Promotions, pay & gender discrimination 1. Background 2. Theory –Tournaments –Glass ceilings & sticky floors 3. Empirical evidence –Promotion –Pay.

Promotions, pay & gender discrimination Promotions, pay & gender discrimination

1. Background1. Background

2. Theory2. Theory– TournamentsTournaments– Glass ceilings & sticky floorsGlass ceilings & sticky floors

3. Empirical evidence3. Empirical evidence– PromotionPromotion– Pay Pay

1. Background1. Background

International evidence on gender pay gaps (table International evidence on gender pay gaps (table 1)1)– Males = largest share of the workforce (Finland)Males = largest share of the workforce (Finland)– Variations between public & private sectorsVariations between public & private sectors– Wide variation in the mean gender gapsWide variation in the mean gender gaps

Pooled: Britain = 25% versus Italy = 6%Pooled: Britain = 25% versus Italy = 6% Public sector lower than private sectorPublic sector lower than private sector

– Variation in gender pay gap across the pay distributionVariation in gender pay gap across the pay distribution E.g. 10E.g. 10thth percentile (low pay) versus 90 percentile (low pay) versus 90thth percentile (high pay) percentile (high pay) Glass ceilings? Glass ceilings?

1.1. Company levelCompany level (Wal Mart) (Wal Mart)Source: ‘Wal Mart and the Glass Ceiling, Feb 2004) Source: ‘Wal Mart and the Glass Ceiling, Feb 2004)

Percent Women in Store Mgmt & Hourly Supervisors, 2001

0102030405060708090

100

% W

omen

in J

ob

Women in Management at Wal-Mart and Women in Management at Wal-Mart and Competitors (Source: See above)Competitors (Source: See above)

Gender pay gaps, 2001 Gender pay gaps, 2001 (Average annual earnings; Source: See above)(Average annual earnings; Source: See above)

Note: Full time staff, includes bonuses, Data from Richard Drogin (2003) ‘Statistical analysis of gender patterns in Wal Mart workforce’.

JobJob % of Women% of Womenin Jobin Job

Male Male Salaries ($)Salaries ($)

FemaleFemaleSalaries ($)Salaries ($)

DifferenceDifferencein Salaries ($)in Salaries ($)

Regional Vice Pres.Regional Vice Pres. 1010 419,400419,400 279,800279,800 139,600139,600

District ManagerDistrict Manager 1010 239,500239,500 177,100177,100 62,40062,400

Store ManagerStore Manager 1414 105,700105,700 89,30089,300 16,40016,400

Assistant ManagerAssistant Manager 3636 39,80039,800 37,30037,300 2,5002,500

Management TraineeManagement Trainee 4141 23,20023,200 22,40022,400 800800

Department HeadDepartment Head 7878 23,50023,500 21,70021,700 1,8001,800

Sales AssociateSales Associate 6868 16,50016,500 15,10015,100 1,4001,400

CashierCashier 9393 14,50014,500 13,80013,800 700700

2. Theory2. Theory

A) Tournaments A) Tournaments – Who gets promoted?Who gets promoted?

ILM – promotion laddersILM – promotion ladders Slots – predetermined pay (prize)Slots – predetermined pay (prize) Relative performanceRelative performance The larger the spread between current position & The larger the spread between current position &

promoted position – greater the effortpromoted position – greater the effort Luck (or noise)Luck (or noise)

– Offsets effortOffsets effort– Larger spreadLarger spread

2. Theory2. Theory Tournament: winner (promoted) =WTournament: winner (promoted) =W11 and and

loser Wloser W22

Two workers Two workers jj and and kk OutputOutput

qqjj = m = mjj + e + ejj (1a)(1a) qqkk = m = mkk + e + ekk (1b)(1b)

Each workers behaviourEach workers behaviour Max WMax W11PP + W + W22 ( (1 - P1 - P) – C(m)) – C(m) (2) (2)

(for given effort)(for given effort) (W(W11 – W – W22) dP/dm = C’(m)) dP/dm = C’(m) (3)(3)

2. Theory2. Theory

Probability that j winsProbability that j wins– PP = Prob(m = Prob(mjj + e + ejj > m > mkk + e + ekk) ) – = Prob(m= Prob(mjj – m – mkk > e > ekk - e - ejj))– = G(m= G(mjj – m – mkk))– G is the distribution function, we want g = the density G is the distribution function, we want g = the density

functionfunction EquilibriumEquilibrium

– (W(W11 – W – W22)g(0) = C’(m))g(0) = C’(m) See Lazear for firms problem – max expected See Lazear for firms problem – max expected

profitprofit

3. Evidence3. Evidence

Glass ceilings & sticky floorsGlass ceilings & sticky floors

– (i) Glass ceilings(i) Glass ceilings Why are gender pay gaps larger at the upper end of Why are gender pay gaps larger at the upper end of

the distribution?the distribution?

– (ii) Sticky floors(ii) Sticky floors Why are gender pay gaps smaller at the lower end of Why are gender pay gaps smaller at the lower end of

the distribution?the distribution?

3. Empirical evidence3. Empirical evidence

Causes of differences in gender gapsCauses of differences in gender gaps– Gender-specific policies – cross sectionalGender-specific policies – cross sectional

a) Parental leave & state provision of child carea) Parental leave & state provision of child care

b) Pay-bargaining institutions b) Pay-bargaining institutions

c) Other factorsc) Other factors– Appointments & promotions Appointments & promotions

3. Empirical evidence3. Empirical evidence

Gender & promotion – single companyGender & promotion – single company ‘‘glass ceiling’ v. other causesglass ceiling’ v. other causes Financial services Financial services

– 4,379 employees - closed ILM4,379 employees - closed ILM– port = low level clericalport = low level clerical– pay = f(grade)pay = f(grade)

Determinants of allocation to each grade?Determinants of allocation to each grade?– Education, tenure, age on entryEducation, tenure, age on entry– simulate distribution of females if treated as malessimulate distribution of females if treated as males

(i) cont.(i) cont.

– Little change in distributionLittle change in distribution Officer grades: rise 14% to 19%Officer grades: rise 14% to 19% Managers: rise 1% to 3%Managers: rise 1% to 3%

– Women have different LM characteristicsWomen have different LM characteristics e.g. lower tenuree.g. lower tenure

– No ‘glass ceiling’No ‘glass ceiling’

ConclusionsConclusions

Promotion is an incentive mechanism in Promotion is an incentive mechanism in ILMsILMs– Understood via tournament theoryUnderstood via tournament theory

Promotion & pay are linkedPromotion & pay are linked Gender differences in payGender differences in pay

– Differences in promotion rates - discriminationDifferences in promotion rates - discrimination– Or other factors?Or other factors?– Evidence is mixedEvidence is mixed