Modelling Public Sector Wage- Employment Behaviour: Evidence from Transitioning Economies Jelena Lausev The University of Belgrade Richard Disney The Institute for Fiscal Studies and University College London LSEE November, 2012
Motivation: Theoretical framework & stylised facts � Public sector a dominant employer in pre- transition economies – dominance declining during transition � Evidence of lower pay inequality in the public than in the private sector � Evidence of growing inequality during transition � The standard interpretation is a ‘solidaristic’ pay policy that is eroded by private sector
Public sector dominated employment pre-transition
Wage inequality lower pre-transition; grew more rapidly in transition
Returns to education lower pre-transition; grew faster in transition
Table 4: Gini coefficients for income inequality by ownership type
Study Country Data Period Method Pay Gap % Motivation 2: Empirical studies Depalo et Italy SHIW 1998-2008 QR men: 10 at lower end, 6 at al (2011) median, 0 at the higher end. Disney et al UK BHPS 1991-1999 OLS 5 men; 17.2 women (2003) FE 0 men; 9.2 women men (women):10 th ;50 th ;90 th Lucifora et France LFS 1998 Decompos al (2006) Italy SHIW F: 9(11);2.4(8.4);-5.5(3.4) UK LFS I: 8(8); 2(5); -2(1.3) UK:13.7(16.3);7.3(8.3);0(0) 17.4 at 10 th ; -6.8 at 90 th Machado et Portugal QP 1982, QR 11.8 at 10 th ; -1.6 at 90 th al (2001) 1994 Melly Germany SOEP 1984-2001 Decompos men (women): 5 (29.6) at 10 th and -17.4 (-7) at 90 th (2005) Monaster. Greece SES 2005 OLS 14.2 12.9 at 10 th ; 3.5 at 90 th et al (2011) QR Albrecht et Sweden LINDA 1998 OLS -9.5 men; -2.9 women men (women):10 th ;50 th ;90 th al (2003) QR -0.9(3.7); -8(-2); -15.5(-10)
Adamchik et Poland LFS 1996 IV -7(-10) al (2000) men (women) Newell Poland LFS 1994; 1998 OLS -12.9; -8.5 (2001) all workers Brainerd Russia CPOR 1993; 1998 OLS -27; -16.5 (2002) all workers Jovanovic et Yugoslavia LFS 2000 IV -9.4(-4) al (2003) men (women) Jovanovic et Moscow 1997 IV -14.3(-18.3) al (2004) men (women) Leping Estonia LFS 1989 QR -23;-31.2;-76.8 (2006) 2004 0;-2.8;-11.4 all 10 th ; 50 th ; 90 th Peter et al Ukraine LMS 1997-2003 OLS -20.5(-30.9) (2007) FE -22.6(-20.4) men (women) Hamori Hungary WS 1994; 2003 QR LS 1, -4; 11, -20 (2007) HS -30, -42,12, -48 men 10 th , 90 th Lausev Serbia LFS 1995-2003 Decomp -7.8(-4.3); -15(-19) (2010) 2004-2008 17(12.2);0(-5.9) men (women)
The model in the paper predicts: � A public sector pay ‘penalty’ relative to competitive market at the start of transition � More compressed pay in the public monopsony case than in the competitive market � Increase in the wage inequality as a result of decline in the public sector monopsony power � Increase in returns to education
Theoretical background � Static models of monopsony, especially in public sector (but typically consider only one kind of labour). � Mortensen (1990) and Burdett and Mortensen (1998), Mortensen (2003) and Manning (2003): � imperfect competition is a necessary explanation for the dispersion of pay � Burdett (2012): cost minimising government offers a single wage after it has chosen to employ a given number of workers in a steady-state.
A Model of Public Sector Monopsony � Non-profit: E E � Employs two kinds of labour: s u � Hires subject to Budget constraint: f E E max ( , ) E E s u s u = + wE w E w E s.t. s s u u � Upward sloping labour supply curve: ′ − 1 > Y w ε = ≥ = wE w E w 0 ' ( ) / ( ) 0 Ew ε w Ew
Model: Public Sector Monopsony continued Y = f E s E ( , ) Cobb Douglas production function: u α − α = Y E s E 1 < α < 0 1 u ∂ Y E w E w E α = α = = s s s s s ∂ + E Y w E w E wE s s s u u E E The elasticity of substitution between and u s E d s ln( ) w E s = d e E ln = ω e s σ = = = u 1 w w E 1 d d u u u ln( ) ln ω w s the slope of demand function with unit elasticity: ω de d = − ω e
Diagrammatic illustration for one type of labour ( E ) Employment Government budget constraint: fixed wage bill wE Labour supply curve E s ( w ) Public Monopsony solution B ( m m E w ) Competitive solution C E * Competitive wage m w ( w = w Y ) * ' Wage rate
Solutions � Two solutions ε → ∞ � Competitive Solution: Ew � Hence: α w E Y − Y w ' ' γ = = = s u s 1 = = 0 − α w E Y 1 ' ε w u s u Ew ε < ∞ � Monopsony Solution: Ew + ε � Hence: 1 u α ε α m w E E ′ − 1 > Y w ω = = = s u u u = 0 + ε − α − α m w E E 1 ε w 1 1 s u s s Ew ε s
Competitive versus Monopsony Solutions Because the monopsony implies that � ε ε and < θ = < < θ = < s u 0 1 0 1 s u + ε + ε 1 1 s u ε < ε ⇒ θ < θ and because : s u s u The wage ratio is lower (more compressed) in the � public monopsony case than in the competitive ω < γ market i.e. The employment ratio is greater in the public � monopsony case than in the competitive e m > e market i.e.
Economic Transition α ω θ e γ = ω m = e e � From: and = = s − α 1 γ θ m e u Transition means a decline in the relative θ public sector monopsony power: → s 1 θ u � This implies: m e e � a decline in towards γ ω � an increase in towards
Empirical studies of labour supply elasticity: � Based on recruitment and separation rates: � Card and Krueger (1995); Manning (2003); Van Der ε ≈ 5 − Berg and Ridder (1993): (all workers) 7 � Individual case studies: � Sullivan (1989): (skilled workers) ε = 1 . 26 � Short-run: s ε = � Long-run: 3 . 86 s � JOLE (2010) (skilled workers) ε = 0 . 1 � Staiger, Spetz and Phibbs: s ε = � Ransom and Sims: 3 . 7 s ε = 1 . 4 � Falch: s � Boal (1995): (unskilled workers) ε = 11 � Short-run: u ε = 30 � Long-run: u � Other: Disney (2011); Elliott et al (2007)
Public sector pay relative to private sector pay in Hungary: unconditional and conditional differences in real gross earnings by gender in period 1992-2003 0.4 Unconditional Men monthly 0.3 Unconditional Women monthly Unconditional 0.2 Men hourly Unconditional Pay Gap a Women hourly 0.1 Conditional Men monthly Conditional 0 Women monthly Conditional Men hourly -0.1 Conditional Women hourly ‘transformational -0.2 recession’ public sector wage reform ‘minimum wage hike’ -0.3 ‘stabilization program’ 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Year of survey
OLS and quantile regression estimates of public sector gross monthly pay premia and penalties, by highest educational qualification for male employees in Hungary Unskilled Skilled High-skilled 1992- 2001- 1992- 2001- 1992- 2001- Change Change Change 1999 2003 1999 2003 1999 2003 (1)-(2) (1)-(2) (1)-(2) (1) (2) (1) (2) (1) (2) -0.087 -0.067 -0.045 Mean -0.146*** -0.059*** -0.167*** -0.100*** -0.338*** -0.293*** (0.001) (0.001) (0.001) (0.001) (0.001) (0.002) -0.039 -0.048 -0.166 10 th 0.025*** 0.064*** 0.025*** 0.073*** -0.014*** 0.152*** (0.001) (0.001) (0.001) (0.001) (0.001) (0.002) -0.061 -0.111 -0.060 25 th -0.089*** 0.028*** -0.123*** -0.012*** -0.209*** -0.149*** (0.001) (0.002) (0.001) (0.001) (0.001) (0.001) -0.108 -0.121 -0.010 50 th -0.169*** -0.061*** -0.228*** -0.107*** -0.372*** -0.362*** (0.001) (0.001) (0.001) (0.001) (0.001) (0.002) -0.103 -0.039 -0.033 75 th -0.235*** -0.132*** -0.254*** -0.215*** -0.533*** -0.500*** (0.001) (0.002) (0.001) (0.002) (0.001) (0.001) -0.112 0.004 -0.009 90 th -0.277*** -0.165*** -0.265*** -0.269*** -0.614*** -0.605*** (0.001) (0.003) (0.001) (0.002) (0.001) (0.001)
Could other models explain the same phenomena? � Solidarity model � wages of skilled and unskilled workers are compressed because of egalitarian concerns � Bureaucratic model � but incremental pay structure does not have the same predictions � SBTC
Thank You Jelena Lausev jelena@ekof.bg.ac.rs
Recommend
More recommend
