Insurance Between Firms The Role of Internal Labor Markets Giacinta Cestone (Cass Business School and ECGI) Chiara Fumagalli (Bocconi University and CEPR) Francis Kramarz (ENSAE, CREST and CEPR) Giovanni Pica (Università di Milano, LdA and CSEF) NBER Summer Institute Boston, 10 July 2017
51
Embed
Insurance Between Firms The Role of Internal Labor Marketsdidattica.unibocconi.it/mypage/dwload.php?nomefile=Internal_Labor... · The Role of Internal Labor Markets ... I Employment
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Insurance Between FirmsThe Role of Internal Labor Markets
Giacinta Cestone (Cass Business School and ECGI)Chiara Fumagalli (Bocconi University and CEPR)
Francis Kramarz (ENSAE, CREST and CEPR)Giovanni Pica (Università di Milano, LdA and CSEF)
NBER Summer InstituteBoston, 10 July 2017
Introduction
This paper in a nutshell
We investigate the claim that INTERNAL LABOR MARKETS allow business groups torespond to both POSITIVE AND ADVERSE SHOCKS calling for labor adjustments, bypassing hiringand firing frictions in external labor markets.
Giacinta Cestone (Cass) ILMs and insurance between firms 1 / 28
Introduction
The theory: How ILMs can create value
External labor markets have FRICTIONS→ stand-alone firms can either (i) perform laboradjustments bearing hiring and firing costs, or (ii) avoid labor adjustments.
Within diversified groups, labor adjustments are less costly to perform within the INTERNALLABOR MARKET (ILM):
I Severance payments, advance notice and dismissal penalties avoided;I Search costs lower (less asymmetric information).
Following an idiosyncratic shock to an individual unit, a value maximizing group OPTIMALLYperforms some labor adjustment using the ILM.
→ Two sources of value: (i) greater flexibility; (ii) saving in hiring/firing costs.
Giacinta Cestone (Cass) ILMs and insurance between firms 2 / 28
Introduction
What we do:
Empirical findings in line with theoretical predictions:
I Business groups use their ILMs more than the external labor market in their responseto both POSITIVE and ADVERSE shocks.
I External labor market FRICTIONS drive the ILM response:
→ Employment protection regulations cause ILM reaction to adverse shocks;
→ Hiring costs for high-skill workers determine use of ILM after positive shocks.
I ILM provides implicit employment INSURANCE to workers.
Focus on BUSINESS GROUPS: networks of independent legal entities, often operating inmultiple industries and locations, controlled by a common owner.
Use administrative fiscal DATA FROM INSEE on universe of French firms and employees→employee year-to-year job moves (DADS); business group structures (LIFI); firms’ balancesheets (FICUS). Period 2002-2010.
Giacinta Cestone (Cass) ILMs and insurance between firms 3 / 28
Introduction
What we do:
Empirical findings in line with theoretical predictions:
I Business groups use their ILMs more than the external labor market in their responseto both POSITIVE and ADVERSE shocks.
I External labor market FRICTIONS drive the ILM response:
→ Employment protection regulations cause ILM reaction to adverse shocks;
→ Hiring costs for high-skill workers determine use of ILM after positive shocks.
I ILM provides implicit employment INSURANCE to workers.
Focus on BUSINESS GROUPS: networks of independent legal entities, often operating inmultiple industries and locations, controlled by a common owner.
Use administrative fiscal DATA FROM INSEE on universe of French firms and employees→employee year-to-year job moves (DADS); business group structures (LIFI); firms’ balancesheets (FICUS). Period 2002-2010.
Giacinta Cestone (Cass) ILMs and insurance between firms 3 / 28
Introduction
What we do:
Empirical findings in line with theoretical predictions:
I Business groups use their ILMs more than the external labor market in their responseto both POSITIVE and ADVERSE shocks.
I External labor market FRICTIONS drive the ILM response:
→ Employment protection regulations cause ILM reaction to adverse shocks;
→ Hiring costs for high-skill workers determine use of ILM after positive shocks.
I ILM provides implicit employment INSURANCE to workers.
Focus on BUSINESS GROUPS: networks of independent legal entities, often operating inmultiple industries and locations, controlled by a common owner.
Use administrative fiscal DATA FROM INSEE on universe of French firms and employees→employee year-to-year job moves (DADS); business group structures (LIFI); firms’ balancesheets (FICUS). Period 2002-2010.
Giacinta Cestone (Cass) ILMs and insurance between firms 3 / 28
Introduction Data
The data
Three data sources from INSEE: DADS (tracks workers year-to-year movements), LIFI (identifiesfirms affiliated with groups) and FICUS (includes firms’ financial statements) for the period2002-2010.
DADS Postes Files: administrative matched employer-employee data
I covers all employed people in the economy.I for each individual, information on plant/firm identifier IN YEAR t and IN YEAR t − 1.I for each year, information on: wage, number of working days, number of hours, type of occupation,
full time/part time status, plant and firm geographical location, industry classification, etc.
LIFI Files: survey collected by INSEE
I unique data set to study BG activityI available information: financial links between firms→ identification of ultimate owner. Group
structure: all firms (directly and indirectly) controlled by a common owner.I covers the vast majority of French BGs.
FICUS: administrative fiscal dataI covers the universe of French firms.I based on mandatory reporting of firms’ financial statements to tax authorities.I available information: usual balance sheet and income statement items.
Back
Giacinta Cestone (Cass) ILMs and insurance between firms 4 / 28
Introduction Literature
Related literature
Are diversified organizations better able to adapt to changing economic conditions in thepresence of frictions?
I INTERNAL CAPITAL MARKETS provide resilience in the face of adverse shocks(Almeida and Kim 2012; Kuppuswamy and Villalonga 2012; Boutin, Cestone,Fumagalli, Pica, Serrano-Velarde 2013), and ability to take advantage of positiveshocks (Giroud and Muller 2015).
I Employment sensitivity to the business cycle is different in business groups vsstand-alone firms (Faccio and O’Brien 2017).
I Earlier evidence of ILM ACTIVITY after plant closures in multi-establishment firms (Tateand Yang 2015).
Labor/Personnel economics: VERTICAL INTERNAL LABOR MARKETS within firms ascareer/incentive mechanisms (Doeringer and Piore 1971, etc.).
ORGANIZATION OF HUMAN CAPITAL and managerial quality are central to firm growth andperformance (Garicano 2000, Caliendo and Rossi-Hansberg 2012, Bertrand and Shoar2003, Bloom et al. 2014)).
Wage and employment INSURANCE WITHIN FIRMS: Guiso, Pistaferri, Schivardi 2005; Sraerand Thesmar 2010; Ellul, Pagano, Schivardi 2014.
Giacinta Cestone (Cass) ILMs and insurance between firms 5 / 28
Introduction Literature
Related literature
Are diversified organizations better able to adapt to changing economic conditions in thepresence of frictions?
I INTERNAL CAPITAL MARKETS provide resilience in the face of adverse shocks(Almeida and Kim 2012; Kuppuswamy and Villalonga 2012; Boutin, Cestone,Fumagalli, Pica, Serrano-Velarde 2013), and ability to take advantage of positiveshocks (Giroud and Muller 2015).
I Employment sensitivity to the business cycle is different in business groups vsstand-alone firms (Faccio and O’Brien 2017).
I Earlier evidence of ILM ACTIVITY after plant closures in multi-establishment firms (Tateand Yang 2015).
Labor/Personnel economics: VERTICAL INTERNAL LABOR MARKETS within firms ascareer/incentive mechanisms (Doeringer and Piore 1971, etc.).
ORGANIZATION OF HUMAN CAPITAL and managerial quality are central to firm growth andperformance (Garicano 2000, Caliendo and Rossi-Hansberg 2012, Bertrand and Shoar2003, Bloom et al. 2014)).
Wage and employment INSURANCE WITHIN FIRMS: Guiso, Pistaferri, Schivardi 2005; Sraerand Thesmar 2010; Ellul, Pagano, Schivardi 2014.
Giacinta Cestone (Cass) ILMs and insurance between firms 5 / 28
Introduction Literature
Related literature
Are diversified organizations better able to adapt to changing economic conditions in thepresence of frictions?
I INTERNAL CAPITAL MARKETS provide resilience in the face of adverse shocks(Almeida and Kim 2012; Kuppuswamy and Villalonga 2012; Boutin, Cestone,Fumagalli, Pica, Serrano-Velarde 2013), and ability to take advantage of positiveshocks (Giroud and Muller 2015).
I Employment sensitivity to the business cycle is different in business groups vsstand-alone firms (Faccio and O’Brien 2017).
I Earlier evidence of ILM ACTIVITY after plant closures in multi-establishment firms (Tateand Yang 2015).
Labor/Personnel economics: VERTICAL INTERNAL LABOR MARKETS within firms ascareer/incentive mechanisms (Doeringer and Piore 1971, etc.).
ORGANIZATION OF HUMAN CAPITAL and managerial quality are central to firm growth andperformance (Garicano 2000, Caliendo and Rossi-Hansberg 2012, Bertrand and Shoar2003, Bloom et al. 2014)).
Wage and employment INSURANCE WITHIN FIRMS: Guiso, Pistaferri, Schivardi 2005; Sraerand Thesmar 2010; Ellul, Pagano, Schivardi 2014.
Giacinta Cestone (Cass) ILMs and insurance between firms 5 / 28
Introduction Literature
Related literature
Are diversified organizations better able to adapt to changing economic conditions in thepresence of frictions?
I INTERNAL CAPITAL MARKETS provide resilience in the face of adverse shocks(Almeida and Kim 2012; Kuppuswamy and Villalonga 2012; Boutin, Cestone,Fumagalli, Pica, Serrano-Velarde 2013), and ability to take advantage of positiveshocks (Giroud and Muller 2015).
I Employment sensitivity to the business cycle is different in business groups vsstand-alone firms (Faccio and O’Brien 2017).
I Earlier evidence of ILM ACTIVITY after plant closures in multi-establishment firms (Tateand Yang 2015).
Labor/Personnel economics: VERTICAL INTERNAL LABOR MARKETS within firms ascareer/incentive mechanisms (Doeringer and Piore 1971, etc.).
ORGANIZATION OF HUMAN CAPITAL and managerial quality are central to firm growth andperformance (Garicano 2000, Caliendo and Rossi-Hansberg 2012, Bertrand and Shoar2003, Bloom et al. 2014)).
Wage and employment INSURANCE WITHIN FIRMS: Guiso, Pistaferri, Schivardi 2005; Sraerand Thesmar 2010; Ellul, Pagano, Schivardi 2014.
Giacinta Cestone (Cass) ILMs and insurance between firms 5 / 28
ILM as response to positive shocks
INTERNAL LABOR MARKETS AS A RESPONSE TO POSITIVE SHOCKS
Giacinta Cestone (Cass) ILMs and insurance between firms 6 / 28
ILM as response to positive shocks
Internal Labor Markets as a response to positive shocks
EMPIRICAL STRATEGY:
Exploit the exit of a large industry competitor→ (unexpected) growth opportunity:
I Focus on competitor exits that cannot be ascribed to industry-wide or macro problems
Study whether group-affiliated subsidiaries rely on their group’s ILM to respond to theexpansion opportunity:
I Do subsidiaries expand their workforce by absorbing more workers from their group’sILM (as opposed to the external labor market)?
I If yes, which workers?
Giacinta Cestone (Cass) ILMs and insurance between firms 7 / 28
ILM as response to positive shocks The Parmalat case
Parmalat exit: an unexpected growth opportunity
CASE STUDY: Parmalat collapse (December 2003)
I Large multinational (130 subsidiaries in 30 countries)I Major competitor for French firms active in the production/sale of milk productsI Unexpected exit due to discovery of accounting fraud
Was Parmalat’s collapse a POSITIVE SHOCK for its main French competitors?
I Yes, in two out of five of the 4-digit sectors where Parmalat was active.I Largest firms in “Wholesale Milk Trade" and “Other Milk Production” sectors expanded
(sales, employment, assets, PPE) significantly more than largest firms in all other4-digits in the French food industry.
Giacinta Cestone (Cass) ILMs and insurance between firms 8 / 28
ILM as response to positive shocks The Parmalat case
Parmalat exit: an unexpected growth opportunity
CASE STUDY: Parmalat collapse (December 2003)
I Large multinational (130 subsidiaries in 30 countries)I Major competitor for French firms active in the production/sale of milk productsI Unexpected exit due to discovery of accounting fraud
Was Parmalat’s collapse a POSITIVE SHOCK for its main French competitors?
I Yes, in two out of five of the 4-digit sectors where Parmalat was active.I Largest firms in “Wholesale Milk Trade" and “Other Milk Production” sectors expanded
(sales, employment, assets, PPE) significantly more than largest firms in all other4-digits in the French food industry.
Giacinta Cestone (Cass) ILMs and insurance between firms 8 / 28
ILM as response to positive shocks The Parmalat case
Evolution of bilateral worker flows
Giacinta Cestone (Cass) ILMs and insurance between firms 9 / 28
ILM as response to positive shocks The Parmalat case
Evolution of bilateral worker flows
Giacinta Cestone (Cass) ILMs and insurance between firms 10 / 28
ILM as response to positive shocks The Parmalat case
Bilateral worker flows following Parmalat shock
Study the bilateral flow of workers within firm pairs (origin/destination) in which thedestination firm is BG-AFFILIATED and OPERATES IN ONE OF THE SHOCKED SECTORS.
DIFF-IN-DIFF APPROACH: compare worker flows after the Parmalat shock with flows beforethe shock, from ILM firms of origin (“SameGroup”) versus external labor market firms oforigin:
I fijt= ratio of workers hired by BG-affiliated firm i (active in the shocked sectors) in year tand previously employed in firm j to total number of job-to-job movers hired by firm i inyear t.
I φij controls for firm pair fixed effects.
Giacinta Cestone (Cass) ILMs and insurance between firms 11 / 28
ILM as response to positive shocks The Parmalat case
Bilateral employment flows following Parmalat shock
NON SHOCKED SECTORS SHOCKED SECTORSVariables (1) (2) (3) (4)Same Group 0.0277*** 0.0230* 0.0135 0.0066
(0.0055) (0.0107) (0.0096) (0.0217)Firm of Origin group affiliated -0.0010 -0.0013 0.0003 -0.0020
(0.0014) (0.0027) (0.0037) (0.0070)Post 2004 × firm of origin group affiliated -0.0009 -0.0002 -0.0040 -0.0038
(0.0017) (0.0018) (0.0046) (0.0054)POST 2004 × SAME GROUP -0.0035 -0.0013 0.0293* 0.0350*
(0.0066) (0.0071) (0.0118) (0.0143)N 50,013 50,013 22,219 22,219Firm of destination FE YES NO YES NOFirm of origin × firm of destination FE NO YES NO YESYear dummies YES YES YES YES
→ Use non shocked sectors (other sectors where Parmalat was present that did not get any boostin 2004) as PLACEBO.
Giacinta Cestone (Cass) ILMs and insurance between firms 12 / 28
ILM as response to positive shocks Large competitor closures
Reaction to unexpected growth opportunity: beyond Parmalat
Identify 115 LARGE CLOSURE EVENTS taking place in 102 sectors. “Large” means 500+employees on average 4+ years prior to closure.
Identify 16 SHOCKED SECTORS: 4-digits in which largest firms’ sales and employmentexpand in DD regressions (i.e., post-closure event, compared to largest firms in other 4 digitsbut same 2-digit)
Use “non shocked sectors” (those in which top industry players’ sales and employment didnot improve upon large competitor closure) to run placebos.
Giacinta Cestone (Cass) ILMs and insurance between firms 13 / 28
ILM as response to positive shocks Large competitor closures
Bilateral employment flows and large competitor closures
(0.00253) (0.00489) (0.00498) (0.00158) (0.00169) (0.00231)Post shock × firm of origin group affiliated -0.00276*** -0.00365*** -0.00009 -0.00367*** -0.00444***
(0.00050) (0.00060) (0.00038) (0.00027) (0.00040)Post shock × Same Group 0.00583* 0.01153*** 0.00128 0.00062 0.00079
(0.00287) (0.00303) (0.00148) (0.00137) (0.00184)Closure year × Same Group 0.00618
(0.00401)Closure year+1 × Same Group 0.01116*
(0.00427)Closure year+2 × Same Group 0.01065*
(0.00416)Closure year +3 × Same Group 0.01996***
(0.00462)Closure year+4 × Same Group 0.01159
(0.00704)Closure year+5 or more × Same Group 0.00782
(0.00687)Post shock × Same Group × Managers/High-Skill 0.00531*
(0.00236)Post shock × Same Group × Intermediate Occupations -0.00095
(0.00197)Post shock × Same Group × Clerical Support 0.00207
(0.00148)
N 575,366 575,366 575,366 2,301,464 3,817,969 1,956,489Firm of destination FE YES NO NO NO NO NOFirm of origin × firm of destination FE NO YES YES YES YES YESYear dummies YES YES YES YES YES YESTime to shock dummies YES YES YES YES YES YES
Giacinta Cestone (Cass) ILMs and insurance between firms 14 / 28
ILM as response to positive shocks Large competitor closures
ILMs allow to bypass hiring frictions
Our finding: following exit of large competitor, group-affiliated firms expand by hiring moreMANAGERS AND SKILLED WORKERS from the ILM relative to the external market.
Hiring costs in the external labor market negligible for low-skilled, but large for skilled workers(Abowd and Kramarz 2003, Kramarz and Michaud 2010).
Search and training frictions mitigated within an organization (Greenwald 1986, Jaeger2016).
BOTTOM LINE: ILMs allow groups to overcome hiring frictions when expanding their skilled humancapital in good times.
Giacinta Cestone (Cass) ILMs and insurance between firms 15 / 28
ILM as response to adverse shocks
INTERNAL LABOR MARKETS AS A RESPONSE TO ADVERSE SHOCKS
Giacinta Cestone (Cass) ILMs and insurance between firms 16 / 28
ILM as response to adverse shocks Closures
Firm closures and mass layoffs within groups
Do groups rely on the ILM when responding to ADVERSE SHOCKS hitting affiliated subsidiaries?
We study displaced workers from group units undergoing closures/mass layoffs:
We identify “firm closures”: firms whose employment drops by more than 90% from one yearto the other.
We remove false closures: cases in which more than 70% of the lost employment ends up inthe same firm/plant.
Closures/mass layoffs preceded by performance slump (2-3 year window pre-closure). Clearsigns of financial distress one year before closure.
Giacinta Cestone (Cass) ILMs and insurance between firms 17 / 28
ILM as response to adverse shocks Empirical specification
Evolution of bilateral worker flows
Giacinta Cestone (Cass) ILMs and insurance between firms 18 / 28
ILM as response to adverse shocks Empirical specification
Evolution of bilateral worker flows
Giacinta Cestone (Cass) ILMs and insurance between firms 19 / 28
ILM as response to adverse shocks Empirical specification
Study the bilateral flow of workers between pairs of firms (origin/destination) where the firmof origin is BG-AFFILIATED and at some point subject to a CLOSURE/MASS LAYOFF.
DIFF-IN-DIFF APPROACH: compare worker flows at closure with flows in normal timestowards ILM (“SameGroup”) partners versus external labor market partners:
I fijt= ratio of workers moving from affiliated firm i (eventually closing) to firm j in year t tototal number of job-to-job movers from firm i in year t
I cit = 1 in last two years of firm i’s activityI φij is firm of origin × firm of destination fixed effect.
Giacinta Cestone (Cass) ILMs and insurance between firms 20 / 28
ILM as response to adverse shocks Descriptive statistics
Closures within the group trigger an ILM response
Outflow of workers from BG firms that eventually close: fraction moving towards ILM partnersversus fraction moving towards external labor market partners:
Giacinta Cestone (Cass) ILMs and insurance between firms 21 / 28
ILM as response to adverse shocks Results
Closures within the group trigger an ILM response
Variables (1) (2)
(3) (4)
Firm of destination group affiliated -0.0013*** 0.0011
-0.0021*** 0.0015***
(0.0003) (0.0007)
(0.000) (0.000)
Same Group 0.0334*** -0.0122**
0.0018 -0.0096***
(0.0019) (0.0041)
(0.001) (0.001)
Closure × firm of destination group affiliated 0.0004 0.0025***
-0.0001 0.0005
(0.0004) (0.0006)
(0.000) (0.000)
Closure × same group 0.1487*** 0.1187***
0.0452*** 0.0378***
(0.0039) (0.0050)
(0.002) (0.002)
Same group × Managers /High-Skill
0.0161*** 0.0161***(0.002) (0.002)
Same group × Intermediate Occupations
0.0093*** 0.0093***(0.001) (0.001)
Same group × Clerical Support
0.0010 0.0010(0.001) (0.001)
Closure × same group × Managers/High-Skill
-0.0082** -0.0082**(0.002) (0.002)
Closure × same group × Intermediate Occupations
-0.0129*** -0.0129***(0.002) (0.002)
Closure × same group × Clerical Support
-0.0112*** -0.0112***(0.002) (0.002)
N 1,171,552 1,171,552
4,686,112 4,686,112
Firm of origin FE YES NO
YES NO
Pair FE NO YES
NO YES
Year dummies YES YES
YES YES
Time to closure dummies YES YES
YES YES
Giacinta Cestone (Cass) ILMs and insurance between firms 22 / 28
ILM as response to adverse shocks Results
Closures within the group trigger an ILM response
Variables (1) (2) (3) (4)Firm of destination group affiliated -0.0013*** 0.0011 -0.0021*** 0.0015***
(0.0003) (0.0007) (0.000) (0.000)Same Group 0.0334*** -0.0122** 0.0018 -0.0096***
(0.0019) (0.0041) (0.001) (0.001)Closure × firm of destination group affiliated 0.0004 0.0025*** -0.0001 0.0005
(0.0004) (0.0006) (0.000) (0.000)Closure × same group 0.1487*** 0.1187*** 0.0452*** 0.0378***
(0.0039) (0.0050) (0.002) (0.002)Same group × Managers /High-Skill 0.0161*** 0.0161***
(0.002) (0.002)Same group × Intermediate Occupations 0.0093*** 0.0093***
(0.001) (0.001)Same group × Clerical Support 0.0010 0.0010
(0.001) (0.001)Closure × same group × Managers/High-Skill -0.0082** -0.0082**
(0.002) (0.002)Closure × same group × Intermediate Occupations -0.0129*** -0.0129***
(0.002) (0.002)Closure × same group × Clerical Support -0.0112*** -0.0112***
(0.002) (0.002)
N 1,171,552 1,171,552 4,686,112 4,686,112Firm of origin FE YES NO YES NOPair FE NO YES NO YESYear dummies YES YES YES YESTime to closure dummies YES YES YES YES
Giacinta Cestone (Cass) ILMs and insurance between firms 22 / 28
ILM as response to adverse shocks The role of EPR
Labor market frictions: the role of EPR
Firms above 50 employees are subject to stricter employment protection regulation (EPR) inFrance→ complex termination procedure for collective dismissals:
Higher firing costs
Procedure negotiated with union representatives
Procedure mandatory also in case of closure
Figure : Firm size distribution around the 50 employee threshold (year 2006)
(a) Stand-alone firms (b) Business group affiliated firms (c) All firms
Giacinta Cestone (Cass) ILMs and insurance between firms 23 / 28
ILM as response to adverse shocks The role of EPR
Labor market frictions: the role of EPR
Study the differential behavior above and below 50.
We address sorting around the threshold by INSTRUMENTING current size with (average) firm sizein normal times, i.e. 4+ years before closure.
FE estimates IV estimatesFirm size window 40-60 40-60 40-60 35-65 45-55Same Group 0.0381*** 0.0073 0.0325** 0.0165 0.0135
(0.0027) (0.0037) (0.0042) (0.0042) (0.0075)Closure × Same Group 0.1211*** 0.0785*** 0.0810*** 0.0970*** 0.0933***
(0.0158) (0.0222) (0.0107) (0.0106) (0.0171)Closure × same group × Firm size> 50 0.0515* 0.0705 0.0817*** 0.0421* 0.0647*
(0.0261) (0.0370) (0.0182) (0.0195) (0.0312)N 53,544 53,544 40,795 56,387 17,855Firm of origin FE YES NO NO NO NOFirm of origin × destination firm FE NO YES YES YES YESYear dummies YES YES YES YES YESTime to closure dummies YES YES YES YES YES
Giacinta Cestone (Cass) ILMs and insurance between firms 24 / 28
ILM as response to adverse shocks Insurance to workers?
Are group workers insured?
ILM PROVIDES JOB STABILITY: Employees of closing group subsidiaries less likely to moveto unemployment than employees of closing stand-alone companies.
I This implicit employment insurance stronger for blue collars.
After a closure, the employees redeployed to an ILM destination firm see a largerREDUCTION IN ANNUAL WAGE (due to fewer hours worked) when compared to displacedemployees who find a job in the external labor market.
Giacinta Cestone (Cass) ILMs and insurance between firms 25 / 28
ILM as response to adverse shocks Which firms absorb workers?
Which firms absorb workers within the ILM?
Employees displaced upon closures/mass layoffs are redeployed more intensely, within the ILM:
towards the group units enjoying BETTER GROWTH OPPORTUNITIES:
I firms in high-growth sectorsI firms with larger Capex in “normal times” (4+ years pre-closure event)
towards MORE EFFICIENT group units (higher TFP)
towards group units that have the FINANCIAL MUSCLE to seize expansion opportunities.
Giacinta Cestone (Cass) ILMs and insurance between firms 26 / 28
Conclusion
Conclusion
TAKEAWAYS:
Firing and hiring frictions in external labor markets trigger the use of ILMs:
I By mitigating “human capital constraints”, ILMs help affiliated firms grasp growthopportunities in good times.
I ILMs allows affiliated firms to bypass employment protection regulation in bad times.
To understand how firms respond to shocks, we must account for the different ways in whichconglomerates and stand-alone firms cope with labor market frictions.
IMPLICATIONS AND FUTURE RESEARCH:
Use of ILM in response to shocks privately beneficial for groups. Not necessarily sociallybeneficial.
Impact of ILM on “absorbing firms”?
Firm boundaries and organizational response to frictions.
Giacinta Cestone (Cass) ILMs and insurance between firms 27 / 28
Appendix Business Groups
What is a Business Group?
A collection of LEGALLY INDEPENDENT firms, possibly operating in multiple industries or locations,partly or wholly owned by a single family or firm that CONTROLS the member firms’ assets.
Giacinta Cestone (Cass) ILMs and insurance between firms Additional material slide 1
Appendix Business Groups
Business Groups are widespread
BGs account for a LARGE FRACTION OF THE ECONOMIC ACTIVITY in both emerging anddeveloped economies:
In India, in 2006 affiliated firms accounted for 80% of total assets of the top 100 firms (Sarkar 2010)
In 1996, the 30 largest chaebols accounted for 40% of South Korea’s total output (Chang and Hong, 2000)
In Germany, from 1996 to 1999, top 15 family groups controlled 25% of listed corporate assets, 22% inItaly (Faccio and Lang 2002).
Many prominent BG-form organizations in US and UK (Berkshire Hathaway, Alphabet, Virgin Group).
Comprehensive data for France:I From 1999 to 2010, affiliated firms (on average 5%) accounted for around 40% of total employment
and 60% of value added.I In manufacturing, such a percentage is as high as 70% (above 90% in automotive and energy).
Back
Giacinta Cestone (Cass) ILMs and insurance between firms Additional material slide 2
Appendix Business Groups
(French) Business Groups are highly heterogenous
The size distribution of French business groups is highly ASYMMETRIC:
FEW LARGE GROUPS, with many large affiliates, that are DIVERSIFIED both from a sectoraland geographical perspective
Groups in the top decile, on average:I have 20 firms (top percentile: more than 100 firms).I employ in each firm from 600 to 1000 workers (in 1999-2010).I operate in 7 different 4-digit industries (top percentile: 15 industries) and in 2 different
macrosectors.I have firms located in 4 different regions (top percentile: more than 7).
MANY SMALL GROUPS, with few small affiliates, that are HARDLY DIVERSIFIED.
Groups in the rest of the population:I have less than 6 firms.I employ less than 50 workers per firm.I operate in less than 3 different 4-digit sectors.I have firms mostly located in the same region.
Back
Giacinta Cestone (Cass) ILMs and insurance between firms Additional material slide 3
Appendix Data
The data
Three data sources from INSEE: DADS (tracks workers year-to-year movements), LIFI (identifiesfirms affiliated with groups) and FICUS (includes firms’ financial statements) for the period2002-2010.
DADS Postes Files: administrative matched employer-employee data
I covers all employed people in the economy.I for each individual, information on plant/firm identifier IN YEAR t and IN YEAR t − 1.I for each year, information on: wage, number of working days, number of hours, type of occupation,
full time/part time status, plant and firm geographical location, industry classification, etc.
LIFI Files: survey collected by INSEE
I unique data set to study BG activityI available information: financial links between firms→ identification of ultimate owner. Group
structure: all firms (directly and indirectly) controlled by a common owner.I covers the vast majority of French BGs.
FICUS: administrative fiscal dataI covers the universe of French firms.I based on mandatory reporting of firms’ financial statements to tax authorities.I available information: usual balance sheet and income statement items.
Back
Giacinta Cestone (Cass) ILMs and insurance between firms Additional material slide 4
Appendix Data
Why do BG affiliated firms close?
Figure : Performance of group affiliated closing firms
(a) ROA and ROS (b) Coverage (c) Sales
ROA denotes return on assets; ROS return on sales; coverage is the ratio of EBITDA over interest payments.
Back
Giacinta Cestone (Cass) ILMs and insurance between firms Additional material slide 5
Appendix Excess probability to originate from a closing firm of the same group
Excess probability to originate from a closing firm of the samegroup
Back
Giacinta Cestone (Cass) ILMs and insurance between firms Additional material slide 6
Appendix Occupations
Classification of occupationsCODE CATEGORY
10 Farmers2 TOP MANAGER/CHIEF OF FIRMS
21 Top managers/chiefs of handicraft firms22 Top managers/chiefs of industrial/commercial firms with less than 10 employees23 Top managers of industrial/commercial firms with more than 10 employees3 MANAGEMENT AND SUPERIOR INTELLECTUAL OCCUPATIONS
31 Healthcare professionals, legal professionals and other professionals33 Managers of the ’Fonction publique’34 Professors, researchers, scientific occupations35 Journalists, media, arts and entertainment occupations37 Administrative and commercial managers38 Engineers and technical managers4 INTERMEDIATE OCCUPATIONS
42 Teachers and other education, training and library occupations43 Healthcare support occupations and social services occupations44 Clergy and religious occupations45 Intermediate administrative occupations in the ’Fonction publique’46 Intermediate administrative and commercial occupations in firms47 Technicians48 Supervisors and ’agents de maitrise’5 WHITE COLLAR OCCUPATIONS
52 White collar occupations in the ’Fonction publique’53 Surveillance and security occupations54 Administrative white collars in firms55 Sales and related occupations56 Personal service occupations6 BLUE COLLAR OCCUPATIONS
62 Industrial qualified workers63 Handicraft qualified workers64 Drivers65 Maintenance, repair and transport qualified workers67 Industrial non qualified workers68 Handicraft non qualified workers69 Agricultural worker
Back to gammas Back
Giacinta Cestone (Cass) ILMs and insurance between firms Additional material slide 7
Appendix First stage
Top French competitors’ performance after Parmalat collapse
Sample: firms in 2-digit sector in which Parmalat was present
Treatment group: firms in 4-digit sectors in which Parmalat was presentControl group: firms in 4-digit sectors in which Parmalat was not present
Variables Sales Employment Total Assets Net PPE(1) (2) (3) (4)
Top 10 ×Wholesale Milk Trade × Post 0.1779*** 0.2383*** 0.1210* 0.1278**(0.0459) (0.0324) (0.0511) (0.0466)
Top 10 × Other Milk Production × Post 0.4343*** 0.2282*** 0.5029*** 0.3438***(0.0466) (0.0324) (0.0509) (0.0473)
Top 10 × Milk Production × Post 0.0124 -0.3459*** 0.2670*** -0.1436**(0.0459) (0.0324) (0.0512) (0.0468)
Top 10 × Butter × Post 0.1058* 0.0637 0.0661 -0.9385***(0.0467) (0.0327) (0.0539) (0.0472)
Top 10 × Cheese × Post -0.1081* 0.0253 -0.1438** -0.0537(0.0465) (0.0324) (0.0511) (0.0471)