High-growth firms: Not so vital after all?

High-growth firms: Not so vitl fter ll? Sven-Olov Dunfeldt, Dniel Hlvrsson, nd On Mihescu Abstrct: High-growth firms hve received considerble interest recently since they crete most of the new jobs in the economy. The purpose of our pper is to investigte the chrcteristics of high-growth firms prior to their growth period, nd whether these chrcteristics differ cross industries. Using dt on lrge smple of limited libility firms in Sweden for the period 2007-2010, we find tht high-growth firms do not hve the chrcteristics tht we typiclly ssocite with successful firms. On the contrry, our results indicte tht high-growth firms hve low profits nd wek finncil position. This might explin why studies hve found tht high-growth firms re seldom cpble of sustining their high growth rtes in subsequent periods, nd thus question policies tht re trgeted towrds these compnies. Keywords: Entrepreneurship; Firm growth; Gzelles; High-growth firms; High-impct firms; Innovtion JEL-codes: L11; L25 1

1. Introduction Studies hve shown tht smll number of high-growth firms (HGFs) re very importnt for job cretion (Henrekson nd Johnsson, 2010; Cod et l., 2014). Nest (2009), for exmple, found tht the 6 percent fstest growing firms in the United Kingdom creted lmost 50 percent of ll jobs during 2002-2008. These firms were given the nme the vitl 6 percent to highlight their remrkble importnce for job cretion. 1 HGFs bility to crete job opportunities hs ttrcted incresing ttention from policymkers (Dunfeldt et l., 2015). Support for high-growth SMEs is, for exmple, stted s politicl objective in the Europe 2020 strtegy of the Europen Commission (Europen Commission, 2010). Some reserchers lso support the ide of trgeting potentil HGFs (e.g., Shne, 2009; Mson nd Brown, 2013), rguing tht policy should be redirected towrds firms with growth spirtions insted of strt-up firms with low survivl rtes. The shre of fst-growing innovtive firms is emphsized by the Europen Commission (2010) s n importnt indictor to mesure whether policies trgeted towrds potentil HGFs re successful. However, the ssumption tht HGFs re overrepresented in high-tech industries seems to hve little empiricl support (Henrekson nd Johnsson, 2010). On the contrry, Dunfeldt et l. (2015) found tht HGFs in Sweden were less common in R&D-intensive industries nd overrepresented in knowledge-intensive service sectors. This points towrds knowledge problem embedded in the politicl inititive to promote the growth of HGFs, suggesting tht we need more reserch on wht ctully chrcterizes the rpidly growing firms in the economy. Previous studies hve indicted tht young (Reichstein et l., 2010; Brb Nvretti et l., 2014) nd smll firms (Birch, 1979; Almus, 2010; Goedhuys nd Sleuwegen, 2010) re more likely to be chrcterized by fst growth thn older nd lrger firms. However, recent studies hve rgued tht no systemtic reltionship exists between firm growth nd firm size once ge is controlled for (Hltiwnger et l., 2013; Lwless, 2014). The importnce of firm ge for explining high-growth events is lso highlighted by Dunfeldt et l. (2014), who identified HGFs in nine different wys nd found tht the common denomintor regrdless of definition ws their reltively young ge. There is lso some evidence indicting tht HGFs do not grow through cquisitions, but re more likely to enter into llinces with other firms (Mohr et l., 2014). HGFs lso seem more likely to engge in export behvior thn non-hgfs (Hölzl nd Friesenbichler, 2007). However, most previous studies on HGFs hve not controlled for the profitbility nd the finncil strength of the firms before they entered their period of fst growth. This is troublesome, considering tht Penrose (1959) hd long go emphsized the importnce of profits for chieving long-term growth. As stted by Brännbck et l. (2009, p. 71): 1 Similr results hve been shown by other reserchers s well. Storey (1994), for exmple, found tht the 4 percent fstest growing firms in the UK contributed to 50 percent of ll jobs, nd Dunfeldt et l. (2013) indicted tht the 6 percent fstest growing firms generted 42 percent of ll jobs in Sweden during 2005-2008. 2

Being profitble clerly seems to be fr more productive, nd in the long run, better pproch to being str firm. Dvidson et l. (2009) lso found tht the bility to grow in subsequent periods is positively ssocited with the firm s profitbility. It ws found tht, if initil growth coincides with high profitbility, firms re more likely to disply growth in future periods lso. Previous studies on the chrcteristics of HGFs hve in most cses lso been bsed on economy-wide dt or dt from selective industries, such s the mnufcturing industry (Cod, 2009). This is unfortunte since we know tht there re lrge differences cross industries tht might influence the likelihood of observing high-growth events (Audretsch et l., 2004). The mnufcturing industry, for exmple, is cpitl intensive nd chrcterized by high sunk costs, which is often interpreted s sign tht smll nd young mnufcturing firms might be forced to grow fst in order to survive. On the other hnd, such scle economies re less likely to be importnt for firms in low-tech service industries, such s the ccommodtion nd food services industry. We contribute to the HGF literture by investigting whether the chrcteristics of HGFs differ cross industries fter controlling for both profits nd finncil strength. Our nlysis is bsed on comprehensive dtset, covering ll limited libility firms in Sweden during 2007-2010. We find tht HGFs re chrcterized by low profits nd low solidity prior to their growth episodes. This finding is perplexing since these chrcterfetures re typicl compred to the ones we ssocite with firms tht chieve long-term growth. Viewed in light of recent studies (Hölzl, 2014; Dunfeldt nd Hlvrsson, 2015), however, it might help to explin why HGFs re unlikely to repet their initil high growth rtes in coming periods. Policies to promote HGFs re often trgeted towrds R&D-intensive industries (Dunfeldt et l., 2015), but we find no evidence of substntil industry differences regrding the chrcteristics of HGFs prior to their growth period. As consequence, we find no support for the view tht HGFs within certin industries re more suitble to trget thn firms in other industries when designing industry policy. In the next section, we present brief theoreticl bckground on why HGFs re seen s importnt job cretors. Dt nd descriptive sttistics re presented in Section 3, nd our empiricl method is described in Section 4. Results for the full smple nd for five selected industries re then presented in Section 5. The finl section concludes the pper with summry nd discussion of our key findings. 2. Theoreticl bckground Which firms re importnt for the cretion of new jobs nd economic growth? This question hs interested reserchers nd policymkers for long time. The nswer, however, hs shifted mrkedly over the pst 100 yers. As fr bck s 1911, Schumpeter (1934/1911) emphsized the importnce of entrepreneurs nd new firms for creting economic growth nd prosperity. According to the young Schumpeter, the entrepreneur ws considered the individul force tht 3

introduced new ides in the economy, often by estblishing new firms. In his view, these young firms, most of which were lso smll, were crucil for economic development nd growth becuse they chllenged the incumbents with new technology. Incumbents tht could not keep up with the progress of these smll, innovtive firms were eventully replced process tht Schumpeter populrly clled cretive destruction. The imge is cler yet powerful, portrying young compnies s the destroyers of old, inefficient technology. The older Schumpeter (1942), on the other hnd, emphsized the importnce of scle economies for both production nd reserch nd development. In the first decdes fter World Wr II, it ws considered self-evident tht it ws the lrge nd, nturlly, the older compnies tht creted jobs nd growth. At this time, economic policy ws ttuned to the economic theories of return-to-scle production s ws reserch nd development. New nd smll businesses were viewed s inefficient. Occsionlly, they were even considered wste of resources (Glbrith, 1956, 1967). As result, economic policies were designed to trget lrge industril compnies. In very influentil report, however, Dvid Birch (1979) cme to question this view. In ccordnce with the previling view t tht time, he found tht lrge firms ccounted for the mjority of ll new jobs. However, when observing these firms over time, he found tht lrge firms lost jobs nd were replced by firms tht hd once been smll but hd grown big. Smll firms thus creted the mjority of ll jobs over time, while lrger businesses reduced their number of employees. The importnt insight nd contribution ws tht the perception tht lrge enterprises were importnt for job cretion ws bsed on sttic pproch, while the importnce of smll firms tkes precedence in dynmic nlysis. Birch s (1979) results were controversil nd criticized in severl studies (e.g., Dvis et l., 1996; Hltiwnger nd Krizn, 1999). Lter studies, however, confirmed most of his initil results (Vn Prg nd Versloot, 2008), but with one importnt ddition: most smll firms were not growing. The new jobs were insted being creted by smll number of HGFs (Birch nd Medoff, 1994; Henrekson nd Johnsson, 2010). Storey (1994) found, for exmple, tht 50 percent of the new jobs in the UK were creted by the 4 percent fstest growing compnies. In recent study, entitled The Vitl 6 per cent, Nest (2009) showed tht it ws rther the 6 percent fstest growing compnies in the UK tht ccounted for hlf of ll new jobs in the economy. The job cretion bility of HGFs hs led to suggestions tht policymkers should devote more resources to supporting these compnies, rther thn investing in strt-ups tht normlly hve no mbitions to grow or cnnot survive mrket competition (Shne, 2009; Mson nd Brown, 2013). The Europen Commission, for exmple, sttes in its strtegy documents tht more efforts should be directed towrds supporting the fstgrowing smll nd medium-sized firms (Europen Commission, 2010). The ide of supporting potentil HGFs hs, however, been criticized recently since the growth of HGFs does not seem to be sustined over time (Hölzl, 2014; Dunfeldt nd Hlvrsson, 2015) nd becuse it seems to be difficult to predict which firms will be 4

chrcterized by high growth in the future (Storey, 1994; Hölzl, 2009). Another problem with the orienttion towrds HGFs is tht it leds to policies tht re focused on R&Dintensive industries (Dunfeldt et l., 2015). However, there is little evidence tht rpidly growing firms re more common in R&D-intensive industries (Hölzl, 2009; Henrekson nd Johnsson, 2010). Dunfeldt et l. (2015), for exmple, present results indicting tht HGFs re less common in R&D-intensive industries nd rther overrepresented in the knowledge-intensive service industries. 3. Dt nd descriptive sttistics All limited libility firms in Sweden re required by lw to submit nnul reports to the Swedish Ptent nd Registrtion Office (PRV). We use dt from PAR, Swedish consulting firm, which gthered this informtion from PRV. The dt include informtion on ll figures in the nnul reports, such s profits, number of employees, industry clssifiction, nd sles. In the dtset, firms re clssified into industries ccording to the Europen Union s NACE stndrd. We use these industry clssifiction codes to select dt on surviving firms in eight different industries during 2007-2010 (Tble 1). We hve selected these industries to induce possible vrition in the likelihood of receiving policy interventions nd the degree of technologicl knowledge. While some industries, such s mnufcturing, re frequently studied (Cod, 2009) nd re of considerble interest mong politicins, industries such s hospitlity nd retil hve received less ttention. Compred to mnufcturing with reltively high R&D expenditures, hospitlity nd retil re often considered to be low-tech industries tht provide jobs for low-qulified workers nd re rrely the focus of policy interventions. A similr observtion cn be mde bout policies tht re trgeted towrds potentil HGFs, which re dominted by R&Dintensive sectors. The reltive lck of interest in retil nd hospitlity cn lso be observed in previous literture, with very few studies tht specificlly investigte the chrcteristics of fst-growing firms within these industries. 2 Our finl dtset consists of 78,937 firms ctive during 2007-2010. [Tble 1 bout here] One inescpble obstcle when investigting the chrcteristics of HGFs is tht there is no consensus definition or wy of identifying these firms (Cod et l., 2014). It is therefore necessry to mke decisions regrding the growth indictor, firm growth mesure, length of growth period, nd growth process (Delmr nd Dvidsson, 1998). Two of the most common growth indictors in the literture re sles nd number of employees (Delmr et l., 2003; Dunfeldt et l., 2014), which re known to be modestly correlted (Shepherd nd Wiklund, 2009). The indictors represent different spects of the production process even if results seem little influenced when choosing one or the other (Dunfeldt et l., 2014). While the number of employees is n input fctor (often 2 A notble exception is Dunfeldt et l. (2013), who investigted firm growth within the Swedish retil nd wholesle trde industries during 2000-2004 using quntile model. 5

considered to be qusi-fixed), sles represents firm s gross output. Thus, in using employment growth, firm growth cptures the rte of chnge of internl resources, wheres using sles growth reflects the product s or service s cceptnce in the mrket (Delmr et l., 2003). We therefore choose to pply both of these growth indictors in the pper. More specificlly, employment growth nd sles growth re clculted by: ln =ln ln, ln =ln ln, (1) where ln is the logrithmic chnge in sles from 2007 to 2010, nd ln the logrithmic chnge in the number of employees during the sme period. 3 It is worth noting here tht the logrithmic difference is one of the most frequently used mesures in the firm growth literture (Cod, 2009), with the convenient property of being symmetric for positive nd negtive growth rtes (Tornqvist et l., 1985). This mens tht rel chnges in either indictor give the sme percentge chnge, whether it is positive or negtive. As stted in eqution (1), we follow the previous literture on HGFs nd consider chnges over the course of 3 yers, more specificlly between 2007 nd 2010 (Cod et l., 2014). By voiding nnul growth rtes we cn void lot of idiosyncrsy tht exists for more nrrowly defined growth rtes. However, s Bjuggren et l. (2013) remrk, results do not seem prticulrly sensitive to which period is chosen. Finlly, it is well known tht growth cn be divided into orgnic growth or cquired growth. Most studies do not hve ccess to dt on mergers nd cquisitions nd must therefore rely on mesures of totl growth, tht is, the sum of orgnic nd cquired growth. This is drwbck since wht is interesting to investigte re the chrcteristics of nd mechnisms by which firms chieve high levels of growth by incresing output nd enhncing sles (orgnic growth), nd not by mergers nd cquisitions growth tht is generted outside the firm. Fortuntely, the PAR dtbse includes informtion on mergers nd cquisitions nd we use this informtion to exclude firms tht hve been subject to merger or cquisition. In contrst to most previous studies, we cn thus focus on orgnic growth nd lso control whether the firm ws subject to merger or cquisition before the study period. Previous studies hve indicted tht HGFs tend to be young nd smll, nd we therefore control for both firm ge (A i2014) nd firm size (R i2006, E i2006) in our s. Firm ge is mesured using informtion on the registered strt yer, nd defined s the observtion yer minus the registered strt yer. Access to registered strt yer is rre (Cod et l., 2015), nd in contrst to previous studies we therefore do not need to work with truncted or censored ge dt. Firm size is mesured using either sles or number of employees in 2006, depending on which growth indictor we use. 3 Note tht reltive growth mesures, such s the one we pply, tend to fvor smller firms, wheres bsolute growth mesures re bised towrds lrger firms (Delmr et l., 2003). 6

Returns on totl ssets (ROAi2006) nd solidity (Si2006) re included s independent vribles in order to investigte the profitbility nd finncil strength of HGFs prior to their growth period. We use returns on totl ssets s our profit mesure since it is not ffected by the type of finncing (Libby et l., 2011), nd lthough multiple profitbility mesures hve been used in the literture (Richrd et l., 2009), ROA is the one tht seems most commonly used (Dvidsson et l., 2009; Steffens et l., 2009). Finncil strength of the firms is mesured by their solidity, tht is, the percentge shre of equity out of totl cpitl in 2006. This shows the reltive proportion of equity tht is used to finnce firm s ssets nd indictes the firm s solvency in the long term. The higher the proportion of equity tht finnces the compny, the higher nd better the solidity nd the lower the finncil risk. Contrry to previous studies, we re lso ble to control whether the firm hs been subject to merger nd cquisition before the study period (C i2006). As we wnt to control whether the initil conditions of the firms influence their subsequent growth rtes, ll of our control vribles (except firm ge, which is monotonic trnsformtion) re mesured in 2006. Finlly, we lso include municiplity-specific fixed effects ( nd ) nd industry-specific fixed effects 4 ( nd ) to control for regionl nd industry time-invrint heterogeneity tht might ffect firm growth rtes. Descriptive sttistics of the vribles tht we include in our empiricl nlysis re presented in Tble 2. [Tble 2 bout here] The descriptive sttistics indicte tht the verge firm in the smple is round 18 yers old, hs eight employees, 5.4 percent in returns on totl ssets, nd solidity of 34.7. Sles re on verge 16,509,000 SEK (corresponds in 31 August 2015 to 1,737,000 EUR), while the medin firm s sles re 2,585,000 SEK (272,000 EUR). Note lso tht more observtions re missing when we investigte employment growth. The reson is tht some firms with positive sles do not hve ny employees, which mens tht they will be excluded when we tke the log difference to clculte firm growth rtes. 4. Empiricl method It is well known tht firm growth tends to follow the tent-shped Lplce distribution (Stnley et l., 1996; Bottzzi nd Secchi, 2006; Bottzzi et l., 2011), with most firms not growing nd few firms growing very fst. The fmilir tent-shped distribution is lso evident in our dtset for both employment growth nd sles growth (Figure 1). This violtes the stndrd lest-squres ssumption of normlly distributed error terms, nd mens tht OLS estimtion becomes less ttrctive. It is of little interest to estimte the verge effect when the verge firm is chrcterized by very mrginl 4 Industry-specific fixed effects re used only for estimtions conducted on the ggregted dtset including ll eight industries considered for nlysis. 7

growth rtes. Neither do we wnt to consider HGFs s outliers, s OLS would, since our focus is on investigting wht determines the growth rtes of these fst-growing firms. [Figure 1 bout here] Medin, which ssumes the error terms to be Lplce distributed, becomes more suitble in our cse. Following Fotopoulos nd Louri (2004), Cod nd Ro (2008), nd Reichstein et l. (2010), we therefore estimte quntile model to investigte wht chrcterizes firms cross the entire growth rte distribution, including the fstest growing firms. The estimted equtions cn be written: = + + + + + + + + = + + + + + + + + (2) where is the ge of firm i; is number of employees in 2006; is firm sles in 2006; is returns on totl ssets in 2006; is the solidity of firm i in 2006; nd is n indictor vrible tht equls one if the compny hs been subject to merger or cquisition in 2006. We lso include industry-specific fixed effects 5, nd, to ccount for time-invrint differences cross industries tht might influence firm growth rtes, nd municiplity-specific fixed effects, nd, to ccount for time-invrint heterogeneity cross municiplities in Sweden; 6 nd re constnts, nd - nd - re prmeters to be estimted. Finlly, nd re rndom error terms. 5. Wht chrcterizes high-growth firms? 5.1 Results, ll firms The results when eqution (2) is estimted for ll firms re presented in Tble 3 (sles growth) nd Tble 4 (employment growth). In order to evlute the ppropriteness of using OLS, we present results both from n OLS model nd from quntile model. The OLS results indicte tht older firms re chrcterized by fster sles growth thn younger firms (Tble 3). However, the quntile results revel tht this result is driven by firms with mrginl growth rtes nd tht sles growth is not significntly relted to firm ge for the fstest growing firms. This shows the importnce of not relying on OLS when investigting determinnts of firm growth rtes. No sttisticlly significnt 5 Idem. 6 We hve lso estimted model without municiplity- nd/or industry-specific (in the cse of ll firms) fixed effects, nd most results remin qulittively similr. The results re vilble in Appendices B nd C. 8

reltionship is observed between firm size nd firm growth in the OLS results, wheres firm growth seems to be negtively relted to firm size for firms in the 0.80 quntile. Profitbility nd solidity re two firm-specific vribles tht hve seldom been controlled for in previous firm growth studies, lthough some reserchers hve rgued tht they might be importnt determinnts for firm growth rtes (Dvidsson et l., 2009; Steffens et l., 2009). The OLS results in Tble 3 indicte tht sles growth is positively relted to initil profits for the verge firm. However, sles HGFs re less likely to be profitble, suggesting tht they tend to grow before chieving profits. Solidity is lso negtively relted to rpid growth, implying tht sles HGFs strt their growth period from wek finncil position. Finlly, sles HGFs re less likely to hve prticipted in merger nd cquisition prior to their growth period. In Tble 4, the corresponding results for growth in number of employees re presented. Firm ge is now positively relted to firm growth for the mjority of the firms in growth rte distribution, but not for the fstest growing firms. According to the results, employment HGFs re younger thn firms tht re growing more slowly. Firm size, on the other hnd, does not seem to influence employment growth for the fstest growing firms. Neither is it sttisticlly significnt when estimting n OLS model. This implies tht firm ge is more importnt determinnt of employment growth thn firm size, supporting Hltiwnger et l. s (2003) findings. The results lso indicte tht employment HGFs re more likely to be chrcterized by low profitbility nd low degree of solidity. Firms tht re growing fst in terms of number of employees thus seem to hve lower initil profits nd less finncil strength thn firms tht re growing more slowly. Note finlly tht the results for the 0.5 quntile re difficult to interpret since in most of the firms in this quntile the number of employees does not chnge, which mens tht the vrition in the dependent vrible is very low. [Tble 3 bout here] [Tble 4 bout here] 5.2 Industry differences mong high-growth firms The nlysis hs so fr been focused on the chrcteristics of HGFs for ll eight industries presented in Tble 1. In order to test whether the chrcteristics of HGFs differ cross industries, we hve lso estimted eqution (2) seprtely for the following five industries: (i) Mnufcturing (NACE-code 25); (ii) Construction (NACE-codes 42 nd 43); (iii) Retil (NACE-code 47); (iv) Hospitlity (NACE-codes 55 nd 56); nd (v) Computer progrmming (NACE-codes 62 nd 63). The results for the fstest growing firms re presented in Tble 5 (sles growth) nd Tble 6 (employment growth), while the results for the other growth quntiles re presented in Appendix A. [Tble 5 bout here] 9

[Tble 6 bout here] The results in Tbles 5 nd 6 indicte tht HGFs in most cses shre the sme chrcteristics cross industries. Only smll differences cross the industries under study cn be observed, nd in most cses they confirm the ggregted results presented in Tbles 3 nd 4. Thus, despite considerble industry differences regrding scle economies, cpitl intensity, nd humn cpitl, HGFs re chrcterized by low initil profits nd hve wek finncil position prior to the growth period. The only differences tht we cn observe re tht firms in the computer industry, tht is, more knowledgeintensive services, seem to be older nd hve tken prt in merger before the study period. 6. Conclusions The purpose of this pper hs been to nlyze the chrcteristics of HGFs in Sweden during 2007-2010, nd to investigte whether they differ for firms ctive in different industries. This question is of importnce since policymkers hve strted to design policies tht re trgeted towrds potentil HGFs in R&D-intensive industries (Dunfeldt et l., 2015). HGFs were found to be chrcterized by low profits prior to their growth period, which is troublesome since profits seem importnt in predicting future growth (Dvidsson et l., 2009; Steffens et l., 2009). The lck of profits would be less of problem if HGFs were finncilly strong, but we found tht HGFs hd lso grown from wek finncil position. This implies tht HGFs do not hve the chrcteristics tht we typiclly ssocite with firms tht re ble to become successful in the long run. We believe tht our results might help to explin why recent studies (e.g., Hölzl, 2014; Dunfeldt nd Hlvrsson, 2015) hve found tht HGFs re one-hit wonders, unlikely to sustin their high growth rtes in subsequent periods. Our study thus dds reson for concern regrding the efficcy of policies tht re trgeted towrds HGFs. Policies tht re trgeted towrds potentil HGFs hve in generl been focused on R&D-intensive industries. However, we did not find ny lrge differences mong HGFs belonging to five industries tht re very different in terms of, for exmple, cpitl intensity, minimum efficient scle, nd shre of educted workers. Thus, our results do not seem to be driven by industry-specific differences. Our results question the current fscintion with HGFs, nd suggest tht policies trgeted towrds these firms re unlikely to be successful. Mybe policymkers should insted focus on improving the generl conditions for firm growth. As noted by Bornhäll et l. (2015), mny profitble firms might choose to grow if the conditions for firm growth become more fvorble. This implies tht politicins should try to remove growth brriers for ll firms insted of trying to pick winners or design policies trgeted towrds those firms tht hve high historicl growth rtes. We believe, therefore, tht there is need for more reserch on the conditions for firm growth, nd on the kinds of policies tht cn promote firm growth tht is sustinble in the long run. 10

Acknowledgments Reserch funding from the R&D Fund of the Swedish Tourism nd Hospitlity Industry (BFUF) is grtefully cknowledged. A specil thnks to Krl Wennberg. 11

REFERENCES Almus, M. (2010). Wht chrcterizes fst-growing firm? Applied Economics, 34(12), 1497-1508. Audretsch, D. B., Klomp, L., Sntrelli, E., nd Thurik, A. R. (2004). Gibrt's lw: re the services different?. Review of Industril Orgniztion, 24(3), 301-324. Brb Nvretti, G., Cstellni, D., nd Pieri, F. (2014). Age nd firm growth: evidence from three Europen countries. Smll Business Economics, 43(4), 823-837. Birch, D. L. (1979), The Job Genertion Process. MIT Progrm on Neighborhood nd Regionl Chnge. Msschusetts Institute of Technology: Cmbridge, MA. Birch, D. L., nd Medoff, J. (1994). Gzelles, in: L. C. Solmon & A. R. Levenson (Eds), Lbor Mrkets, Employment Policy nd Job Cretion, 159-167, Boulder, CO: Westview. Bjuggren, C. M., Dunfeldt, S. O., nd Johnsson, D. (2013). High-growth firms nd fmily ownership. Journl of Smll Business & Entrepreneurship, 26(4), 365-385. Bornhäll, A., Dunfeldt, S-O., nd Rudholm, N. (2015). Sleeping gzelles: The unseen job cretors?. In Corbett, A.C., Ktz, J.A., nd McKelvie, A. Entrepreneuril growth: Individul, firm, nd region, Advnces in Entrepreneurship, Firm Emergence, nd Growth, 17, 161-185. Bottzzi G., nd Secchi A. (2006). Explining the distribution of firm growth rtes. Rnd Journl of Economics, 37(2), 235-256. Bottzzi, G., Cod, A., Jcoby, N., & Secchi, A. (2011). Corporte growth nd industril dynmics: Evidence from French mnufcturing. Applied Economics, 43(1), 103-116. Brännbck, M., Crsrud, A. L., nd Kiviluoto, N. (2014). Understnding the Myth of High Growth Firms. New York: Springer. Cod, A. (2009). The Growth of Firms: A Survey of Theories nd Empiricl Evidence. Edwrd Elgr, Cheltenhm: United Kingdom. Cod, A., Dunfeldt, S-O., Hölzl, W., Johnsson, D., nd Nightingle, P. (2014). High-growth firms: Introduction to the specil issue. Industril nd Corporte Chnge, 23(1), 91-112. Cod, A., Dunfeldt, S-O., nd Hlvrsson, D. (2015). Bursting into life: Firm growth nd growth persistence by ge. HUI Working Pper 112, HUI Reserch: Stockholm. Cod, A., nd Ro, R. (2008). Innovtion nd firm growth in high-tech sectors: A quntile pproch. Reserch Policy, 37(4), 633 648. 12

Dunfeldt, S-O., Lng, Å., Mcuchov, Z., nd Rudholm, N. (2013). Firm growth in the retil nd wholesle trde sectors: Evidence from Sweden, The Service Industries Journl, 33, 1193-1205. Dunfeldt, S-O., Elert, N., nd Johnsson, D. (2014). Economic contribution of highgrowth firms: Do policy implictions depend on the choice of growth indictor? Journl of Industry, Competition nd Trde, 14(3), 337-365. Dunfeldt, S-O., Elert, N., nd Johnsson, D. (2015). Are high-growth firms overrepresented in high-tech industries? Forthcoming Industril nd Corporte Chnge. Dunfeldt, S-O., nd Hlvrsson, D. (2015). Are High-Growth Firms One-Hit Wonders? Evidence from Sweden. Smll Business Economics, 44(2), 361-383. Dvidsson, P., Steffens., P., nd Fitzsimmons, J. (2009). Growing profitble or growing from profits? Putting the horse in front of the crt? Journl of Business Venturing, 24(4), 388-406. Dvis, S. J., Hltiwnger, J., & Schuh S. (1996). Smll business nd job cretion: dissecting the myth nd ressessing the fcts. Smll Business Economics, 8(4), 297-315. Delmr, F., Dvidsson., P., nd Grtner, W.B. (2003). Arriving t the high-growth firm. Journl of Business Venturing, 18(2), 189 216. Delmr, F., nd Dvidsson, P. (1998). A txonomy of high-growth firms. 399-413 in: Reynolds, P.D., et l, (eds.), Frontiers of Entrepreneurship Reserch. Bbson College, Wellesley, MA. Europen Commission (2010). Europe 2020: A strtegy for smrt, sustinble nd inclusive growth: Communiction from the Commission. Reserch report. Fotopoulos, G., nd Louri, H. (2004). Firm growth nd FDI: Are multintionls stimulting locl industril development?. Journl of Industry, Competition nd Trde, 4(3), 163 189. Glbrith, J.K. (1956). Americn cpitlism: The theory of counterviling power. Boston, MA: Houghton Mifflin. Glbrith, J.K. (1967). The new industril stte. Princeton, NJ: Princeton University Press. Goedhuys, M. nd Sleuwegen, L. (2010). High-growth entrepreneuril firms in Afric: quntile pproch. Smll Business Economics, 34(1), 31-51. Hltiwnger, J., nd Krizn, C. J. (1999). Smll business nd job cretion in the United Sttes: The role of new nd young businesses. In Are smll firms importnt? Their role nd impct (s. 79-97). Springer US. 13

Hltiwnger J., Jrmin R.S., nd Mirnd, J. (2013). Who cretes jobs? Smll versus Lrge versus Young. Review of Economics nd Sttistics, 95(2), 347 361. Henrekson M., nd Johnsson, D. (2010). Gzelles s job cretors: survey nd interprettion of the evidence. Smll Business Economics, 35(2), 227-244. Hölzl, W., nd Friesenbichler, K. (2007). Are gzelles more innovtive thn other firms?. WIFO, Austri (mimeo). Hölzl, W. (2009). Is the R&D behviour of fst-growing SMEs different? Evidence from CIS III dt for 16 countries. Smll Business Economics, 33(1), 59-75. Hölzl, W. (2014). Persistence, survivl, nd growth: closer look t 20 yers of fstgrowing firms in Austri. Industril nd Corporte Chnge, 23(1), 199-231. Lwless, M., (2014). Age or size? Contributions to job cretion. Smll Business Economics, 42(4), 815 830. Libby, R., Libby, P., nd Short, D.G. (2011). Finncil Accounting, 7 ed, McGrw- Hill/Irwin: New York. Mson, C., nd Brown, R. (2013). Creting good public policy to support high-growth firms. Smll Business Economics, 40(2), 211-225. Mohr, V., Grnsey, E., nd Theyel, G. (2014). The role of llinces in the erly development of high-growth firms. Industril nd Corporte Chnge, 23(1), 233-259. NESTA (2009), The Vitl 6 Per Cent. NESTA: London, UK. Penrose, E. (1959). A Theory of the Growth of the Firm. 4th edition. Oxford University Press: New York. Reichstein, T., Dhl, M.S., Ebersberger, B., nd Jensen, M.B. (2010). The devil dwells in the tils: A quntile pproch to firm growth. Journl of Evolutionry Economics, 20(2), 219-231. Richrd, P.J., Devinney,T.T., Yip, G., nd Johnson, G. (2009). Mesuring orgniztionl performnce: Towrds best methodologicl prctice. Journl of Mngement, 34(3), 718-804. Schumpeter, J. (1911/1934). The Theory of Economic Development. Hrvrd University Press, Cmbridge. Schumpeter, J. (1942). Cpitlism, Socilism nd Democrcy. Hrper, New York. Shne, S. (2009). Why encourging more people to become entrepreneurs is bd public policy. Smll Business Economics, 33(2), 141-149. 14

Shepherd, D., nd Wiklund, J. (2009). Are we compring pples with pples or pples with ornges? Appropriteness of knowledge ccumultion cross growth studies. Entrepreneurship Theory nd Prctice, 33(1), 105-123 Stnley, M.H.R., Amrl, L.A.N., Buldyrev, S.V., Hvlin, S., Leschhorn, H., Mss, P., Slinger, M.A., nd Stnley, H.E. (1996). Scling behvior in the growth of compnies. Nture, 379(6568), 804-806. Storey, D. (1994). Understnding the Smll Business Sector. Routledge, London. Tornqvist L., Vrti P., nd Vrti Y.O. (1985). How should reltive chnges be mesured? Americn Sttisticin 39(1), 43-46. Vn Prg, M., nd Versloot, P. (2008). The economic benefits nd costs of entrepreneurship: A review of the reserch. Foundtions nd Trends in Entrepreneurship, 4(2), 65-154. 15

Tble 1. Industries nd number of firms included in the nlysis NACE Industry Frequency Percent 43 Specilized construction ctivities 24,961 31.62 47 Retil trde, except motor vehicles nd 24,767 31.38 Motorcycles 62 Computer progrmming 11,472 14.53 56 Food nd beverge serving ctivities 8,654 10.96 25 Mnufcture of fbricted metl products, 6,373 8.07 except mchinery nd equipment 55 Accommodtion 1,157 1.47 42 Civil engineering 826 1.05 63 Informtion service ctivities 727 0.92 Totl 78,937 100 Tble 2. Descriptive sttistics Vrible Obs. Men Medin Std. Dev. Min Mx 50,729 0.0689-0.0244 1.0881-10.350 11.397 45,169-0.0171 0 0.468-4.736 5.193 78,937 18.293 15 13.073 4 116 60,292 16,509 2,585 250,401 0 33,700,000 59,387 8.274 3 79.233 0 10,856 58,596 5.417 6.2 44.999-999 999 58,591 34.674 37 59.877-999 622 78,937 0.212 0 0.409 0 1 16

Tble 3. Estimtion results, sles growth ( ), 2007-2010, municiplity nd industry fixed effects. All firms. OLS nd quntile s. Stndrd errors in prentheses. Vrible OLS Quntile 0.0035*** 0.0060*** 0.0043*** 0.0024*** 0.0005-0.0008 (0.0004) (0.000) (0.000) (0.000) (0.000) (0.001) -0.0001 0.00005 0.0000*** -0.0001** -0.0001*** -0.0002 (0.000) (0.000) (0.000) (0.000) (0.000) (0.001) 0.00005*** 0.0002*** 0.0001*** 0.000001-0.0001** -0.0002** -0.0003*** 0.0001 0.0000-0.0001-0.0006*** -0.0011*** (0.0001) (0.000) (0.000) (0.000) (0.000) (0.000) -0.0306*** -0.0273* -0.0205*** -0.0196*** -0.0387*** -0.0842*** (0.0114) (0.014) (0.007) (0.005) (0.012) (0.029) constnt 0.0754-0.7426*** -0.3712*** 0.0634* 0.5650*** 1.1260*** (0.0877) (0.090) (0.081) (0.036) (0.053) (0.091) Observtions 45,507 46,507 46,507 46,507 46,507 46,507 R-squred 0.0103 0.0025 0.0035 0.0052 0.0013 0.0011 squred 0.0039 0.0448 0.0268 0.0106 0.0156 0.0247 Adjusted R-squred. Tble 4. Estimtion results, employment growth ( ), 2007-2010, municiplity nd industry fixed effects. All firms. OLS nd quntile s. Stndrd errors in prentheses. Vrible OLS Quntile 0.00267*** 0.0062*** 0.0039*** 0.0000 0.0009*** -0.0008* 0.0000 0.0000 0.0000-0.0000 0.0000*** -0.0000-0.00003*** 0.000006-0.00007*** -0.0000-0.0001*** -0.0002*** -0.0002*** 0.0002 0.0006*** 0.0000-0.0012*** -0.0014*** -0.0089* -0.0274*** -0.0441*** -0.0000 0.0474*** -0.0002 (0.005) (0.010) (0.009) (0.001) (0.009) (0.020) constnt -0.0244-0.6316*** -0.3166*** -0.0000 0.2415*** 0.5381*** (0.039) (0.032) (0.059) (0.004) (0.028) (0.049) Observtions 42,133 42,133 42,133 42,133 42,133 42,133 R-squred 0.0176 0.0105 0.0098 0.00009 0.0067 0.0021 squred 0.0105 0.0521 0.0335 0.0000 0.0246 0.0262 Stndrd errors in prentheses. Adjusted R-squred. 17

Tble 5. Estimtion results sles-hgfs (0.90 percentile), per industry, 2007-2010. Stndrd errors in prentheses. Industry Vrible Mnufcturing Construction Retil Hospitlity Computer -0.0005 0.00006-0.0024** -0.0012 0.0077* (0.001) (0.001) (0.001) (0.003) (0.004) -0.0004-0.0003*** -0.00002*** -0.0003*** -0.0005*** (0.000) (0.000) (0.000) (0.000) (0.000) -0.0003* -0.0002-0.0001-0.0002** -0.0002*** (0.000) (0.000) (0.000) (0.000) (0.000) -0.0001-0.0004-0.0015*** -0.0008*** 0.00013 (0.001) (0.001) (0.001) (0.000) (0.001) -0.0710-0.0723** 0.0047 0.0924-0.0831 (0.052) (0.035) (0.050) (0.077) (0.086) constnt 1.0965*** 0.6443*** 1.0441*** 1.9468*** 0.8408*** (0.191) (0.152) (0.109) (0.107) (0.285) Observtions 4,728 16,744 14,528 4,619 5,888 R-squred 0.0273 0.0056 0.0102 0.0309 0.0176 squred 0.1190 0.0319 0.0543 0.1340 0.0786 Tble 6. Estimtion results employment-hgfs (0.90 percentile), per industry, 2007-2010. Stndrd errors in prentheses. Industry Vrible Mnufcturing Construction Retil Hospitlity Computer -0.0011* -0.0005-0.00001 0.0005 0.0013 (0.001) (0.001) (0.001) (0.002) (0.001) -0.000** 0.000*** -0.000-0.000* -0.0000*** (0.000) (0.000) (0.000) (0.000) (0.000) -0.0002** -0.0002*** -0.0002*** -0.0001*** -0.0001* (0.000) (0.000) (0.000) (0.000) (0.000) -0.0003-0.0007* -0.0013*** -0.0011** -0.0012*** (0.001) (0.000) (0.000) (0.000) (0.000) -0.0265-0.0296-0.0095 0.0686 0.1672** (0.023) (0.026) (0.018) (0.048) (0.066) constnt 0.1860*** 0.3414*** 0.4085*** 0.4410*** 0.7504*** (0.062) (0.081) (0.056) (0.047) (0.121) Observtions 4,412 15,403 13,187 4,125 5,006 R-squred 0.0328 0.0070 0.0053 0.0324 0.0108 squred 0.1480 0.0457 0.0487 0.1210 0.1010 18

Figure 1. Frequency distributions, chnge in totl sles nd number of employees, 2007-2010. Logrithmic chnge in totl sles Logrithmic chnge in number of employees 19

APPENDIX A. Estimtion results, per industry, with municiplity fixed effects Tble A1. Estimtion results, sles growth ( ), 2007-2010, municiplity fixed effects. OLS nd quntile. NACE-code 25, Mnufcturing. Stndrd errors in prentheses. Vrible OLS Quntile 0.0019* 0.0036** 0.0028*** 0.0014** -0.0001-0.0005 (0.001) (0.002) (0.001) (0.001) (0.001) (0.001) -0.0002-0.00002-0.0001 0.00007-0.0002-0.0004 0.000008 0.0003** 0.0001-0.00001-0.0001-0.0003* 0.0007** 0.0005 0.0001 0.00029 0.0003-0.0001 (0.000) (0.001) (0.001) (0.000) (0.001) (0.001) 0.0025 0.04057 0.02947 0.0149-0.0393-0.07097 (0.028) (0.035) (0.023) (0.018) (0.031) (0.052) constnt 0.2131-0.5512** -0.1948** 0.0891 0.5379*** 1.0965*** (0.2167) (0.222) (0.095) (0.065) (0.137) (0.191) Observtions 4,728 4,728 4,728 4,728 4,728 4,728 R-squred 0.0565 0.0111 0.01402 0.0209 0.0258 0.0273 squred -0.0059 0.0934 0.0523 0.0398 0.0593 0.119 Adjusted R-squred. Tble A2. Estimtion results, employment growth ( ), 2007-2010, municiplity fixed effects. OLS nd quntile, NACE-code 25, Mnufcturing. Stndrd errors in prentheses. Vrible OLS Quntile 0.0013*** 0.0027*** 0.0013*** 0.000002-0.0004-0.0011* (0.000) (0.001) (0.000) (0.000) (0.001) (0.001) 0.0000 0.0000 0.0000 0.0000** -0.0000-0.0000** 0.0000-0.00002-0.00003-0.0000-0.0001-0.0002** -0.0003 0.0004 0.0004-0.0000-0.0009** -0.0003 (0.000) (0.000) (0.000) (0.000) (0.000) (0.001) -0.0031-0.0022-0.0165 0.0257*** 0.0040-0.0265 (0.0137) (0.020) (0.013) (0.007) (0.019) (0.023) constnt -0.0054-0.2708** -0.1474-0.0001 0.1346*** 0.1860*** (0.1082) (0.137) (0.129) (0.022) (0.040) (0.062) Observtions 4,412 4,412 4,412 4,412 4,412 4,412 R-squred 0.0716 0.0159 0.0230 0.0323 0.0370 0.0328 squred 0.0058 0.122 0.0805 0.0226 0.0922 0.148 Adjusted R-squred. 20

Tble A3. Estimtion results, sles growth ( ), 2007-2010, municiplity fixed effects. OLS nd quntile. NACE-codes 42 nd 43, Construction. Stndrd errors in prentheses. Vrible OLS Quntile 0.0050*** 0.0085*** 0.0067*** 0.0033*** 0.0016** 0.00006 (0.001) (0.001) (0.000) (0.000) (0.001) (0.001) -0.0001 0.0001 0.00004-0.00005-0.0002*** -0.0003*** (0.0001) (0.000) (0.000) (0.000) (0.000) (0.000) 0.0001*** 0.0002** 0.00009*** -0.000006-0.00005-0.0002-0.0006*** -0.0003-0.0002-0.0002-0.0005* -0.0004 (0.0002) (0.000) (0.000) (0.000) (0.000) (0.001) -0.0623*** -0.0434* -0.0217-0.0290*** -0.0518** -0.0723** (0.0186) (0.023) (0.013) (0.009) (0.022) (0.035) constnt -0.0126-0.7000*** -0.4522*** -0.0713 0.3307*** 0.6443*** (0.1106) (0.075) (0.125) (0.091) (0.065) (0.152) Observtions 16,744 16,744 16,744 16,744 16,744 16,744 R-squred 0.0200 0.0075 0.0082 0.0089 0.0078 0.0056 squred 0.0025 0.0416 0.0279 0.0124 0.0171 0.0319 Adjusted R-squred. Tble A4. Estimtion results, employment growth ( ), 2007-2010, municiplity fixed effects. OLS nd quntile. NACE-codes 42 nd 43, Construction. Stndrd errors in prentheses. Vrible OLS Quntile 0.0031*** 0.0078*** 0.0048*** -0.0000 0.0008** -0.0005 (0.000) (0.001) (0.000) (0.000) (0.000) (0.001) 0.0000 0.0000 0.0000* 0.0000 0.0000*** 0.0000*** -0.0001*** -0.0000-0.0002*** 0.0000-0.00004** -0.0002*** -0.0001 0.0002 0.0011*** -0.0000-0.0006** -0.0007* -0.0151-0.01509-0.0250* 0.0000 0.0341* -0.0296 (0.009) (0.018) (0.013) (0.002) (0.020) (0.026) constnt -0.1455*** -0.7753*** -0.4992*** -0.0000 0.0387 0.3414*** (0.0534) (0.033) (0.104) (0.007) (0.025) (0.081) Observtions 15,403 15,403 15,403 15,403 15,403 15,403 R-squred 0.0252 0.0126 0.0120 0.0010 0.0092 0.0070 squred 0.0063 0.0660 0.0452 0.0007 0.0337 0.0457 Adjusted R-squred. 21

Tble A5. Estimtion results, sles growth ( ), 2007-2010, municiplity fixed effects. OLS nd quntile. NACE-code 47, Retil. Stndrd errors in prentheses. Vrible OLS Quntile 0.0025*** 0.0043*** 0.0030*** 0.0019*** 0.0004-0.0024** (0.001) (0.000) (0.000) (0.000) (0.001) (0.001) -0.00003 0.00007*** 0.00002-0.00008*** -0.00001* -0.00002*** 0.00006** 0.0002*** 0.00008*** 0.000007-0.0001** -0.0001-0.00056*** 0.0004-0.00003-0.0002-0.0013*** -0.0015*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.001) -0.02846-0.0639*** -0.0434*** -0.0225*** -0.0306 0.0047 (0.021) (0.020) (0.010) (0.008) (0.019) (0.050) constnt -0.0376-1.1206*** -0.5492*** -0.1560** 0.3688** 1.0441*** (0.186) (0.160) (0.161) (0.076) (0.178) (0.109) Observtions 14,528 14,528 14,528 14,528 14,528 14,528 R-squred 0.0229 0.0032 0.0039 0.0055 0.0065 0.0102 squred 0.0028 0.0498 0.0232 0.0094 0.0189 0.0543 Adjusted R-squred. Tble A6. Estimtion results, employment growth ( ), 2007-2010, municiplity fixed effects. OLS nd quntile. NACE-code 47, Retil. Stndrd errors in prentheses. Vrible OLS Quntile 0.0021*** 0.0044*** 0.0025*** 0.0009** -0.00002 (0.000) (0.001) (0.000) (0.000) (0.001) 0.0000 0.0000*** 0.0000* 0.0000*** -0.0000 (0.000) (0.000) (0.000) (0.000) (0.000) -0.00003** 0.00002-0.00007*** -0.0001*** -0.0002*** (0.000) (0.000) (0.000) (0.000) (0.000) -0.000219** 0.0004 0.0005*** -0.0010*** -0.0013*** (0.000) (0.000) (0.000) (0.000) (0.000) -0.0220** -0.0468** -0.0609*** 0.0330** -0.0095 (0.009) (0.018) (0.013) (0.013) (0.018) constnt -0.0385-0.4621*** -0.2746*** 0.2810*** 0.4085*** (0.076) (0.058) (0.055) (0.038) (0.056) Observtions 13,187 13,187 13,187 13,187 13,187 R-squred 0.0253 0.0103 0.0104 0.0074 0.0053 squred 0.0031 0.0642 0.0433 0.0355 0.0487 Adjusted R-squred. 22

Tble A7. Estimtion results, sles growth ( ), 2007-2010, municiplity fixed effects. OLS nd quntile. NACE-codes 55 nd 56, Hospitlity. Stndrd errors in prentheses. Vrible OLS Quntile 0.0008 0.0034*** 0.0037*** 0.0008-0.0011-0.0012 (0.002) (0.001) (0.001) (0.001) (0.001) (0.003) 0.000002 0.0003*** 0.0002 0.00004-0.00005-0.0003*** -0.00008* 0.00002 0.00001-0.00002-0.00008-0.0002** (0.00005) (0.000) (0.000) (0.000) (0.000) (0.000) -0.0006** -0.0003-0.0002-0.0002-0.0004-0.0008*** (0.0003) (0.001) (0.000) (0.000) (0.000) (0.000) 0.0833** 0.0274 0.0132 0.0205 0.0511 0.0924 (0.0407) (0.039) (0.024) (0.016) (0.034) (0.077) constnt 0.1461-3.7921*** -0.1864** -0.0325 1.5020*** 1.9468*** (0.4042) (0.148) (0.085) (0.091) (0.138) (0.107) Observtions 4,619 4,619 4,619 4,619 4,619 4,619 R-squred 0.0633 0.0092 0.0142 0.0248 0.0316 0.0309 squred 0.0033 0.107 0.0575 0.0273 0.0672 0.134 Adjusted R-squred. Tble A8. Estimtion results, employment growth ( ), 2007-2010, municiplity fixed effects. OLS nd quntile. NACE-codes 55 nd 56, Hospitlity. Stndrd errors in prentheses. Vrible OLS Quntile 0.0043*** 0.0077*** 0.0059*** -0.0000 0.0017 0.0005 (0.001) (0.001) (0.001) (0.001) (0.001) (0.002) 0.0000002 0.0000*** 0.0000 0.0000 0.0000-0.0000* -0.00003 0.00001-0.000007-0.0000-0.00009** -0.0001*** -0.0008*** -0.0006-0.0005*** -0.0000-0.0010** -0.0011** 0.0324 0.0539 0.0141 0.0000 0.0092 0.0686 (0.022) (0.038) (0.027) (0.008) (0.022) (0.048) constnt -0.3324-0.7697*** -0.7422*** -0.1335 0.4183*** 0.4410*** (0.2797) (0.068) (0.054) (0.116) (0.047) (0.047) Observtions 4,125 4,125 4,125 4,125 4,125 4,125 R-squred 0.0763 0.0266 0.0399 0.0317 0.0381 0.0324 squred 0.0113 0.131 0.0939 0.0259 0.0858 0.121 Adjusted R-squred. 23

Tble A9. Estimtion results, sles growth ( ), 2007-2010, municiplity fixed effects. OLS nd quntile. NACE-codes 62 nd 63, Computer. Stndrd errors in prentheses. Vrible OLS Quntile 0.0069*** 0.0087*** 0.0071*** 0.0065*** 0.0065 0.0077* (0.002) (0.003) (0.002) (0.001) (0.004) (0.004) -0.0002 0.0002*** 0.00007-0.0001-0.0003*** -0.0005*** 0.0001*** 0.0003*** 0.0002* 0.0001-0.0001** -0.0002*** 0.00007-0.0004 0.0005 0.0006 0.0005 0.0001 (0.000) (0.000) (0.001) (0.001) (0.001) (0.001) -0.0429-0.0141-0.0585-0.0732*** -0.0684-0.0831 (0.038) (0.053) (0.041) (0.023) (0.052) (0.086) constnt -0.3995-1.443229*** -0.8442** -0.2435 0.1102 0.8408*** (0.315) (0.311) (0.386) (0.164) (0.134) (0.285) Observtions 5,888 5,888 5,888 5,888 5,888 5,888 R-squred 0.0472 0.0138 0.0200 0.0298 0.0241 0.0176 squred 0.0032 0.0830 0.0505 0.0296 0.0520 0.0786 Adjusted R-squred. Tble A10. Estimtion results, employment growth ( ), 2007-2010, municiplity fixed effects. OLS nd quntile. NACE-codes 62 nd 63, Computer. Stndrd errors in prentheses. Vrible OLS Quntile 0.0047*** 0.0086*** 0.0060*** 0.0000* 0.000002 0.0013 (0.001) (0.001) (0.001) (0.000) (0.000) (0.001) -0.0000 0.0000-0.0000-0.0000*** -0.0000*** -0.0000*** -0.00004** 0.00002-0.00002 0.0000** -0.0000-0.0001* 0.0002 0.0004** 0.0010*** -0.0000-0.000002-0.0012*** -0.0016-0.0735*** -0.1717*** -0.000002 0.2140*** 0.1672** (0.016) (0.028) (0.039) (0.000) (0.031) (0.066) constnt -0.0655-0.8070*** -0.2297*** 0.000004 0.0002 0.7504*** (0.119) (0.142) (0.087) (0.000) (0.017) (0.121) Observtions 5,006 5,006 5,006 5,006 5,006 5,006 R-squred 0.052 0.0246 0.0227 0.0189 0.0116 0.0108 squred 0.0016 0.1120 0.1070 0.0139 0.0699 0.1010 Stndrd errors in prentheses. Adjusted R-squred. 24

APPENDIX B. Estimtion results, per industry, without municiplity fixed effects Tble B1. Estimtion results, sles growth ( ), 2007-2010, no fixed effects. OLS nd quntile s. NACE-code 25, Mnufcturing. Stndrd errors in prentheses. Vrible OLS Quntile 0.0022** 0.0040*** 0.0025*** 0.0018*** 0.00002 0.0018 (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) -0.0002-0.00004 0.00002 0.00007-0.0002* -0.0006*** (0.000) (0.000) (0.001) (0.000) (0.000) (0.000) 0.000002 0.0004** 0.00009-0.000009-0.0002*** -0.0003*** 0.0009*** 0.0008 0.0004 0.0003 0.0003 0.0008 (0.000) (0.002) (0.000) (0.000) (0.000) (0.001) -0.0080 0.0370 0.0243 0.0196-0.0214-0.1330*** (0.027) (0.038) (0.026) (0.015) (0.023) (0.051) constnt 0.1258*** -0.6240*** -0.3048*** 0.0721*** 0.5331*** 0.8814*** (0.028) (0.109) (0.030) (0.018) (0.030) (0.059) Observtions 4,728 4,728 4,728 4,728 4,728 4,728 R-squred 0.0035 0.0017 0.0022 0.0018 0.0000 0.0002 squred 0.0024 0.0110 0.0048 0.0020 0.0021 0.0070 Adjusted R-squred. Tble B2. Estimtion results, employment growth ( ), 2007-2010, no fixed effects. OLS nd quntile. NACE-code 25, Mnufcturing. Stndrd errors in prentheses. Vrible OLS Quntile 0.0013*** 0.0035*** 0.0019*** 0.000001-0.0003-0.0012 (0.000) (0.001) (0.000) (0.000) (0.000) (0.002) 0.00000 0.0000*** 0.0000 0.0000-0.0000*** -0.0000* 0.0000 0.0000-0.0001** -0.0000-0.0001* -0.0003*** -0.0002 0.0014** 0.0014*** -0.0000-0.0005*** -0.0002 (0.000) (0.001) (0.000) (0.000) (0.000) (0.001) -0.0080 0.01482-0.0218 0.0508*** -0.0001-0.1366** (0.013) (0.025) (0.018) (0.014) (0.017) (0.056) constnt 0.0533*** -0.4667*** -0.2158*** -0.00003 0.3245*** 0.6566*** (0.014) (0.028) (0.025) (0.006) (0.018) (0.049) Observtions 4,412 4,412 4,412 4,412 4,412 4,412 R-squred 0.0022 0.0003 0.0002 0.0000 0.0001 0.0000 squred 0.0011 0.0142 0.0127 0.0026 0.0014 0.0128 Adjusted R-squred. 25

Tble B3. Estimtion results, sles growth ( ), 2007-2010, no fixed effects. OLS nd quntile s. NACE-codes 42 nd 43, Construction. Stndrd errors in prentheses. Vrible OLS Quntile 0.0050*** 0.0090*** 0.0067*** 0.0033*** 0.0014** -0.0001 (0.001) (0.001) (0.000) (0.000) (0.001) (0.001) -0.0001 0.000* 0.000003-0.00005** -0.0001*** -0.0002*** (0.0001356) (0.000) (0.000) (0.000) (0.000) (0.000) 0.0001** 0.0002*** 0.00009*** -0.000004-0.00006* -0.0002*** -0.0006*** -0.0003-0.0001-0.0003*** -0.0005** -0.0005-0.0644*** -0.0522** -0.0296** -0.0217*** -0.0658*** -0.0795** (0.0183) (0.021) (0.013) (0.008) (0.018) (0.036) constnt -0.0281* -0.8063*** -0.5127*** -0.0936*** 0.37182*** 0.8027*** (0.0167) (0.024) (0.014) (0.009) (0.019) (0.037) Observtions 16,744 16,744 16,744 16,744 16,744 16,744 R-squred 0.0049 0.0041 0.0041 0.0043 0.0016 0.00002 squred 0.0046 0.0188 0.0140 0.0038 0.0015 0.0019 Adjusted R-squred. Tble B4. Estimtion results, employment growth ( ), 2007-2010, no fixed effects. OLS nd quntile. NACE-codes 42 nd 43, Construction. Stndrd errors in prentheses. Vrible OLS Quntile 0.0031*** 0.0087*** 0.0055*** 0.0000 0.0015*** -0.0009 (0.0003) (0.001) (0.000) (0.000) (0.000) (0.001) 0.0000 0.0000-0.0000 0.0000 0.0000*** 0.0000*** -0.00007*** -0.0001* -0.0002*** 0.0000-0.0001** -0.0002*** -0.0001 0.0002 0.0012*** 0.0000-0.0019*** -0.0007* -0.0160* -0.0227-0.0307** 0.0000 0.0402*** -0.0477** (0.0090) (0.036) (0.012) (0.002) (0.013) (0.022) constnt -0.0800*** -0.6875*** -0.4034*** 0.00000 0.1569*** 0.4990*** (0.008) (0.050) (0.020) (0.003) (0.023) (0.037) Observtions 15,403 15,403 15,403 15,403 15,403 15,403 R-squred 0.0083 0.0078 0.0057 0.0011 0.0002 squred 0.0080 0.0295 0.0219 0 0.0075 0.0023 Adjusted R-squred. 26