Rethinking the Welfare State (Preliminary) Nezih Guner, Remzi - - PowerPoint PPT Presentation

Rethinking the Welfare State (Preliminary)

Nezih Guner, Remzi Kaygusuz and Gustavo Ventura Oslo-Penn-Toronto Conference, 2019

This Project

This Project

• We depart from standard one-earner, life-cycle framework with

incomplete markets.

This Project

• We depart from standard one-earner, life-cycle framework with

incomplete markets.

• We develop equilibrium framework with uninsurable shocks, a

realistic demographic structure and labor supply decisions in two-earner households. We use this framework for policy analysis.

This Project

• We depart from standard one-earner, life-cycle framework with

incomplete markets.

• We develop equilibrium framework with uninsurable shocks, a

realistic demographic structure and labor supply decisions in two-earner households. We use this framework for policy analysis. Questions:

This Project

• We depart from standard one-earner, life-cycle framework with

incomplete markets.

• We develop equilibrium framework with uninsurable shocks, a

realistic demographic structure and labor supply decisions in two-earner households. We use this framework for policy analysis. Questions:

• What are the roles of public policy and household decisions in

shaping economic inequality?

This Project

• We depart from standard one-earner, life-cycle framework with

incomplete markets.

• We develop equilibrium framework with uninsurable shocks, a

realistic demographic structure and labor supply decisions in two-earner households. We use this framework for policy analysis. Questions:

• What are the roles of public policy and household decisions in

shaping economic inequality?

• Do households value current social insurance/redistributive

programs in the U.S.? What are the effects of policy reforms? – focus today.

What we do

What we do

• Document facts on inequality over the life-cycle for different

types of households – married, single, skilled, unskilled.

What we do

• Document facts on inequality over the life-cycle for different

types of households – married, single, skilled, unskilled.

• Develop a life-cycle economy that has the potential to

account for these facts

What we do

• Document facts on inequality over the life-cycle for different

types of households – married, single, skilled, unskilled.

• Develop a life-cycle economy that has the potential to

account for these facts

• Use this framework to evaluate/understand quantitatively:

(i) how households value current welfare system; (ii) a system that replaces current taxes and transfers with a Negative Income Tax (iii) a system that replaces current transfers with a Universal Basic Income transfer.

Why we care

Why we care

• Inequality of earnings over the life-cycle.

Why we care

• Inequality of earnings over the life-cycle.

→ Marital status/gender differences. Policy analysis should be consistent with observed heterogeneity.

Why we care

• Inequality of earnings over the life-cycle.

→ Marital status/gender differences. Policy analysis should be consistent with observed heterogeneity.

• Current welfare system encompasses multiple programs that

transfer to poor and middle-income households. Can simple alternatives do better?

Why we care

• Inequality of earnings over the life-cycle.

→ Marital status/gender differences. Policy analysis should be consistent with observed heterogeneity.

• Current welfare system encompasses multiple programs that

transfer to poor and middle-income households. Can simple alternatives do better?

• Interplay between two-earner households, non-linear taxation

and the transfer system. Largely unexplored. → Social insurance/redistribution policy recommendations with two (potential) earners likely different than in single-earner economies.

Model - big picture

Model - big picture

• Ex-ante heterogenous married and single households hit by

uninsurable productivity shocks; → Permanent differences in endowments (education). → Permanent and persistent shocks to labor endowments.

Model - big picture

• Ex-ante heterogenous married and single households hit by

uninsurable productivity shocks; → Permanent differences in endowments (education). → Permanent and persistent shocks to labor endowments.

• Labor supply decisions at intensive and extensive margins;

Model - big picture

• Ex-ante heterogenous married and single households hit by

uninsurable productivity shocks; → Permanent differences in endowments (education). → Permanent and persistent shocks to labor endowments.

• Labor supply decisions at intensive and extensive margins;
• Skill depreciation for females associated to non-participation;

Model - big picture

• Ex-ante heterogenous married and single households hit by

uninsurable productivity shocks; → Permanent differences in endowments (education). → Permanent and persistent shocks to labor endowments.

• Labor supply decisions at intensive and extensive margins;
• Skill depreciation for females associated to non-participation;
• Equilibrium model with imperfect substitutability of skills in

production;

Model - big picture

• Ex-ante heterogenous married and single households hit by

uninsurable productivity shocks; → Permanent differences in endowments (education). → Permanent and persistent shocks to labor endowments.

• Labor supply decisions at intensive and extensive margins;
• Skill depreciation for females associated to non-participation;
• Equilibrium model with imperfect substitutability of skills in

production;

• Policy → tax credits, transfers and non-linear taxes

conditional on income and number of children

Model - big picture

• Ex-ante heterogenous married and single households hit by

uninsurable productivity shocks; → Permanent differences in endowments (education). → Permanent and persistent shocks to labor endowments.

• Labor supply decisions at intensive and extensive margins;
• Skill depreciation for females associated to non-participation;
• Equilibrium model with imperfect substitutability of skills in

production;

• Policy → tax credits, transfers and non-linear taxes

conditional on income and number of children

• Model extension of prior work; Guner, Kaygusuz, and Ventura

(2012a, 2012b, 2018).

Model – Demographics and Heterogeneity

• Life-cycle economy, j = 1, ...., JR, ....J. [25,26,.......,65,.....,80]
• Males (m) and females (f ), who differ in their permanent

types – skilled and unskilled (i = s, u).

• Male types map into productivity profiles, ̟m(i, j).
• Female types map into initial productivity levels.
• Agents can be single or married. Marital status is exogenous,

and does not change over the life-cycle.

Model – Female Skills

• Female types map into initial productivity levels, h1 = η(i).
• After age 1, labor market productivity of females evolves

endogenously: hi

j+1 = exp[ln hi j +

αi

e

• growth

χ(l) − δi

• depreciation

(1 − χ(l))], e : labor market experience.

Model – Idiosyncratic Productivity Shocks

Model – Idiosyncratic Productivity Shocks

• For an age-j single male of type i = s, u, earnings are given by

wi

• wage by skill

∗ ̟(j, i) ∗ exp(ηj + ν)

• labor endowment

∗ lm

• labor supply

Model – Idiosyncratic Productivity Shocks

• For an age-j single male of type i = s, u, earnings are given by

wi

• wage by skill

∗ ̟(j, i) ∗ exp(ηj + ν)

• labor endowment

∗ lm

• labor supply
• Persistent shock is governed by an AR(1) process

ηs,m

j+1 = ρηs,m j

+ εs,m

j+1,

with ηs,m

1

= 0, εs,m

j+1 ∼ N(0, σ2 εs,m).

Model – Idiosyncratic Productivity Shocks

• For an age-j single male of type i = s, u, earnings are given by

wi

• wage by skill

∗ ̟(j, i) ∗ exp(ηj + ν)

• labor endowment

∗ lm

• labor supply
• Persistent shock is governed by an AR(1) process

ηs,m

j+1 = ρηs,m j

+ εs,m

j+1,

with ηs,m

1

= 0, εs,m

j+1 ∼ N(0, σ2 εs,m).

• Permanent shock is a Gaussian draw:

ν ∼ N(0, σ2

νs,m)

Model – Idiosyncratic Productivity Shocks

Model – Idiosyncratic Productivity Shocks

• For a single female of age-j who has human capital hj,

earnings are given by wi

• wage by skill

∗ hj ∗ exp(ηj + ν)

• labor endowment

∗ lf

• labor supply

Model – Idiosyncratic Productivity Shocks

• For a single female of age-j who has human capital hj,

earnings are given by wi

• wage by skill

∗ hj ∗ exp(ηj + ν)

• labor endowment

∗ lf

• labor supply
• Persistent shock :

ηs,f

j+1 = ρηs,f j

+ εs,f

j+1

with ηs,m

1

= 0 and εs,f

j+1 ∼ N(0, σ2 εs,f ).

Model – Idiosyncratic Productivity Shocks

• For a single female of age-j who has human capital hj,

earnings are given by wi

• wage by skill

∗ hj ∗ exp(ηj + ν)

• labor endowment

∗ lf

• labor supply
• Persistent shock :

ηs,f

j+1 = ρηs,f j

+ εs,f

j+1

with ηs,m

1

= 0 and εs,f

j+1 ∼ N(0, σ2 εs,f ).

• Permanent shock:

ν ∼ N(0, σ2

νs,f )

Model – Idiosyncratic Productivity Shocks

Model – Idiosyncratic Productivity Shocks

• For married couples, earnings are given by

wif hi

j exp(ηm,f j

+ νm,f )

• labor endowment

lf + wim̟(im, j) exp(ηm,m

j

+ νm,m)

• labor endowment

∗ lm

Model – Idiosyncratic Productivity Shocks

• For married couples, earnings are given by

wif hi

j exp(ηm,f j

+ νm,f )

• labor endowment

lf + wim̟(im, j) exp(ηm,m

j

+ νm,m)

• labor endowment

∗ lm

• For j > 1, the bivariate AR(1) process is

ηm,m

j+1 = ρηm,m j

+ εm,m

j+1

, ηm,f

j+1 = ρηm,f j

+ εm,f

j+1

with ηm,m

1

= ηm,f

1

= 0 and (εm,m

j+1 , εm,f j+1) ∼ N

0 , σ2

εm,m

σεf εm σεf εm σ2

εf ,f

• ,

Model – Idiosyncratic Productivity Shocks

• For married couples, earnings are given by

wif hi

j exp(ηm,f j

+ νm,f )

• labor endowment

lf + wim̟(im, j) exp(ηm,m

j

+ νm,m)

• labor endowment

∗ lm

• For j > 1, the bivariate AR(1) process is

ηm,m

j+1 = ρηm,m j

+ εm,m

j+1

, ηm,f

j+1 = ρηm,f j

+ εm,f

j+1

with ηm,m

1

= ηm,f

1

= 0 and (εm,m

j+1 , εm,f j+1) ∼ N

0 , σ2

εm,m

σεf εm σεf εm σ2

εf ,f

• ,
• Permanent shocks:

(νm,m, νm,f ) ∼ N

• 0 , σ2

νm,m

σνmνf

1

σνmνf σ2

νf ,f

Model – Idiosyncratic Productivity Shocks

Comments:

Model – Idiosyncratic Productivity Shocks

Comments:

• Many parameters.

Model – Idiosyncratic Productivity Shocks

Comments:

• Many parameters.
• Variances and covariances depend on gender and marital

status.

Model – Idiosyncratic Productivity Shocks

Comments:

• Many parameters.
• Variances and covariances depend on gender and marital

status.

• We infer these variances and covariances from data –

inequality in wages and correlations in wages between spouses at different stages in life cycle.

Model – Idiosyncratic Productivity Shocks

Comments:

• Many parameters.
• Variances and covariances depend on gender and marital

status.

• We infer these variances and covariances from data –

inequality in wages and correlations in wages between spouses at different stages in life cycle.

• Specification of shocks is a mixture of RIP and HIP.

Model – Demographics and Heterogeneity

• Married households and single females differ in terms of the

number of children attached to them.

• Three possibilities: without, early, late (b = 0, 1, 2)
• If a female with children works, married or single, then the

household has to pay for child care costs, that vary with the age of children

• Joint market work for married couples also implies a utility

cost. → Residual heterogeneity in labor force participation.

Model – Transfers

Model – Transfers

• Welfare Programs: we use the Survey of Income and Program

Participation (SIPP), 1995-2013.

• Effective transfer functions from Rauh, Guner and Ventura

(2019).

• Include AFDC/TANF, SSI, Food Stamps/SNAP, WIC and

Housing Assistance.

Model – Transfers

• Welfare Programs: we use the Survey of Income and Program

Participation (SIPP), 1995-2013.

• Effective transfer functions from Rauh, Guner and Ventura

(2019).

• Include AFDC/TANF, SSI, Food Stamps/SNAP, WIC and

Housing Assistance.

• Child-related transfers: Child Tax Credit (CTC), Childcare

Credit (CDCTC) and CCDF (childcare subsidies).

Model – Transfers

• Welfare Programs: we use the Survey of Income and Program

Participation (SIPP), 1995-2013.

• Effective transfer functions from Rauh, Guner and Ventura

(2019).

• Include AFDC/TANF, SSI, Food Stamps/SNAP, WIC and

Housing Assistance.

• Child-related transfers: Child Tax Credit (CTC), Childcare

Credit (CDCTC) and CCDF (childcare subsidies).

• Earned Income Tax Credit (EITC)

Model – Transfers

• Welfare Programs: we use the Survey of Income and Program

Participation (SIPP), 1995-2013.

• Effective transfer functions from Rauh, Guner and Ventura

(2019).

• Include AFDC/TANF, SSI, Food Stamps/SNAP, WIC and

Housing Assistance.

• Child-related transfers: Child Tax Credit (CTC), Childcare

Credit (CDCTC) and CCDF (childcare subsidies).

• Earned Income Tax Credit (EITC)
• Total transfer functions: TRM(I, b, j) and TRS(I, b, j)

Model – Taxation

• Income tax functions: T M(I, b) and T S(I, b)
• We estimate these functions from Internal Revenue Service

(IRS) micro data – Guner, Kaygusuz and Ventura (2014)

• There is a social security system financed by a flat payroll tax,

τp, plus additional flat capital income tax τk.

Model – Preferences

• Single males and single females:

US

m (c, l) = log(c) − l1+ 1

γ ,

US

f (c, l) = log(c) − Bf l1+ 1

γ .

Model – Preferences

• Single males and single females:

US

m (c, l) = log(c) − l1+ 1

γ ,

US

f (c, l) = log(c) − Bf l1+ 1

γ .

• Married couples

UM(c, lf , lm, θ, q) = 2 log(c) − l

1+ 1

γ

m

− θ Bf l

1+ 1

γ

f

− χ{lf }q.

Model – Preferences

• Single males and single females:

US

m (c, l) = log(c) − l1+ 1

γ ,

US

f (c, l) = log(c) − Bf l1+ 1

γ .

• Married couples

UM(c, lf , lm, θ, q) = 2 log(c) − l

1+ 1

γ

m

− θ Bf l

1+ 1

γ

f

− χ{lf }q. θ takes two values at start of life; θ ∈ {θL, θH}

Decisions – Big Picture

Decisions – Big Picture

• Households have access to one-period, risk-free asset. They

decide how much to consume, save and the work of their members.

Decisions – Big Picture

• Households have access to one-period, risk-free asset. They

decide how much to consume, save and the work of their members.

• Given their state, married households decide whether the

female member should work.

• Costs of work: child care expenses, additional taxes.
• Benefits: higher household income, future human capital.

Decisions – Big Picture

• Households have access to one-period, risk-free asset. They

decide how much to consume, save and the work of their members.

• Given their state, married households decide whether the

female member should work.

• Costs of work: child care expenses, additional taxes.
• Benefits: higher household income, future human capital.
• Taxation plus presence and generosity of transfers affect the

cost and benefits of work.

Model and Data Statistic Data Model Capital Output Ratio 2.93 2.93 LFP of Married Females (%), 25-54 Unskilled 68.2 67.7 Skilled 77.4 77.3 Total 71.8 71.5 Variance log-wages (Married Males, age 40) 0.37 0.37 Variance log-wages (Married Females, age 40) 0.33 0.35 Variance log-hours (Married Females, age 40) 0.13 0.14 Correlation Between Wages of Spouses (age 25) 0.27 0.27 Correlation Between Wages of Spouses (age 40) 0.31 0.31 Skill Premium 1.8 1.8 Variance log-consumption (Age 50-54 vs 25-29) 0.08 0.07

Model and Data Statistic Data Model Capital Output Ratio 2.93 2.93 LFP of Married Females (%), 25-54 Unskilled 68.2 67.7 Skilled 77.4 77.3 Total 71.8 71.5 Variance log-wages (Married Males, age 40) 0.37 0.37 Variance log-wages (Married Females, age 40) 0.33 0.35 Variance log-hours (Married Females, age 40) 0.13 0.14 Correlation Between Wages of Spouses (age 25) 0.27 0.27 Correlation Between Wages of Spouses (age 40) 0.31 0.31 Skill Premium 1.8 1.8 Variance log-consumption (Age 50-54 vs 25-29) 0.08 0.07

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Age

Var-Log Married Female Earnings

Model Data

0.05 0.1 0.15 0.2 0.25 0.3 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Age

Var-Log Married Female Hours

Data Model

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55

Age

Married Female Labor Force Participation

Model Data

Rethinking the Welfare State

Rethinking the Welfare State

• What are the effects of abolishing the welfare state? Do

households value the current scheme?

Rethinking the Welfare State

• What are the effects of abolishing the welfare state? Do

households value the current scheme? → Eliminate all transfers. Taxes reduced for all.

Rethinking the Welfare State

• What are the effects of abolishing the welfare state? Do

households value the current scheme? → Eliminate all transfers. Taxes reduced for all.

• Replace all taxes and transfers with a Negative Income Tax

(NIT)

• Each household receives a transfer per member (including

children) in all dates and states.

• All households face same proportional income tax.

Rethinking the Welfare State

• What are the effects of abolishing the welfare state? Do

households value the current scheme? → Eliminate all transfers. Taxes reduced for all.

• Replace all taxes and transfers with a Negative Income Tax

(NIT)

• Each household receives a transfer per member (including

children) in all dates and states.

• All households face same proportional income tax.
• Replace all transfers with a Universal Basic Income (UBI)

transfer.

• Each adult receives a transfer in all dates and states.
• Existing taxes unchanged. Additional resources via

proportional income tax if needed.

Eliminating Welfare State

No Transfer System Output (%) 1.7 Married Females LFP (%) 3.4 Married Females LFP (U, %) 4.8 Married Females LFP (S, %) 1.7 Aggregate Hours (MF, %) 3.8 Aggregate Hours (%) 2.8 Variance Log-Earnings (benchmark value: 0.524) 0.486 Welfare (CV, %)

• 2.9

Winning Households (%) 62.0

Eliminating Welfare State

No Transfer System Output (%) 1.7 Married Females LFP (%) 3.4 Married Females LFP (U, %) 4.8 Married Females LFP (S, %) 1.7 Aggregate Hours (MF, %) 3.8 Aggregate Hours (%) 2.8 Variance Log-Earnings (benchmark value: 0.524) 0.486 Welfare (CV, %)

• 2.9

Winning Households (%) 62.0

• Asymmetric welfare effects. Large welfare losses – but majority

support for eliminating current scheme.

Eliminating Welfare State

No Transfer System Output (%) 1.7 Married Females LFP (%) 3.4 Married Females LFP (U, %) 4.8 Married Females LFP (S, %) 1.7 Aggregate Hours (MF, %) 3.8 Aggregate Hours (%) 2.8 Variance Log-Earnings (benchmark value: 0.524) 0.486 Welfare (CV, %)

• 2.9

Winning Households (%) 62.0

• Asymmetric welfare effects. Large welfare losses – but majority

support for eliminating current scheme.

• Elimination of welfare transfers (AFDC, etc) leads to largest losses.

Rethinking the Welfare State

NIT NIT NIT (0%) (2%) (4%) Output (%) 2.9 1.6 0.1 Married Females LFP (U, %) 7.3 2.7

• 2.3

Married Females LFP (S, %) 3.2 1.4

• 0.8

Aggregate Hours (MF, %) 6.9 3.0

• 1.3

Aggregate Hours (%) 4.4 2.3 0.0 Variance Log-Earnings (benchmark value: 0.524) 0.49 0.50 0.52 Tax Rate (%) 7.0 11.7 17.2 Welfare (CV, %)

• 4.0
• 1.2

0.1 Winning Households (%) 50.4 63.4 73.8

Rethinking the Welfare State

NIT NIT NIT (0%) (2%) (4%) Output (%) 2.9 1.6 0.1 Married Females LFP (U, %) 7.3 2.7

• 2.3

Married Females LFP (S, %) 3.2 1.4

• 0.8

Aggregate Hours (MF, %) 6.9 3.0

• 1.3

Aggregate Hours (%) 4.4 2.3 0.0 Variance Log-Earnings (benchmark value: 0.524) 0.49 0.50 0.52 Tax Rate (%) 7.0 11.7 17.2 Welfare (CV, %)

• 4.0
• 1.2

0.1 Winning Households (%) 50.4 63.4 73.8 → NIT can lead to welfare gains and majority support. But requires sizeable transfers.

‐4.5 ‐4 ‐3.5 ‐3 ‐2.5 ‐2 ‐1.5 ‐1 ‐0.5 0.5 10 20 30 40 50 60 70 80 0.5 1 1.5 2 3 4 5 6

Welfare (%) Winners (%)

Percentage of Mean Household Income

NIT, Welfare (right) and Winners (left)

Winners Welfare

‐4.5 ‐4 ‐3.5 ‐3 ‐2.5 ‐2 ‐1.5 ‐1 ‐0.5 0.5 ‐2 ‐1 1 2 3 4 0.5 1 1.5 2 3 4 5 6

Welfare (%) Output (%)

Percentage of Mean Household Income

NIT, Welfare (right) and Output (left)

Output Welfare

‐2 ‐1 1 2 3 4 0.46 0.47 0.48 0.49 0.5 0.51 0.52 0.53 0.54 0.5 1 1.5 2 3 4 5 6

Output (%) Var‐Log Earnings

Percentage of Mean Household Income

NIT, Output (right) and Var‐Log Earnings (left)

Var‐Log Earnings Output

Rethinking the Welfare State

NIT UBI No (Optimal) (Optimal) Transfers Output (%)

• 0.8
• 0.9

1.7 Married Females LFP (U, %)

• 5.1
• 3.2

4.8 Married Females LFP (S, %) 3.2 1.4 1.7 Welfare (CV, %) 0.2

• 1.0
• 2.9

Winning Households (%) 59.3 54.8 62.0 Transfer 5.0 5.15 0.0 (% Household Income) (per person) (per adult) – Tax Rate (%) 20.2 4.5 – (all income) (additional) –

Rethinking the Welfare State

NIT UBI No (Optimal) (Optimal) Transfers Output (%)

• 0.8
• 0.9

1.7 Married Females LFP (U, %)

• 5.1
• 3.2

4.8 Married Females LFP (S, %) 3.2 1.4 1.7 Welfare (CV, %) 0.2

• 1.0
• 2.9

Winning Households (%) 59.3 54.8 62.0 Transfer 5.0 5.15 0.0 (% Household Income) (per person) (per adult) – Tax Rate (%) 20.2 4.5 – (all income) (additional) – → UBI does NOT lead to welfare gains. Dominated by NIT.

Rethinking the Welfare State

NIT UBI No (Optimal) (Optimal) Transfers Output (%)

• 0.8
• 0.9

1.7 Married Females LFP (U, %)

• 5.1
• 3.2

4.8 Married Females LFP (S, %) 3.2 1.4 1.7 Welfare (CV, %) 0.2

• 1.0
• 2.9

Winning Households (%) 59.3 54.8 62.0 Transfer 5.0 5.15 0.0 (% Household Income) (per person) (per adult) – Tax Rate (%) 20.2 4.5 – (all income) (additional) – → UBI does NOT lead to welfare gains. Dominated by NIT. Optimal NIT transfer: $ 4,500 in current dollars.

Conclusions

Conclusions

• We develop life-cycle model suitable for policy analysis. It

goes a long way towards reproducing patterns of life-cycle inequality (all and new).

Conclusions

• We develop life-cycle model suitable for policy analysis. It

goes a long way towards reproducing patterns of life-cycle inequality (all and new).

• Overall, it is hard to improve over the existing welfare system.

Conclusions

• We develop life-cycle model suitable for policy analysis. It

goes a long way towards reproducing patterns of life-cycle inequality (all and new).

• Overall, it is hard to improve over the existing welfare system.
• Revenue-neutral elimination of all transfers leads to large

welfare losses BUT is supported by a majority of newborn households.

Conclusions

• We develop life-cycle model suitable for policy analysis. It

goes a long way towards reproducing patterns of life-cycle inequality (all and new).

• Overall, it is hard to improve over the existing welfare system.
• Revenue-neutral elimination of all transfers leads to large

welfare losses BUT is supported by a majority of newborn households.

• NIT arrangements can improve upon the status quo and be

supported by a majority. However, ex-ante gains are not large.

Conclusions

• We develop life-cycle model suitable for policy analysis. It

goes a long way towards reproducing patterns of life-cycle inequality (all and new).

• Overall, it is hard to improve over the existing welfare system.
• Revenue-neutral elimination of all transfers leads to large

welfare losses BUT is supported by a majority of newborn households.

• NIT arrangements can improve upon the status quo and be

supported by a majority. However, ex-ante gains are not large.

• NIT dominates UBI. KEY: larger redistribution is possible

under NIT via lower distortions.

Conclusions

• We develop life-cycle model suitable for policy analysis. It

goes a long way towards reproducing patterns of life-cycle inequality (all and new).

• Overall, it is hard to improve over the existing welfare system.
• Revenue-neutral elimination of all transfers leads to large

welfare losses BUT is supported by a majority of newborn households.

• NIT arrangements can improve upon the status quo and be

supported by a majority. However, ex-ante gains are not large.

• NIT dominates UBI. KEY: larger redistribution is possible

under NIT via lower distortions.

• More to come...

EXTRA SLIDES

The Structure of Shocks

The Structure of Shocks

• Permanent Shocks.

Variance single males: σ2

νs,m = 0.255

Variance single females: σ2

νs,f = 0.226

Variance married males: σ2

νm,m = 0.220

Variance married females: σ2

νm,f = 0.216

Correlation (married males, married females): 0.216

The Structure of Shocks

• Permanent Shocks.

Variance single males: σ2

νs,m = 0.255

Variance single females: σ2

νs,f = 0.226

Variance married males: σ2

νm,m = 0.220

Variance married females: σ2

νm,f = 0.216

Correlation (married males, married females): 0.216

• Persistent Shocks.

The Structure of Shocks

• Permanent Shocks.

Variance single males: σ2

νs,m = 0.255

Variance single females: σ2

νs,f = 0.226

Variance married males: σ2

νm,m = 0.220

Variance married females: σ2

νm,f = 0.216

Correlation (married males, married females): 0.216

• Persistent Shocks.

Common persistence: ρ = 0.958 – Kaplan (2012) Variance single males: σ2

ǫs,m = 0.005

Variance single females: σ2

ǫs,f ∼ 0

Variance married males: σ2

ǫm,m = 0.008

Variance married females: σ2

ǫm,f = 0.0006

Correlation (married males, married females): 0.44

0.000 0.100 0.200 0.300 0.400 0.500 0.600 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Age

Var-Log Male Earnings - Married

Model Data

Facts

Facts

• Current Population Survey (CPS), 1980-2005.
• Household heads and their spouses between ages 25 to 60;
• For earnings and hours → drop all observations with (i) hourly

wage lower than federal minimum wage; (ii) hours lower than than 520 hours per year;

• Two groups: skilled (college educated and higher) and

unskilled (less than college).

Facts

• Current Population Survey (CPS), 1980-2005.
• Household heads and their spouses between ages 25 to 60;
• For earnings and hours → drop all observations with (i) hourly

wage lower than federal minimum wage; (ii) hours lower than than 520 hours per year;

• Two groups: skilled (college educated and higher) and

unskilled (less than college).

• Consumption Expenditure Survey (CEX) → non-durable

consumption expenditure.

Facts

• Current Population Survey (CPS), 1980-2005.
• Household heads and their spouses between ages 25 to 60;
• For earnings and hours → drop all observations with (i) hourly

wage lower than federal minimum wage; (ii) hours lower than than 520 hours per year;

• Two groups: skilled (college educated and higher) and

unskilled (less than college).

• Consumption Expenditure Survey (CEX) → non-durable

consumption expenditure.

• For all variables, we estimate age effects controlling for time

(year) effects.

0.50 0.75 1.00 1.25 1.50 1.75 2.00 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Age

Hourly Wages, Males

ALL UnSkilled Skilled

0.30 0.35 0.40 0.45 0.50 0.55 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Age

Variance of Log Earnings, Males

ALL Married

0.40 0.42 0.44 0.46 0.48 0.50 0.52 0.54 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Age

Variance of Log Earnings, Females

ALL Married

0.30 0.35 0.40 0.45 0.50 0.55 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Age

Variance of Log Earnings, Married Males and Females

Married Males Married Females

0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Age

Variance of Log Household Earnings

ALL Married

0.08 0.09 0.10 0.11 0.12 0.13 0.14 0.15 0.16 0.17 0.18 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Ages

Variance of Log Yearly Hours, Females

ALL Married

‐0.05 0.05 0.1 0.15 0.2 25 30 35 40 45 50 55 60

Age

Var‐Log Household Consumption (Change)

Summary

Summary

• Hourly wages grow faster for skilled than for unskilled workers;

Summary

• Hourly wages grow faster for skilled than for unskilled workers;
• Variance of log earnings for all males increases non-trivially
• ver the life-cycle;

Summary

• Hourly wages grow faster for skilled than for unskilled workers;
• Variance of log earnings for all males increases non-trivially
• ver the life-cycle;
• For females, married or not, we do not observe such increase.

Summary

• Hourly wages grow faster for skilled than for unskilled workers;
• Variance of log earnings for all males increases non-trivially
• ver the life-cycle;
• For females, married or not, we do not observe such increase.
• Variance of log-hours is flat over the life cycle;

Summary

• Hourly wages grow faster for skilled than for unskilled workers;
• Variance of log earnings for all males increases non-trivially
• ver the life-cycle;
• For females, married or not, we do not observe such increase.
• Variance of log-hours is flat over the life cycle;
• The variance of log consumption increases over the life-cycle;

→ But much less than the increase in the variance of household earnings.

Decision Problem – Married Households

Decision Problem – Married Households

Let sM ≡ (if , im, q, b, νm,f , νm,m, θ), with if , im ∈ {s, u}.

Decision Problem – Married Households

Let sM ≡ (if , im, q, b, νm,f , νm,m, θ), with if , im ∈ {s, u}. Let η ≡ (ηm,f

j

, ηm,m

j

)

Decision Problem – Married Households

Let sM ≡ (if , im, q, b, νm,f , νm,m, θ), with if , im ∈ {s, u}. Let η ≡ (ηm,f

j

, ηm,m

j

) (sM, η) → ’exogenous’ states.

Decision Problem – Married Households

Decision Problem – Married Households

V M

j (a, h, e, η; sM)

= max

a′, lf , lm

{[UM

f (c, lf , q) + UM m (c, lm, q)]

+ βEV M

j+1(a′, h′, e′, η′; sM),

Decision Problem – Married Households

V M

j (a, h, e, η; sM)

= max

a′, lf , lm

{[UM

f (c, lf , q) + UM m (c, lm, q)]

+ βEV M

j+1(a′, h′, e′, η′; sM),

subject to (with kids) c + a′ + wud(if , j, b)χ(lf )

• child care costs

+ T M(I, b)

• taxes

− TRM(I, j, b)

• transfers

= wim̟m(im, j) exp(νm,m) + ηm,m

j

)lm(1 − τp) + wif h exp(νm,f + ηm,f

j

)lf (1 − τp) + a(1 + r(1 − τk))

Decision Problem – Married Households

V M

j (a, h, e, η; sM)

= max

a′, lf , lm

{[UM

f (c, lf , q) + UM m (c, lm, q)]

+ βEV M

j+1(a′, h′, e′, η′; sM),

subject to (with kids) c + a′ + wud(if , j, b)χ(lf )

• child care costs

+ T M(I, b)

• taxes

− TRM(I, j, b)

• transfers

= wim̟m(im, j) exp(νm,m) + ηm,m

j

)lm(1 − τp) + wif h exp(νm,f + ηm,f

j

)lf (1 − τp) + a(1 + r(1 − τk)) h′ = G(x, h, lf , e), a′ ≥ 0

Decision Problem – Married Households

V M

j (a, h, e, η; sM)

= max

a′, lf , lm

{[UM

f (c, lf , q) + UM m (c, lm, q)]

+ βEV M

j+1(a′, h′, e′, η′; sM),

subject to (with kids) c + a′ + wud(if , j, b)χ(lf )

• child care costs

+ T M(I, b)

• taxes

− TRM(I, j, b)

• transfers

= wim̟m(im, j) exp(νm,m) + ηm,m

j

)lm(1 − τp) + wif h exp(νm,f + ηm,f

j

)lf (1 − τp) + a(1 + r(1 − τk)) h′ = G(x, h, lf , e), a′ ≥ 0 with I ≡ ra + wim̟m(z, j) exp(νm,m) + ηm,m

j

)lm + wif h exp(νm,f + ηm,f

j

)lf

The Structure of Shocks

The Structure of Shocks

• Permanent Shocks.

Variance single males: σ2

νs,m = 0.255

Variance single females: σ2

νs,f = 0.226

Variance married males: σ2

νm,m = 0.220

Variance married females: σ2

νm,f = 0.216

Correlation (married males, married females): 0.047

The Structure of Shocks

• Permanent Shocks.

Variance single males: σ2

νs,m = 0.255

Variance single females: σ2

νs,f = 0.226

Variance married males: σ2

νm,m = 0.220

Variance married females: σ2

νm,f = 0.216

Correlation (married males, married females): 0.047

• Persistent Shocks.

The Structure of Shocks

• Permanent Shocks.

Variance single males: σ2

νs,m = 0.255

Variance single females: σ2

νs,f = 0.226

Variance married males: σ2

νm,m = 0.220

Variance married females: σ2

νm,f = 0.216

Correlation (married males, married females): 0.047

• Persistent Shocks.

Common persistence: ρ = 0.958 – Kaplan (2012) Variance single males: σ2

ǫs,m = 0.005

Variance single females: σ2

ǫs,f ∼ 0

Variance married males: σ2

ǫm,m = 0.008

Variance married females: σ2

ǫm,f = 0.0006

Correlation (married males, married females): ∼ 0

W lf S

4500 5000

Welfare System

3500 4000 3000

ers ($) married, children single female, children single male, no children

2000 2500

Transfe

1000 1500 500 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2

Household Income (as a fraction of mean household income)

• 0.1
• 0.05

0.05 0.1 0.15 0.2 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4

Multiples of Mean Income

Average Tax Rates

Married, 2 children Single, 2 children

0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5

25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Age

Hours/Worker, Females

Model Data

0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5

25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Age

Hours/Worker, Males

Model Data

0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Age

Correlation of Earnings (positive)

Model Data

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55

Age

Correlation of Wages

Data Model