Catching the Train of the Knowledge Era

Workshop on Improving Statistics on SME's and
Entrepreneurship, Paris, 17-19 September 2003
Differences in entry and exit in European countries –
findings and impact of methodological differences
NICOLA BRANDT
Economic Analysis and Statistics Division
Organisation for Economic Co-operation and Development
1 1
This presentation



Describes the findings from some recent OECD work
on firm entry and exit, survival and growth using new
cross-country data from Eurostat.
Discusses some of the methodological difficulties
associated with international business demographics
data.
Draws conclusions and provides an outlook on OECD
work linking firm dynamics to economic performance
2 2
Results of entry rate regressions
Constant
country fixed
effects:
Finland
Belgium
Netherlands
Sweden
Spain
Portugal
Italy
UK
Output gap
Adjusted R
2
industry- and
year-dummies
7.77***
-2.73***
-3.37***
-0.89**
-3.86***
-0.55
-2.23***
-0.82**
-2.11***
0.59
also country-specific
ICT-effects
6.14***
ICT fixed
effects:
-0.94*
-7.91***
-1.49**
-6.14***
0.15
-5.76***
-0.93
-12.42***
1.60**
-0.94
-0.09
-10.50***
1.36*
-6.55***
0.28
-10.24***
0.39**
0.63
Indicates significance at the ***1 %, ** 5 %, * 10 % level
Denmark is the reference
3 3
The cross-country variation in entry rates
is mainly due to ICT-related industries
Country-specific ICT-industries effects
Sweden
Portugal
UK
Finland
Italy
Belgium
Netherlands
Spain
-15
-10
-5
0
4 4
…which are particularly dynamic
Fixed effects for selected industries
Telecommunications***
Data base activities***
Hardware consultancy***
Labour recruitment ***
Management consulting***
Market research***
Office mach. & computers***
Advertising***
Transport equipment***
Chemicals
0
2
4
6
8
10
12
14
16
18
20
%
Indicates significance at the ***1 %, ** 5 %, * 10 % level
Manufacture of food & beverages is the reference
5 5
Entry and exit are highly correlated across
industries
Cross-industry correlation of entry and exit
Correlation
Denmark
0.81***
Finland
0.65***
Belgium
0.52***
Netherlands 0.65***
Sweden
0.39***
Spain
0.50***
Portugal
0.60***
Italy
0.78***
UK
0.70***
T-Statistic
10.98
6.65
4.36
6.77
2.95
2.88
5.93
9.48
7.77
Obs.
64
64
54
64
49
27
64
61
64
6 6
New firms are small - but results differ
between Eurostat and OECD data...
Average size of entering firms
Manufacturing Eurostat
Manufacturing OECD
Services OECD
Number of
employees
Services Eurostat
24
22
20
18
16
14
12
10
8
6
4
2
n
ed
e
Sw
um
Be
lg
i
ai
n
It a
ly
Sp
et
he
r
N
Po
rtu
la
nd
s
ga
l
K
U
k
ar
en
m
D
Fi
n
la
n
d
0
7 7
...because firms without employees play an
important role in both entry and exit
Entry and exit: proportion of firms with less than five employees
%
Entry 0 employees
Entry 1-4 employees
Exit 0 employees
Exit 1-4 employees
100
80
60
40
20
0
Sw eden
Finland
Denmark
Belgium
Spain
Portugal
Netherlands
UK
Italy
8 8
…few large firms enter and exit the market
Entry: proportion of firms with more than 20 employees
%
OECD Manufacturing
OECD Services
Eurostat Manufacturing
Eurostat Services
14
12
10
8
6
4
2
0
Denmark
Portugal
Finland
UK
Italy
Netherlands
Belgium
Spain
Sw eden
9 9
… especially when entry and exit are cleaned
from other demographic events
%
Manufacturing OECD
Manufacturing Eurostat
Services OECD
Services Eurostat
16
14
12
10
8
6
4
2
0
Portugal
UK
Finland
Denmark
Netherlands
Italy
Sw eden
Belgium
Spain
10 10
Results of hazard rate regressions
I
with
industry
fixed
effects
Constant
Finland
Netherlands
Sweden
Spain
Portugal
Italy
UK
2-YearDummy
GAP
Adjusted R
2
17.70***
-5.06***
-3.57***
-15.15***
-4.67***
-11.21***
-5.07***
-10.20***
4.84***
0.42
II
also with
an output
gap
variable
III
differentiated by size class*
21.44***
-6.47***
-3.76***
-18.82***
-9.46***
-13.56***
-11.05***
-14.34***
5.69***
11.60***
-5.00***
3.22***
-13.28***
-5.96***
-7.99***
-9.43***
-5.02***
5.22***
-1.83***
-1.93***
0.43
0.20
8.51***
-2.70***
-2.10***
-0.24
Size class:
0 Empl.
5-9 Empl.
10-19 Empl.
> 20 Empl.
Indicates significance at the ***1 %, ** 5 %, * 10 % level
Denmark is the reference country
1-4 employees is the reference size class
11 11
Few new firms survive, but those that do
often grow very fast
Two-year employment gains among surviving entrants
%
OECD data
Eurostat data
2.0
1.6
1.2
0.8
0.4
s
K
U
tu
ga
l
Po
r
D
en
m
ar
k
n
ed
e
Sw
Sp
ai
n
It a
ly
d
la
n
Fi
n
et
he
rla
nd
N
U
ni
te
d
St
at
es
0.0
12 12
Conclusions




Thresholds and the ability to identify genuine firm
entry and exit can affect results.
Hazard rates differ considerably across countries,
while differences in entry rates are sizeable only
among ICT-related industries.
This indicates that policy may be important
mainly for firm survival and for entry in young
and dynamic sectors.
To judge whether high firm entry, growth or
survival are desirable, the link with economic
performance has to be studied. There is ongoing
OECD work on this.
13 13