Transcript Project Europe: Greece

Finance and Employment
M arc o P agan o
University of Naples Federico II, CSEF, EIEF
and CEPR
Giovanni Pica
University of Salerno and CSEF
E c o n o m i c P o l i c y , 5 3 rd p a n e l m e e t i n g
1 5 a n d 1 6 A p r i l 2 0 11 - M a g y a r N e m z e t i B a n k
Is finance the enemy of labor?
 Textbook vision of the role of finance:
 efficient allocation of capital
 optimal risk sharing
 Probably most workers instead view finance as:
 creating employment risk via corporate restructuring,
bankruptcies, and financial crises
 enabling or spurring firms to maximize share value at labor’s
expenses
 allowing bankers to earn astronomical bonuses, etc.
 The crisis has reinforced this negative vision of finance.
Good time to think about this…
Outline
1. How does financial development (FD) affect
employment, wage and productivity growth?
2. How does it affect the variability of employment?
3. Does it magnify the employment losses at times of
banking crisis?
Toy model of labor market response to FD
 For a start, consider case of identical firms
 1-size continuum of firms with Cobb-Douglas technology:
Y   K  L1
 Entrepreneur has wealth A
 He can “steal” (at most) fraction 1 of revenues (net of
wages: no stealing from workers)
 Better investor protection
  more funding to firms
(henceforth  = degree of FD)
 Competitive credit and labor markets
FD also amplifies employment response to shocks
 FD raises the response of employment to changes in:


growth opportunities  : firms can better exploit them  hire more
labor, offer higher wages
initial cash flow A: it allows firms to lever more on its cash  hire more
labor, offer higher wages
 This result hinges on firms being finance-constrained.
True with CRS: firms always want to expand
 If there is an efficient scale K*, once firms are past K*:

effect of FD abates as economy grows

FD no longer affects employment response to cash-flow shocks
Evidence on finance, employment and wages
 We extend the approach by Rajan and Zingales (1998): FD
should matter more for industries that are more
“dependent on external finance”
 External dependence = reliance on external finance by U.S.
listed companies in the Compustat database
 Baseline specification:
Y jc   ( FDc  ED j )   SHARE1970
jc   j  c   jc
Data
 Value added, employment and wage bill (Yj c): UNIDO
INDSTAT3 2006 database, 1970-2003 yearly data for

28 three-digit-industries

63 countries
 External dependence (EDj): Rajan and Zingales (1998)
 Financial development (FDc):

private credit/GDP

stock market capitalization/GDP (1980–95 averages)
Finance, employment and wages: all countries
Dependent variable:
Industry’s share in 1970
External dependence 
stock market capitalization
(80-95)
Growth of Value
Added
Wage Growth
Labor Productivity
Growth
-0.156*** 0.204*** 0.141*** 0.167*** 0.020*** 0.022*** 0.002*** 0.002***
(0.030)
(0.027)
(0.026)
(0.029)
(0.003)
(0.003)
(0.001)
(0.001)
0.026*
(0.014)
External dependence 
claims of banks and other
fin. inst. (80-95)
Observations
R-squared
Employment Growth
0.037***
(0.013)
0.034**
(0.016)
1533
0.32
1637
0.33
0.00004
(0.004)
0.055***
(0.014)
1447
0.42
1526
0.39
0.002
(0.011)
0.008
(0.013)
0.002
(0.004)
1293
0.72
1370
0.68
1428
0.29
Effect is between 0.23% and 0.83% as FD rises from 25th to 75th percentile
1505
0.30
Finance, employment and wages: OECD
Dependent variable:
Industry’s share in 1970
External dependence  stock
market capitalization (80-95)
Growth of Value
Added
0.212***
(0.054)
0.212***
(0.055)
0.022
(0.018)
External dependence  claims of
banks and other fin. inst. (8095)
Employment Growth
0.153***
(0.044)
0.155***
(0.045)
0.011
(0.012)
0.011
(0.011)
Wage Growth
Labor Productivity
Growth
0.022***
(0.004)
0.022***
(0.004)
0.001
(0.001)
0.010
(0.007)
0.002
(0.004)
0.021**
(0.011)
0.001
(0.001)
0.012*
(0.007)
0.009
(0.008)
Observations
628
628
624
624
594
594
622
622
R-squared
0.48
0.48
0.55
0.55
0.70
0.70
0.34
0.34
Finance, employment and wages: non-OECD
Dependent variable:
Industry’s share in 1970
External dependence  stock
market capitalization (80-95)
Growth of Value
Added
Employment Growth
Wage Growth
0.161*** 0.213***
(0.032)
(0.030)
0.163*** 0.185***
(0.033) (0.037)
0.021*** 0.023***
(0.003)
(0.003)
0.041***
(0.015)
0.003
(0.004)
0.037**
(0.016)
External dependence  claims
of banks and other fin. inst. (8095)
Observations
R-squared
0.091**
(0.036)
905
0.30
1009
0.32
0.133***
(0.033)
823
0.31
902
0.30
Labor Productivity
Growth
0.003***
(0.001)
0.008
(0.013)
0.000
(0.030)
-0.000
(0.010)
699
0.64
0.003***
(0.001)
776
0.62
806
0.24
883
0.27
Effect on employment reallocation

Extend model to 2 industries with different prospects:
 strong industry H with high expected profitability H
 weak industry L has low expected profitability L

Labor flows freely between them: single equilibrium wage w

Now FD affects not only total employment but also its
distribution between industries – in favor of industry H !

With higher FD, industry H attracts more funds than L:

employment in industry H grows by more than in industry L

employment in industry L may drop (if Ls is sufficiently inelastic)

sufficiently high FD will eventually “shut off” industry L
Response to growth shocks and to cash flow shocks
 With greater FD, sectoral growth shocks entail more cross-
industry employment reallocations
 But as FD proceeds, more and more firms achieve their
efficient scale and become unconstrained:

these firms stop reacting to cash flow shocks…
 As FD rises, it lowers the effect of cash flow shocks on job
reallocations (eventually eliminates it)
Evidence on finance and labor reallocation
 Strategy: regress a measure of inter-industry reallocation
on measures of


FD
FD  cross-industry dispersion of stock returns: FD should amplify
response of sd to growth shocks but lower it to cash flow shocks
sd (Y jct )   FDct   sd (DRI jct )  FDct  c  t   jct
where sd = cross-industry st. dev. deviation of Yjct
(industry j’s growth in VA, L or w) in country c and year t
Finance and reallocation: augmented specification
Dependent
variable: standard
deviation by year
and country of
Private credit by
deposit money
banks and other
financial
institutions to
GDP
Stock market
capitalization to
GDP
Stock market total
value traded to
GDP
Standard deviation
of continental
price shocks  Fin.
Dev.
Observations
R-squared
Value
added
growth
Value
added
growth
Value
added
growth
0.097
(0.064)
Employment
growth
Employment
growth
Employment
growth
0.013
(0.043)
0.072***
(0.025)
Wage
growth
Wage
growth
Wage
growth
0.054*
(0.029)
0.053***
(0.017)
0.089*
(0.045)
0.011
(0.008)
0.069***
(0.024)
0.029**
(0.012)
0.152
(0.191)
0.236**
(0.118)
0.383
(0.250)
0.069
(0.123)
0.182**
(0.074)
0.247**
(0.110)
0.153**
(0.073)
0.063**
(0.032)
0.100*
(0.054)
1281
0.02
874
0.03
893
0.03
1246
0.02
857
0.04
875
0.04
1207
0.03
824
0.03
846
0.03
Crises: the “dark side” of financial development?
 FD may become a handicap in a crisis because they create
“dependence”: the more financial markets are trusted in
normal times, the greater the damage to output and
employment when a crisis hits
 Most clearly seen by looking at liquidity provision: in
normal times banks allow firms to save on liquidity 
deploy more resources to production
 But when banks are hit by a liquidity shortage, the damage
can be more severe
Evidence on banking crises
Two empirical strategies:
1. Kroszner et al. (2007): re-estimate Rajan-Zingales
regressions before, during and after a financial crisis:
o 1 crisis observation per country, averaging crisis episodes for
countries that experience more than one crisis
2. Panel data approach similar to Braun & Larrain (2005):
Y jct   0 SHAREcjt 1  1 ( ED j  crisisct )   2 ( FDc  ED j )
  3 ( FDc  ED j  crisisct )  ct   j   jct
Financial crisis data from Laeven and Valencia (2010): universe
of banking crises (1970-2009)
Panel data evidence on severe banking crises
Value
Added
growth
Employment
growth
Wage
growth
Value
Added
growth
Employment
growth
Wage
growth
Lagged industry's share of value
added (cols 1-4) / Lagged industry's
share of employment (cols 2-5) /
Lagged ratio of industry's wage to
average wage (cols 3-6)
-0.547***
(0.085)
-0.234***
(0.053)
-0.098***
(0.011)
-0.660***
(0.089)
-0.233***
(0.048)
-0.101***
(0.011)
External dependence × Banking
crisis
External dependence × Stock market
capitalization
0.026
(0.027)
0.052**
(0.021)
0.023
(0.015)
0.033**
(0.016)
-0.001
(0.012)
0.002
(0.007)
0.068*
(0.036)
0.049**
(0.023)
-0.014
(0.016)
External dependence × Stock market
capitalization × Banking crisis
0.024
(0.069)
-0.015
(0.030)
0.006
(0.021)
0.065***
(0.019)
0.046***
(0.013)
0.005
(0.006)
-0.055*
(0.030)
-0.050**
(0.022)
0.016
(0.015)
47431
0.01
45533
0.01
44265
0.05
External dependence × banks and
other fin. inst. claims
External dependence × banks and
other fin. inst. claims × Banking
crisis
Observations
R-squared
44856
0.01
43293
0.01
42033
0.05
Conclusions
 Financial development is associated with

more employment growth, but only in non-OECD countries

less employment reallocation (cross-industry dispersion of
employment growth)

but more employment reallocation in response to greater variability of
shocks to growth opportunities (cross-industry dispersion of stock
returns)
 Some evidence of a “dark side” of financial development:

during crises, employment growth drops more in financially
dependent sectors of countries with more developed financial markets