Public Employee Pensions and Collective BargainingRights: Evidence from State and Local Government
FinancesBrookings Municipal Finance Conference
Brigham R. Frandsen Michael Webb
BYU
July 18, 2017
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 1 / 21
05
1015
20
2000 2005 2010 2015year
Source: U.S. Courts (2016) Bankruptcy Filings, table F2
Chapter 9 Bankruptcy Filings
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 2 / 21
01,
000
2,00
03,
000
4,00
0
1960 1970 1980 1990 2000 2010year
collective bargaining required by 1996collective bargaining not required by 1996
$ millionsPublic Employee Pension Benefit Payouts
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 3 / 21
050
01,
000
1,50
0
1960 1970 1980 1990 2000 2010year
collective bargaining required by 1996collective bargaining not required by 1996
$ millionsGovernment Contributions to Public Employee Pensions
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 4 / 21
Institutional Background
Before 1950s: public sector collective bargaining prohibited
1961: AFT organized NYC teachers
1962: JFK signs Executive Order 10988
Starting early 1960s: States begin to allow public sector collectivebargaining
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 5 / 21
Institutional Background
Before 1950s: public sector collective bargaining prohibited
1961: AFT organized NYC teachers
1962: JFK signs Executive Order 10988
Starting early 1960s: States begin to allow public sector collectivebargaining
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 5 / 21
Institutional Background
Before 1950s: public sector collective bargaining prohibited
1961: AFT organized NYC teachers
1962: JFK signs Executive Order 10988
Starting early 1960s: States begin to allow public sector collectivebargaining
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 5 / 21
Institutional Background
Before 1950s: public sector collective bargaining prohibited
1961: AFT organized NYC teachers
1962: JFK signs Executive Order 10988
Starting early 1960s: States begin to allow public sector collectivebargaining
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 5 / 21
010
2030
4050
1960 1980 2000 2020year
Number of States RequiringPublic Sector Collective Bargaining
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 6 / 21
Data
Collective bargaining rights status for each state and year (Freemanand Valletta + Frandsen)
State retirement pension payouts and contributions (U.S. CensusAnnual Surveys of State Government Finances)
Employee contributionsGovernment contributionsBenefit payments
State covariates (population growth, employment, etc.)
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 7 / 21
Data
Collective bargaining rights status for each state and year (Freemanand Valletta + Frandsen)
State retirement pension payouts and contributions (U.S. CensusAnnual Surveys of State Government Finances)
Employee contributionsGovernment contributionsBenefit payments
State covariates (population growth, employment, etc.)
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 7 / 21
Data
Collective bargaining rights status for each state and year (Freemanand Valletta + Frandsen)
State retirement pension payouts and contributions (U.S. CensusAnnual Surveys of State Government Finances)
Employee contributions
Government contributionsBenefit payments
State covariates (population growth, employment, etc.)
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 7 / 21
Data
Collective bargaining rights status for each state and year (Freemanand Valletta + Frandsen)
State retirement pension payouts and contributions (U.S. CensusAnnual Surveys of State Government Finances)
Employee contributionsGovernment contributions
Benefit payments
State covariates (population growth, employment, etc.)
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 7 / 21
Data
Collective bargaining rights status for each state and year (Freemanand Valletta + Frandsen)
State retirement pension payouts and contributions (U.S. CensusAnnual Surveys of State Government Finances)
Employee contributionsGovernment contributionsBenefit payments
State covariates (population growth, employment, etc.)
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 7 / 21
Data
Collective bargaining rights status for each state and year (Freemanand Valletta + Frandsen)
State retirement pension payouts and contributions (U.S. CensusAnnual Surveys of State Government Finances)
Employee contributionsGovernment contributionsBenefit payments
State covariates (population growth, employment, etc.)
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 7 / 21
Table: Sample Means by Public Employee Collective Bargaining Law Status
AllRequired for >50% of
Public EmployeesNot required for >50% of Public Employees
N (number of state-years) 1,383 629 754
Fraction of public employees covered by collective bargaining requirement
0.46 0.97 0.03
Population 5,016,064 5,429,966 4,670,779
Total public employee retirement benefit expenditures ($1,000)
340,446 506,887 201,598
Total public employee retirement benefit expenditures / population ($1,000)
0.064 0.087 0.045
Total contributions to public employee retirement system ($1,000)
484,425 672,156 327,608
Employer contributions ($1,000) 330,435 481,261 204,613
Employee contributions ($1,000) 153,990 190,895 122,995
Fraction of contributions made by employees
0.36 0.32 0.40
CB law
Notes: number of state-year observations and means for the variables listed in the left-hand column. Data are from Valletta and Freeman (1988) and the Census Bureau's Historical Database on Individual Government Finances.
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 8 / 21
Identification strategy
0.2
.4.6
.81
1950 1960 1970 1980 1990 2000year
IowaEmployee Contribution Fraction by Year
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 9 / 21
Identification strategy
Yst = δCBst + X ′stβ + α0s + γt + f (s, t) + εst
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 10 / 21
-100
,000
010
0,00
020
0,00
030
0,00
0es
timat
ed c
oeffi
cien
t
-5 0 5year relative to law change
Total Contributions to Public Employee Pension
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 11 / 21
-100
,000
010
0,00
020
0,00
030
0,00
0es
timat
ed c
oeffi
cien
t
-5 0 5year relative to law change
Total Employer Contribution to Public Employee Pension
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 12 / 21
-100
,000
-50,
000
050
,000
100,
000
estim
ated
coe
ffici
ent
-5 0 5year relative to law change
Total Employee Contribution to Public Employee Pension
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 13 / 21
-100
,000
010
0,00
020
0,00
030
0,00
0es
timat
ed c
oeffi
cien
t
-5 0 5year relative to law change
Total Public Employee Pension Benefit Expenditure
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 14 / 21
Results
Table: Estimated Effects of Collective Bargaining Requirements on PublicEmployee Pensions
X-sectionDependent variable (1) (2) (3) (4)
ln(retirement contributions) 0.198** 0.125* 0.172* 0.160*(0.0746) (0.0732) (0.0864) (0.0844)
ln(employer contributions) 0.287*** 0.230*** 0.290*** 0.278***(0.0919) (0.0817) (0.0912) (0.0905)
ln(employee contributions) -0.169 -0.157 -0.0921 -0.0987(0.197) (0.131) (0.139) (0.137)
Fraction contributed by employee -0.0630** -0.0685*** -0.0751*** -0.0742***(0.0308) (0.0213) (0.0218) (0.0221)
ln(retirement benefits) 0.330*** 0.0947 0.138* 0.134*(0.0868) (0.0733) (0.0744) (0.0764)
Baseline controls? Y Y Y YState effects? N Y Y YState trends? N Y Y YRegion x year effects? N N Y YControls for right-to-work and urbanicity? N N N Y
Difference-in-differences
Notes: Regression coefficients and clustered standard errors (by state) on the collective bargaining (CB) law variable. All regressions control for year effects, state unemployment rate, state GDP, the natural log of the population, and the factors indicated in the bottom rows. Data are from Valletta and Freeman (1988) and the Census Bureau's Historical Database on Individual Government Finances. *Statistically significant at the .10 level; **at the .05 level; ***at the .01 level.
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 15 / 21
Results
Table: Estimated Effects of Collective Bargaining Requirements on PublicEmployee Pensions
X-sectionDependent variable (1) (2) (3) (4)
ln(retirement contributions) 0.198** 0.125* 0.172* 0.160*(0.0746) (0.0732) (0.0864) (0.0844)
ln(employer contributions) 0.287*** 0.230*** 0.290*** 0.278***(0.0919) (0.0817) (0.0912) (0.0905)
ln(employee contributions) -0.169 -0.157 -0.0921 -0.0987(0.197) (0.131) (0.139) (0.137)
Fraction contributed by employee -0.0630** -0.0685*** -0.0751*** -0.0742***(0.0308) (0.0213) (0.0218) (0.0221)
ln(retirement benefits) 0.330*** 0.0947 0.138* 0.134*(0.0868) (0.0733) (0.0744) (0.0764)
Baseline controls? Y Y Y YState effects? N Y Y YState trends? N Y Y YRegion x year effects? N N Y YControls for right-to-work and urbanicity? N N N Y
Difference-in-differences
Notes: Regression coefficients and clustered standard errors (by state) on the collective bargaining (CB) law variable. All regressions control for year effects, state unemployment rate, state GDP, the natural log of the population, and the factors indicated in the bottom rows. Data are from Valletta and Freeman (1988) and the Census Bureau's Historical Database on Individual Government Finances. *Statistically significant at the .10 level; **at the .05 level; ***at the .01 level.
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 16 / 21
Results
Table: Estimated Effects of Collective Bargaining Requirements on PublicEmployee Pensions
X-sectionDependent variable (1) (2) (3) (4)
ln(retirement contributions) 0.198** 0.125* 0.172* 0.160*(0.0746) (0.0732) (0.0864) (0.0844)
ln(employer contributions) 0.287*** 0.230*** 0.290*** 0.278***(0.0919) (0.0817) (0.0912) (0.0905)
ln(employee contributions) -0.169 -0.157 -0.0921 -0.0987(0.197) (0.131) (0.139) (0.137)
Fraction contributed by employee -0.0630** -0.0685*** -0.0751*** -0.0742***(0.0308) (0.0213) (0.0218) (0.0221)
ln(retirement benefits) 0.330*** 0.0947 0.138* 0.134*(0.0868) (0.0733) (0.0744) (0.0764)
Baseline controls? Y Y Y YState effects? N Y Y YState trends? N Y Y YRegion x year effects? N N Y YControls for right-to-work and urbanicity? N N N Y
Difference-in-differences
Notes: Regression coefficients and clustered standard errors (by state) on the collective bargaining (CB) law variable. All regressions control for year effects, state unemployment rate, state GDP, the natural log of the population, and the factors indicated in the bottom rows. Data are from Valletta and Freeman (1988) and the Census Bureau's Historical Database on Individual Government Finances. *Statistically significant at the .10 level; **at the .05 level; ***at the .01 level.
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 17 / 21
Results
Table: Estimated Effects of Collective Bargaining Requirements on PublicEmployee Pensions
X-sectionDependent variable (1) (2) (3) (4)
ln(retirement contributions) 0.198** 0.125* 0.172* 0.160*(0.0746) (0.0732) (0.0864) (0.0844)
ln(employer contributions) 0.287*** 0.230*** 0.290*** 0.278***(0.0919) (0.0817) (0.0912) (0.0905)
ln(employee contributions) -0.169 -0.157 -0.0921 -0.0987(0.197) (0.131) (0.139) (0.137)
Fraction contributed by employee -0.0630** -0.0685*** -0.0751*** -0.0742***(0.0308) (0.0213) (0.0218) (0.0221)
ln(retirement benefits) 0.330*** 0.0947 0.138* 0.134*(0.0868) (0.0733) (0.0744) (0.0764)
Baseline controls? Y Y Y YState effects? N Y Y YState trends? N Y Y YRegion x year effects? N N Y YControls for right-to-work and urbanicity? N N N Y
Difference-in-differences
Notes: Regression coefficients and clustered standard errors (by state) on the collective bargaining (CB) law variable. All regressions control for year effects, state unemployment rate, state GDP, the natural log of the population, and the factors indicated in the bottom rows. Data are from Valletta and Freeman (1988) and the Census Bureau's Historical Database on Individual Government Finances. *Statistically significant at the .10 level; **at the .05 level; ***at the .01 level.
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 18 / 21
Results
Table: Estimated Effects of Collective Bargaining Requirements on PublicEmployee Pensions
X-sectionDependent variable (1) (2) (3) (4)
ln(retirement contributions) 0.198** 0.125* 0.172* 0.160*(0.0746) (0.0732) (0.0864) (0.0844)
ln(employer contributions) 0.287*** 0.230*** 0.290*** 0.278***(0.0919) (0.0817) (0.0912) (0.0905)
ln(employee contributions) -0.169 -0.157 -0.0921 -0.0987(0.197) (0.131) (0.139) (0.137)
Fraction contributed by employee -0.0630** -0.0685*** -0.0751*** -0.0742***(0.0308) (0.0213) (0.0218) (0.0221)
ln(retirement benefits) 0.330*** 0.0947 0.138* 0.134*(0.0868) (0.0733) (0.0744) (0.0764)
Baseline controls? Y Y Y YState effects? N Y Y YState trends? N Y Y YRegion x year effects? N N Y YControls for right-to-work and urbanicity? N N N Y
Difference-in-differences
Notes: Regression coefficients and clustered standard errors (by state) on the collective bargaining (CB) law variable. All regressions control for year effects, state unemployment rate, state GDP, the natural log of the population, and the factors indicated in the bottom rows. Data are from Valletta and Freeman (1988) and the Census Bureau's Historical Database on Individual Government Finances. *Statistically significant at the .10 level; **at the .05 level; ***at the .01 level.
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 19 / 21
Results
Table: Estimated Effects of Collective Bargaining Requirements on PublicEmployment and Payroll
X-sectionDependent variable (1) (2) (3) (4)
ln(state government payroll) 0.0841 0.0293 0.0169 0.0131(0.0550) (0.0211) (0.0227) (0.0216)
ln(other government payroll) 0.0793 -0.0221 -0.0106 -0.0106(0.0682) (0.0188) (0.0200) (0.0204)
ln(total government payroll) 0.0972* -0.0209 -0.00938 -0.0171(0.0496) (0.0191) (0.0250) (0.0237)
ln(state government FTE employment) -0.0659 -0.0122 -0.0272 -0.0312*(0.0507) (0.0205) (0.0174) (0.0184)
ln(other government FTE employment) -0.0905* -0.0240 -0.0154 -0.0143(0.0489) (0.0204) (0.0223) (0.0227)
ln(total government FTE employment) -0.0639** -0.0317 -0.0216 -0.0279(0.0295) (0.0191) (0.0229) (0.0223)
Baseline controls? Y Y Y YState effects? N Y Y YState trends? N Y Y YRegion x year effects? N N Y YControls for right-to-work and urbanicity? N N N Y
Difference-in-differences
Notes: Regression coefficients and clustered standard errors (by state) on the collective bargaining (CB) law variable. All regressions control for year effects, state unemployment rate, state GDP, the natural log of the population, and the factors indicated in the bottom rows. Data are from Valletta and Freeman (1988) and the Census Bureau's Historical Database on Individual Government Finances. *Statistically significant at the .10 level; **at the .05 level; ***at the .01 level.
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 20 / 21
What do these results mean?
Dollar impact of collective bargaining rights:
∆X = nCB X̄ δ̂/(1 + δ̂
). (1)
$11.6B additional annual government pension payouts (29.2 percentof difference in growth between CB and non-CB states
$9.4B additional annual government contributions (61.9 percent ofdifference in growth)
$2B less employee contributions annually
$2517 additional compensation per worker (no evidence of crowd outof wages)
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 21 / 21
What do these results mean?
Dollar impact of collective bargaining rights:
∆X = nCB X̄ δ̂/(1 + δ̂
). (1)
$11.6B additional annual government pension payouts (29.2 percentof difference in growth between CB and non-CB states
$9.4B additional annual government contributions (61.9 percent ofdifference in growth)
$2B less employee contributions annually
$2517 additional compensation per worker (no evidence of crowd outof wages)
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 21 / 21
What do these results mean?
Dollar impact of collective bargaining rights:
∆X = nCB X̄ δ̂/(1 + δ̂
). (1)
$11.6B additional annual government pension payouts (29.2 percentof difference in growth between CB and non-CB states
$9.4B additional annual government contributions (61.9 percent ofdifference in growth)
$2B less employee contributions annually
$2517 additional compensation per worker (no evidence of crowd outof wages)
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 21 / 21
What do these results mean?
Dollar impact of collective bargaining rights:
∆X = nCB X̄ δ̂/(1 + δ̂
). (1)
$11.6B additional annual government pension payouts (29.2 percentof difference in growth between CB and non-CB states
$9.4B additional annual government contributions (61.9 percent ofdifference in growth)
$2B less employee contributions annually
$2517 additional compensation per worker (no evidence of crowd outof wages)
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 21 / 21
What do these results mean?
Dollar impact of collective bargaining rights:
∆X = nCB X̄ δ̂/(1 + δ̂
). (1)
$11.6B additional annual government pension payouts (29.2 percentof difference in growth between CB and non-CB states
$9.4B additional annual government contributions (61.9 percent ofdifference in growth)
$2B less employee contributions annually
$2517 additional compensation per worker (no evidence of crowd outof wages)
Frandsen and Webb (BYU) Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government FinancesJuly 18, 2017 21 / 21