
Swedish service sector in comparison to the US service sector. Thus, it might be that the US
has been early in switching into a more service sector dominated economy and that Sweden
soon will follow the US in facing a slowdown in entrepreneurial activity.
However, Sweden has been in the forefront of the ICT-driven industrial
reorganization. For instance, according to EU’s “Digital and society index”, Sweden ranks as
number two, following Denmark, in Europe.14 In the World Economic Forum’s “The
Networked Readiness Index”, Sweden ranks as number three after Singapore and Finland15.
Thus, the later ICT development in Sweden does not seem to be the main explanation of the
difference in the development in business dynamism in Sweden and in the US.

Differences in changes in the share of immigrants?

Another potential explanation is that immigrants are more likely to start new
businesses and that immigration flows have decreased in the U.S., while immigration flows
have been increasing in Sweden during the period under study. However, this explanation is
unlikely, since immigrants in Sweden are underrepresented as owners of firms with
employees and overrepresented as self-employed.16

14
European Commission (2016), The Digital Economy and Society Index, https://ec.europa.eu/digital-single-
market/en/desi.
15
World Economic Forum (2015), The global information technology report 2015,
http://www3.weforum.org/docs/WEF_Global_IT_Report_2015.pdf.
16
Only approximately 13% of all firms with 0-49 employees have a foreign-born owner (Tillväxtanalys (2012)),
whereas Alden (2014) shows that most immigrant groups are overrepresented as self-employed.

21
Differences in changes in age structures or in population growth?
Yet another explanation is that the older societies become, the less entrepreneurial
they are, and the age structure in the U.S. may have changed in such a way that the U.S. is
now less prone to entrepreneurship compared to Sweden. Lazear et al. (2014) analyze the
relationship between demographics and entrepreneurship using a human capital framework.
Based on cross-country data from the Global Entrepreneurship Monitor, they find that a one-
standard-deviation decrease in the median age of a country increases the rate of
entrepreneurship by 2.5 percentage points, which is approximately 40 percent of the mean
rate.
This seems to be a potential part of the explanation for the slowdown of the
entrepreneurial activity in the US. Nevertheless, we see no evidence of a decrease in
entrepreneurial activity is Sweden despite an aging population. The median age has increased
in Sweden from 38.4 years in 1990 to 40.7 years in 2010. The corresponding numbers for the
US are 32.9 in year 1990 and 37.2 in 2010.17
What role do population and working age population growth play for the evolution of
start-up rates in Sweden? Karahan et al. (2016) suggested that demographic shifts in the
growth rates of the labor supply are important factors behind the declining start-up rates in the
US. They show empirically that the decline in growth rates of the labor supply explains a
significant share of the startup-up decline in the US during the last 30 years. While a detailed
empirical analysis is beyond the scope of this paper, we have correlated Swedish start-up rates
during the period of 1996-2013 with different measures of the labor supply. Using time-series
data on growth in the Swedish labor force and the Swedish working age population, we find
no clear correlations.18 However, it should be noted that the growth in the Swedish labor force
and the Swedish working age population has been stable over the period compared to the US,
where these factors have varied more. Thus, Sweden’s high start-up rates are likely supported
by a stable population and labor force growth. A stable population and labor force growth
might, however, mask substantial regional differences.

17
https://www.statista.com/statistics/375459/average-age-of-the-population-in-sweden/ and
https://www.statista.com/statistics/241494/median-age-of-the-us-population/.
18
Analysis available on request from the authors.

22
8. Summary and Conclusions

We take as our point of departure the more than 30-year decline in entrepreneurship
dynamism and activity by young firms documented in the U.S. (see, e.g., Decker et al., 2014
and Hathaway and Litan, 2014). Using detailed Swedish employer-employee data over the
period of 1990–2013, we observe no decline or a smaller and later decline in Sweden for
different measures of entrepreneurial activity. In particular, we find an increase in the start-up
activity. We also find an increase in job destruction rates in the year after entry (firms of age
one). This increase in job destruction starts to dominate the increase in start-up activity around
2005, giving rise to a small decrease in employment share for young firms in the business
sector. Moreover, we find that most of the job creation by young firms occurs in the
expanding service sector.
What can explain this different pattern? We have argued that a likely explanation for
the high entrepreneurial activity in the Swedish business sector was a set of policy reforms
that were initiated in the late 1980s but mostly implemented in the 1990s. Sweden
transformed from having one of the most regulated, static business sectors in the developed
world in the late 1980s to one of the more pro-entrepreneurial business regulation systems in
the developed world. A potential weakness in the dynamics of the Swedish business sector is
the relatively high job destruction rates of firms of age 1 and 2. Moreover, it should also be
noted that the reform pace in business policy has substantially slowed in the Swedish business
sector the last decades, suggesting that the recent trend of less dynamism in the Swedish
business sector might be here to stay.
A welfare analysis of the observed pattern would be of interest. If the different pace of
adjustment in Sweden and the US is due to differences in burdensome regulations and
institutions, this has potentially large adverse consequences for long-run growth in jobs and
productivity. On the other hand, if the driving force is the implementation of ICT, then the
welfare consequences will be different. Research along these lines seems warranted.
In our analysis of Sweden, we focused on the total effect of the extensive and wide-
ranging business sector reforms. An interesting avenue for future research would be to
examine in greater detail the effects of the different reforms on entrepreneurial activity in the
business sector. Moreover, it would be interesting to see whether such reforms in isolation
might be sufficient to improve entrepreneurial activity or if it is a package that is needed to
create a strong, lasting effect.

23
References

Acemoglu, Daron (2008). “Oligarchic Versus Democratic Societies”. Journal of the European
Economic Association 6(1), 1–44.

Acs, Zoltan J. and David B. Audretsch (2005). “Entrepreneurship, Innovation and

Technological Change”. Foundations and Trends in Entrepreneurship 1(4), 149–195.

Aldén, Lina (2014). Företagande bland utrikes födda – en översikt och en policydiskussion.

Allard, Gayle J. (2005). “Measuring job security over time: in search of a historical indicator
for EPL (Employment Protection Legislation)”. Working Paper 05-17. Instituto de
Empresa, Madrid.

Andersson, Jan and Gunnar Arvidson (2001). Företagens och arbetsställenas dynamik (FAD).
Örebro: Statistics Sweden.

Anyadike-Danes, Michael, Carl-Magnus Bjuggren, Sandra Gottschalk, Werner Hölzl, Dan

Johansson, Mika Maliranta and Anja Myrann (2015). “An international cohort comparison
of size effects on job growth”. Small Business Economics, 44(4), 821-844.

Arrunada, Benito (2007). “Pitfalls to avoid when measuring institutions: Is Doing Business
damaging business?”. Journal of Comparative Economics 35(4), 729–47.

Autio, Erkko and Heikki Rannikko (2016). “Retaining winners: Can policy boost high-
growth entrepreneurship?”. Research Policy 45(1), 42-55.

Bijnens, Gert and Jozef Konings (2018). “Declining Business Dynamism”. CEPR Discussion
Paper No. DP12615.

Branstetter, Lee, Francisco Lima and Lowell J. Taylor (2014). “Do Entry Regulations Deter
Entrepreneurship and Job Creation? Evidence from Recent Reforms in Portugal”.
Economic Journal 124(577), 805–832.

Braunerhjelm, Pontus and Bo Carlsson (1993). “Entreprenörskap, småföretag och industriell

förnyelse 1968–1991”. Research Institute of Industrial Economics (IFN) Working Paper
No. 372. IFN, Stockholm.

Bravo-Biosca, Albert, Chiara Criscuolo and Carlo Menon (2016). “What drives the dynamics
of business growth?”. Economic Policy 31(88), 703-742.

24
Calmfors, Lars (2012). “Sweden: from macroeconomic failure to macroeconomic success”.
Working Paper 3790, Category 6: Fiscal Policy, Macroeconomics and Growth. CESifo,
Munich.

Calvino, Flavino, Chaiara Criscuolo and Carlo Menon (2015). “Cross-country evidence on
start-up dynamics”. OECD Science, Technology and Industry Working Papers 2015 (6).

Carroll, Robert , Douglas Holtz-Eakin, Mark Rider and Harvey S. Rosen (2000). “Income
taxes and entrepreneurs' use of labor”. Journal of Labor Economics, 18(2), 324–351.

Carroll, Robert , Douglas Holtz-Eakin, Mark Rider and Harvey S. Rosen (2001). “Personal
income taxes and the growth of small firms”. Tax Policy and the Economy, 15, 121–148.

Criscuolo, Chiara, Peter N. Gal and Carlo Menon (2014). “The Dynamics of Employment
Growth: New Evidence from 18 Countries”. CEP Discussion Paper, Vol. 1274.

Ciccone, Antonio and Elias Papaioannou (2007). “Red Tape and Delayed Entry”. Journal of
the European Economic Association 5(2–3), 444–458.

Da Rin, Marco, Marina Di Giacomo, and Alessandro Sembenelli (2011). “Entrepreneurship,

firm entry, and the taxation of corporate income: evidence from Europe”. Journal of Public
Economics, 95(9), 1048–1066.

Davis, Steven J., John Haltiwanger and Scott Schuh (1997). Job Creation and Destruction.
Cambridge, MA: MIT Press.

Davis, Steven J. and Magnus Henrekson (2000). ‘Wage-setting institutions as industrial

policy’, NBER Working Paper No. 7502.

Decker, Ryan, John Haltiwanger, Ron Jarmin and Javier Miranda (2014). “The Role of
Entrepreneurship in US Job Creation and Economic Dynamism”. Journal of Economic
Perspectives 28(3), 3–24.

Decker, Ryan A., John Haltiwanger, Ron S. Jarmin and Jacier Miranda (2016). “Where has all
the skewness gone? The decline in high-growth (young) firms in the US”. European
Economic Review 86, 4-23.

De Loecker, Jan and Jan Eeckhout (2017). “The Rise of Market Power and the
Macroeconomic Implications”. Working Paper No. 23687, National Bureau of Economic
Research.

Devereux, Michael P., Rachel Griffith, and Alexander Klemm (2002). “Corporate Income
Tax Reforms and Tax Competition”. Economic Policy 17(35), 449–495.

25
Djankov, Simeon, Rafael La Porta, Florencio Lopez-de-Silanes and Andrei Shleifer (2002).
“The Regulation of Entry”. Quarterly Journal of Economics 117(1), 1–37.

Djankov, Simeon (2009). “The Regulation of Entry: A Survey”. World Bank Research
Observer 24(2), 183–203.

Edquist, Harald, and Magnus Henrekson (2013). “Product market reforms and incentives to
innovate in Sweden”. In M. Choupres, and H. Edquist, eds., What can we learn from
economic reforms in Greece and Sweden? Brussels: European Liberal Forum.

European Commission (2016). The Digital Economy and Society Index.

Färnstrand Damsgaard, Erika, Per Hjertstrand, Pehr-Johan Norbäck, Lars Persson and Helder
Vasconcelos (2017). “Why Entrepreneurs Choose Risky R&D Projects–But Still Not
Risky Enough”. The Economic Journal 127(605), F164-F199.

Haltiwanger, John, Ron S. Jarmin and Javier Miranda (2013). “Who Creates Jobs? Small
Versus Large Versus Young”. The Review of Economics and Statistics 95(2), 347–361.

Hathaway, Ian and Robert Litan (2014). “The Other Aging of America: The Increasing
Dominance of Older Firms”. Brookings Institution Report July 2014. Washington, D.C.:
Brookings Institution.

Henrekson, Magnus and Dan Johansson (1999). “Institutional Effects on the Evolution of the
Size Distribution of Firms”. Small Business Economics 12(1), 11–23.

Henrekson, Magnus and Dan Johansson (2009). “Competencies and Institutions Fostering
High-Growth Firms”. Foundations and Trends in Entrepreneurship 5(1), 1–82.

Heyman, Fredrik, Pehr-Johan Norbäck and Lars Persson (2015). “The Turnaround of Swedish
Economy: Lessons from Large Business Sector Reforms”. Research Institute of Industrial
Economics (IFN) Policy Paper No. 73, forthcoming World Bank Research Observer.

Heyman, Fredrik, Pehr-Johan Norbäck and Lars Persson (2018). “Who Creates Jobs and Who
Creates Productivity? Small versus Large versus Young versus Old”. Economics Letters 164
(March), 50–57.

Karahan, Fatih, Benjamin Pugsley, and Aysegul Sahin (2016). “Demographic Origins of the
Startup Deficit.” Technical Report. New York Fed, mimeo.

King, Mervyn A. and Don Fullerton (1984). The Taxation of Income from Capital: A
Comparative Study of the United States, the United Kingdom, Sweden and West Germany.
Chicago: University of Chicago Press.

26
Klapper, Leora, Luc Laeven and Raghuram Rajan (2006). “Entry Regulation as a Barrier to
Entrepreneurship”. Journal of Financial Economics 82(3), 591–629.

Kneller, Richard, and Danny McGowan (2012). “Tax policy and firm entry and exit
dynamics: evidence from OECD countries”. Working Papers No. 12006, Bangor Business
School, Prifysgol Bangor University (Cymru / Wales).

Lazear, Edward, James Liang and Hui Jackie Wang (2014). “Demographics and
Entrepreneurship”. Stanford Graduate School of Business Working Paper No. 4320.
Stanford, CA: Stanford Graduate School of Business.

Lindbeck, Assar, Per Molander, Torsten Persson, Olof Petersson, Agnar Sandmo, Birgitta
Swedenborg and Niels Thygesen (1994). Turning Around Sweden. Cambridge, MA: MIT
Press.

Mason, Colin and Ross Brown (2013). “Creating Good Public Policy to Support High-Growth
Firms”. Small Business Economics 40(2), 211–225.

Minniti, Maria (2008). “The Role of Government Policy on Entrepreneurial Activity:

Productive, Unproductive, or Destructive?”. Entrepreneurship Theory and Practice 32(5),
779–790.

Mthanti, Thanti and Kalu Ojah (2017). “Entrepreneurial orientation (EO): Measurement and
policy implications of entrepreneurship at the macroeconomic level”. Research Policy
46(4), 724-739.

Neumark, David, Brandon Wall and Junfu Zhang (2011). “Do Small Businesses Create More
Jobs? New Evidence for the United States from the National Establishment Time Series”.
The Review of Economics and Statistics 93(1), 16-29.

Nycander, Svante (2008). Makten över arbetsmarknaden: ett perspektiv på Sveriges 1900-tal,
SNS Publishing, Stockholm.

OECD (2014). The Dynamics of Employment Growth. Paris: OECD.

Parker, Simon C. (2007). “Law and the Economics of Entrepreneurship”. Comparative Labor
Law and Policy Journal 28(4), 695–716.

Perotti, Enrico and Paolo Volpin (2005). “Lobbying On Entry”. CEPR Discussion Paper No.
4519. London: CEPR.

Prantl, Susanne (2012). “The Impact of Firm Entry Regulation on Long-Living Entrants”.
Small Business Economics 39(1), 61–76.

27
Román, Concepción, Emilio Congregado and José María Millán (2013). “Start-Up Incentives:
Entrepreneurship Policy or Active Labour Market Programme?”. Journal of Business
Venturing 28(1), 151–175.

Santarelli, Enrico and Marco Vivarelli (2007). “Entrepreneurship and the Process of Firms’
Entry, Survival and Growth”. Industry and Corporate Change 16(3), 455–488.

Shane, Scott (2009). “Why Encouraging More People to Become Entrepreneurs is Bad Public
Policy”. Small Business Economics 33(2), 141–149.

Skedinger, Per (2010). Employment protection legislation. Evolution, effects, winners and
losers. Northampton, MA: Edward Elgar.

SOU (1993:16). Nya villkor för ekonomi och politik. Stockholm: Allmänna förlaget.

Stenkula, Mikael, Dan Johansson, and Gunnar Du Rietz (2015). “Capital income taxation of
Swedish households, 1862 to 2010”. Scandinavian Economic History Review 63 (2) : 154–
77.

Stigler, George J. (1971). “The Theory of Economic Regulation”. Bell Journal of Economics
and Management Science 2(1), 3–21.

Tillväxtanalys (2012). Statligt stöd till näringslivet 2011. Report 2012:05. Östersund: Growth
Analysis.

Van Stel, André, David J. Storey and A. Roy Thurik (2007). “The Effect of Business
Regulations on Nascent and Young Business Entrepreneurship”. Small Business
Economics 28(2), 171–186.

Viscusi, W. Kip, John M. Vernon and Joseph E. Harington (2005). Economics of Regulation
and Antitrust. Cambridge, MA: MIT Press.

World Economic Forum (2015). The Global Information Technology Report 2015.

Wölfl, A., I. Wanner, T. Kozluk, and G. Nicoletti (2009). “Ten years of product market
reform in OECD countries: insights from a revised PMR indicator”. Working Paper 695.
OECD Economic Department, OECD Publishing, Paris.

Zingales, Luigi (2012). A Capitalism for the People. New York, NY: Basic Books.

28
Figures

45 18
43 16
41 14
39
12
37

Percent
10
Percent

35
8
33
6
31
29 Share of firms, (US, left) 4
Share of job creation (US, left) 2
27
Share of employment (US, right)
25 0

Fig. 1. Declining Share of Activity from Young Firms in the US (firms aged 5 years or less). Firms with at least
one employee.
Source: Decker et al. (2014).

Fig. 2. New Firms in the Swedish Manufacturing Sector (% of stock).

Source: Braunerhjelm and Carlsson (1993)

29
Fig. 3. Share of Activity from Young Firms in Sweden.

30
Fig. 4. Share of Activity from Young Firms in Sweden. Share of employment and share of firms by firm age.

Fig. 5. Share of Activity from Young Firms in Sweden. Share of job creation and job destruction by firm age.

31
Fig. 6. Annual & employment-weighted annual firm entry and firm exit rates by firm age.

Fig.7. Annual job destruction rates for continuing firms by firm age.

32
Fig. 8. Share of Activity from Young Firms in Sweden by Sector, 1996–2013.

33
Fig. 9. Up or Out Dynamics for Young Firms in Sweden, 2002–2013.
Note: In the right panel, we exclude firms aged 1 year or less for aesthetic reasons—the rates are much higher
than for the remaining groups.

Fig. 10. Net Employment Growth and Growth Rates for Surviving Firms.

34
Fig. 11. Firm growth rate differentials and firm growth rates for high-growth firms separated by young and
mature firms (employment-weighted).
Note: The 90-10, 90-50 and 50-10 differentials are the difference between 90th and 10th, 90th and 50th and 50th
and 10th percentile, respectively. Further, HP trends are used with parameter set to 100. Firm age for young
firms is 5 years or less, while mature firms have a firm age of 6 years or more.

35
Tables

Table 1
Net Job Creation by Age and Size, Percentages, 2000–2013.
Size
Age 3-9 10-49 50-199 200-499 >499

0-1 31,42 11,89 9,86 9,14 6,93

2-3 -15,19 0,54 1,87 1,49 0,22
4-8 -5,27 1,84 1,96 1,60 0,08
9-13 -3,00 1,57 1,21 0,81 0,30
>13 -2,48 0,41 0,29 -0,41 -0,18

Table 2
Gross Job Creation and Destruction and Net Growth Rate of Firms by Size and Age Groups, 2000–2013.
(1) (2) (3) (4) (5) (6)
Gross Job Gross Job Net Growth Gross Job Gross Job Net Growth
Creation Ratio Destruction Ratio Rate Creation Ratio Destruction Ratio Rate

Young 0.367*** 0.236*** 0.131***

(0.000888) (0.000895) (0.00141)
Small 0.213*** 0.188*** 0.0249*** 0.0266*** 0.0682*** −0.0417***
(0.000933) (0.000944) (0.00145) (0.000808) (0.000899) (0.00132)
Medium 0.0387*** 0.0239*** 0.0148*** −0.0512*** −0.0340*** −0.0172***
(0.00115) (0.00113) (0.00173) (0.00121) (0.00116) (0.00175)

Observations 1,484,760 1,484,760 1,484,760 1,484,760 1,484,760 1,484,760

R-squared 0.006 0.005 0.001 0.108 0.050 0.006
Year FE Yes Yes Yes Yes Yes Yes
Notes: The reference categories are Old (age>5) and Large (size 250+). Young-large firms are dropped, as this group is likely to be affected
by measurement error. Robust standard errors are in parentheses. *** p<0.01, ** p<0.05, * p<0.1.

Table 3
Gross Job Creation and Destruction and Net Growth Rate of Firms by Size-Age Groups, 2000–2013.
(1) (2) (3)
Gross Job Creation Ratio Gross Job Destruction Ratio Net Growth Rate

Young × Small 0.398*** 0.308*** 0.0909***

(0.00120) (0.00117) (0.00187)
Young × Medium 0.120*** 0.0782*** 0.0421***
(0.00319) (0.00295) (0.00462)
Old × Small 0.0218*** 0.0652*** −0.0434***
(0.000806) (0.000900) (0.00132)
Old × Medium 0.0122*** 0.00627*** 0.00596***
(0.000892) (0.000926) (0.00141)

Observations 1,484,760 1,484,760 1,484,760

R-squared 0.110 0.050 0.006
Year FE Yes Yes Yes
Notes: The reference category is Old-Large (age>5; size 250+). Young-large firms are dropped, as this
group is likely to be affected by measurement error. Robust standard errors are in parentheses.
*** p<0.01, ** p<0.05, * p<0.1.

36
Table 4
Gross Job Creation and Destruction and Net Growth Rate of Firms by Size and Age Groups Excluding Entrants,
2000–2013.
(1) (2) (3) (4) (5) (6)
Gross Job Creation Gross Job Net Gross Job Creation Gross Job Net
Ratio Destruction Ratio Growth Ratio Destruction Ratio Growth
Rate Rate

Young 0.0140*** 0.314*** −0.300***

(0.000216) (0.00104) (0.00112)
Small 0.0277*** 0.213*** −0.185*** 0.0213*** 0.0695*** −0.0482***
(0.000789) (0.000970) (0.00137) (0.000793) (0.000904) (0.00132)
Medium 0.0188*** 0.0246*** −0.0057*** 0.0155*** −0.0497*** 0.0652***
(0.000876) (0.00113) (0.00156) (0.000875) (0.00120) (0.00162)

Observations 1,349,168 1,349,168 1,349,168 1,349,168 1,349,168 1,349,168

R-squared 0.002 0.006 0.005 0.005 0.078 0.063
Year FE Yes Yes Yes Yes Yes Yes
Notes: See Table 2.

Table 5
Gross Job Creation and Destruction and Net Growth Rate of Firms by Size-Age Groups Excluding Entrants,
2000–2013.
(1) (2) (3)
Gross Job Creation Ratio Gross Job Destruction Ratio Net Growth Rate

Young × Small 0.0350*** 0.388*** −0.353***

(0.000801) (0.00129) (0.00164)
Young × Medium 0.0405*** 0.0835*** −0.0431***
(0.00130) (0.00306) (0.00354)
Old × Small 0.0216*** 0.0652*** −0.0436***
(0.000793) (0.000905) (0.00132)
Old × Medium 0.0121*** 0.00631*** 0.00582***
(0.000878) (0.000931) (0.00142)

Observations 1,349,168 1,349,168 1,349,168

R-squared 0.006 0.080 0.064
Year FE Yes Yes Yes
Notes: See Table 3.

37