erc demography presentation. july 8th 2015
TRANSCRIPT
E: [email protected]: enterpriseresearch.ac.uk
Demography Theme Reference Group London July 2015
Getting in under the hood:firm and job dynamics in the UKbefore, during, and after the Great Recession
Michael Anyadike-Danes & Mark HartAston Business School and Enterprise Research Centre
approach & overview
approach
• most of the work using job creation and destruction accountshas focused on flows, particularly job flows: the principalinterest has been in the scale of labour market adjustment,especially the role of ’churn’ in the re-allocation of jobs
• instead we put stocks, in particular the stock of firms, centrestage and combine it with the number of jobs per firms toaccount for the evolution of jobs
overview
• using conventional job creation and destruction data wearticulate the accounts slightly differently to emphasise theconnection between ’startups’ in one year and the stock of’continuing’ firms in the years which follow: today’s startup istomorrow’s continuing firm
• we show – using UK data for 1998 to 2014 – that for most ofthe period, and particularly during the ’Great Recession’ (GR),it is changes in the stock of firms, rather than jobs per firm,which account for the bulk of the contraction and that mostof the change in the stock of firms can be traced back to a’collapse’ in startups in 2009
data sources & construction
data sources & construction
1. sourcesI Inter-Departmental Business Register
(IDBR) is a ’live’ register updated for jobsfrom HMRC (VAT and PAYE) and Business RegisterEmployment Survey
I the BSD comprises extracts from ’snapshots’ of the IDBRtaken each March (1997 to 2014)
2. constructionI focus on firm and job dynamics with firms linked year-to-year
by IDI appearance of first job ≡ birth of firmI dis-appearance of last job ≡ death of firmI private sector (defined by industrial activity)I the database which underpins this study – the Longitudinal
Business Structure Database – can be accessed by approvedresearchers through the ONS Virtual Microdata Laboratory orthe UK Data Service Secure Lab
1. dating: we have adoped the convention ofreferring to the data by the date of the snapshot,so for example the observations labelled 2014 are from theMarch 2014 snapshot
I the data on the register in March are not likely to refer activityas at March or the period end-March; to be entered onto theregister in March, they are likely to refer to some previous time
I the administrative processes underpinning the data collectionmake it likely that the period to which the data refers varies,at least to some extent, across firms – in practice the GRperiod seems clearly visible in the behaviour of the firm andjobs aggregates
2. warning: The statistical data used here is from the Office ofNational Statistics (ONS) and is Crown copyright andreproduced with the permission of the controller of HMSOand Queen’s Printer for Scotland.The use of the ONSstatistical data in this work does not imply the endorsement ofthe ONS in relation to the interpretation or analysis of thestatistical data.
the measurement framework
accounst (1)
1. firms
usual : closing ft ≡ closing f
t−1 − deaths ft + births f
t
ours : continuing ft ≡ continuing f
t−1 + births ft−1 − deaths f
t
closing ft ≡ continuing f
t + births ft
2. jobs
closing jt ≡ closing j
t−1 − deaths jt + births j
t + continuing_net jt
continuing jt ≡ continuing j
t−1+births jt−1+continuing_net j
t−deaths jt
closing jt ≡ continuing j
t + births jt
accounts (2)jobs per firm
continuing jt
continuing ft
≡continuing j
t−1 + births jt−1 + continuing_net j
t − deaths jt
continuing ft
continuing jt
continuing ft
−continuing j
t−1continuing f
t−1≡ −∆continuing f
tcontinuing f
t×
continuing jt−1
continuing ft−1
+ continuing_net jt
continuing ft
+births j
t−1births f
t−1×
birthft−1
continuing ft
− deaths jt
deaths ft
× deaths ft
continuing ft
the facts of firm & job dynamics,1998–2014
births: firms (’000) & jobs per firm (jperf), 1998–2014 (log scales)
firm
s '0
00 (
log
scal
e)
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
148
181
221
270
330
403
2.5
3
3.7
4.5
5.5
6.7
jper
f log
sca
le
firmsjperf
births: firms & jobs, (’000), 1998–2014 (log scales)
firm
s '0
00 (
log
scal
e)
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
148
181
221
270
330
403
446
545
665
812
992
1212
jobs
'000
(lo
g sc
ale)
firmsjobs
deaths: firms (’000) & jobs per firm (jperf), 1998–2014 (log scales)
firm
s '0
00 (
log
scal
e)
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
148
181
221
270
330
403
4.1
5
6
7.4
9
11
jper
f log
sca
le
firmsjperf
deaths: firms & jobs, (’000), 1998–2014 (log scales)
firm
s '0
00 (
log
scal
e)
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
148
181
221
270
330
403
735
898
1097
1339
1636
1998
jobs
'000
(lo
g sc
ale)
firmsjobs
continuing firm numbers, contributions to change (’000),
1999–2014
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
●
●● ● ●
●
●● ●
●●
●● ●
●●
firm
s '0
00
−350
−300
−250
−200
−150
−100
−50
0
50
100
150
200
250
300
350 ● births(t−1)deaths(t)
∆continuing
continuing firms (million) & jobs per firm (jperf),
1998–2014 (log scales)fir
ms
mill
ion
(log
scal
e)
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
0.99
1.1
1.21
1.34
1.48
11
12.2
13.5
14.9
16.4
jper
f (lo
g sc
ale)
firms jperf
continuing firms & jobs (million), 1998–2014 (log scales)
firm
s m
illio
n (lo
g sc
ale)
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
0.99
1.1
1.21
1.34
1.48
13.4
14.8
16.3
18
19.9
jobs
mill
ion
(log
scal
e)
firmsjobs
continuing firm jobs, contributions to change (’000), 1999–2014
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
jobs
& c
ontr
ibut
ions
'000
−10
00−
500
050
010
0015
00
firmsjperf
∆jobs
net job creation, contribution of continuing firms & births (’000),
1999–2014
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
cont
ribut
ions
to jo
bs '0
00
−1000
−750
−500
−250
0
250
500
750
1000∆birth∆cont
net jcr
what have we learned?
methodological conclusions
• the stock of jobs can be written as the stock of firmsmultiplied by average jobs per firm
• we found for births, deaths and to the stock of continuingfirms, the average jobs per firm varied relatively slowly (and inmuch the same way in all three cases)
• over the last decade and a half in the UK, variations in thenumber of firms (being born, continuing and dying) appearedto play a more active role in driving variations in job creationthan did variations in the average size of firms
• focusing on stocks rather than flows provides a more coherent,and slightly different, account of firm and job dynamics in theUK over the period since 1998
substantive conclusions
• the role of startups, both their immediateimpact on job creation in the year of their birth, and theirsubsequent lagged effect through continuing firms, emerged asa critical part of the Great Recession story. Slower job growthplayed little role in accounting for the loss of jobs, there waslittle detectable effect of increased business failure rates after2008
• the onset of the recovery was signalled by two developments.I birth numbers turned up quite dramatically – the increase
almost matched the decrease which marked the onset of theGR
I deaths dropped steeplythe contemporaneous impact of deaths, and the lagged effectof births, can be tracked into the time pattern of changes inthe stock of continuing firms. It rises: in 2012 due to the fallin deaths, in 2013 due to 2012 births