2. Labour Supply And Payroll Taxes

Question 1

Payroll taxes

Answer 1

Comprise of:
Income taxes
Social security contributions

All modern economies use payroll taxes:
-to generate revenue for general expenditure (income taxes) - used for infrastructure, military, energy…
-to fund public programs (taxes and SSC) e.g. unemployment benefits, family benefits, pensions…
-to redistribute income

Question 2

Issue with payroll taxes

Answer 2

Introduce inefficiency (deadweight loss) into labour market:
-drive wedge between cost of labour to the firm (increase) and net wage of worker (decrease)

Drives employment below the optimal level
optimal level: employment without tax - no DWL

Question 3

Measures when classifying tax schedules

Answer 3

Average tax rate
Marginal tax rate

Question 4

Average tax rate

Answer 4

For each £ earned, how many pence are paid in taxes on average?

Question 5

Marginal tax rate

Answer 5

For the next £ earned, how many pence does the individual have to pay in taxes?

Question 6

Properties of tax schedules

Answer 6

Proportional
Progressive
Regressive

Question 7

Properties of tax schedules: Proportional

Answer 7

Independent of the level of income, every worker pays the same tax tax rate

Question 8

Properties of tax schedule: Progressive

Answer 8

The average tax rate is increasing in income

UK tax schedule is progressive

Question 9

Properties of tax schedule: Regressive

Answer 9

The average tax rate is decreasing in income

Question 10

Minimum Guaranteed Income Scheme (MGI)

Answer 10

Social assistance scheme topping up income to guarantee individuals a consumption/ income floor (subsistence income)

Effects: MGI increases the reservation wage - some low income individuals choose not to work and claim MGI - without MGI these individuals would have worked (extensive margin decision)

To offset effect that discourages employment - MGI programs often integrated with employment-conditional incentives

Question 11

Key questions

Answer 11

How high should taxes be?

Discussion so far: taxes distort labour supply decisions (reduce labour supply), so why do governments impose taxes? :
-need source of revenue to fund programs (transfers, public goods…)

Trade off: How to raise revenue for these programs while keeping distortions to a minimum?

Question 12

Optimal linear tax rate

Answer 12

Depends on social value of what the government does with taxes

if you believe social value of government tax revenue close to 0 —–> optimal tax rate is 0

Maximizing tax revenue not the goal of the policy maker

Question 13

Actual Incidence of taxation

Answer 13

Refers to economic burden of a tax, who actually bears the cost of the tax?
in our context:
-which side of the labour market bares the cost of taxes?
-who actually pays for the tax?

Question 14

Formal incidence of taxation

Answer 14

The formal incidence of taxation refers to the legal assignment of the tax burden, which is determined by the laws and regulations governing the tax. It is the party or entity, as specified by tax laws, that is responsible for paying the tax.

They may pass on the cost of the tax (actual incidence of taxation) to someone else, but they are the one that is technically required to pay for the taxation

Question 15

Income effect (tax context)

Answer 15

Taxes reduce net-income holding labour supply fixed
As individuals are poorer —–> work more hours to make up for lost income

Question 16

Substitution effect (tax context)

Answer 16

Taxes reduce price of leisure (forgone net wage)
You lose less from not working because you are not earning as much as you were before tax was put in place
—–> people may consume more leisure by reducing their hours worked

Question 17

How do individuals/ aggregated labour supply respond to income taxation?

Answer 17

If tax affects labour supply along the extensive margin —-> reduce participation (substitution effect)

If taxes affect number of hours worked, total effect on hours depends on relative size of income and substitution effect:
- substitution effect > income effect - progressive tax reduces labour supply
- income effect > substitution effect - progressive tax increases labour supply

Determining these effects and their size is an empirical question

Question 18

Three margins of labour supply adjustment in response to taxes (Meghir and Phillips, 2010)

Answer 18

Extensive margin
Intensive margin
Effort of worker

Question 19

Extensive margin

Answer 19

Whether to work or not

Question 20

Intensive margin

Answer 20

Hours worked conditional on working

Question 21

Effort of worker

Answer 21

Amount of effort that a worker puts forth can be affected by taxation

Differ by demographic and education group:
-Male hours of work not responding to male incentives
-Males with low/medium income respond to tax incentives
-women with young children / single mothers are sensitive to financial incentives at both the extensive and intensive margin
-Highly educated insensitive to work incentives

Question 22

Eissa, 1995 - US TAX REFORM ACT (1996)

Answer 22

Natural experiment that allows us to estimate labour supply responses to tax reform:

Description:
-large tax cuts for high income earners
-small tax cuts for low income earners

Study: compare women married to husbands with different income levels

Results:
-labour supply of high income married women increased at least half along the extensive margin
-married women may have more elastic labour supply

Caveat: three groups used to estimate responses are very different - especially high income group

Follow up:
-studies using more data and tax reform

result: found elasticities around 0.1 - 0.4

Question 23

Bastani and Waldenström, 2021 - ESTIMATE ABILITY GRADIENT IN TAX RESPONSIVENESS

Answer 23

How do responses to income taxation differ by cognitive ability?

Description:
-Estimate responsiveness to income taxation using a bunching estimator:
-use first “kink” point of Swedish central gov tax schedule —-> MTR increases by 20% at this point for income above 33K

Idea: individuals with income slightly above the threshold have an incentive to locate below it
-income distribution around the kink point should show excess mass below the threshold and missing mass above

Data: Linked administrative tax and military enlistment data for Sweden (focus on men)

Question 24

Bunching estimator

Answer 24

Based on idea that individuals and firms will adjust their behaviour in response to changes in tax incentives - such as tax brackets or tax credits - creating “bunches” of tax payers at certain income levels

Uses micro-data from tax returns to identify these bunches of taxpayers and compare the behaviour before and after the tax change - allows researchers to estimate behavioural responses of individuals and firms to changes in tax incentives - such as changes in labour supply, savings or investment

Goal:
-estimate the (compensated) earnings elasticity of individuals to changes in the net-of-tax rate i.e. how does people’s income respond to taxation?

Idea: use discrete increase in marginal tax rate at some income - creates convex kink in budget set

Question 25

Bunching estimator: identifying assumptions

Answer 25

1) Smoothness
2) Shape of counterfactual
3) Aggregation bias in elasticity e

Question 26

Bunching estimator: smoothness

Answer 26

Counterfactual distribution is smooth around the kink

Threats:
Several policies change at the same threshold - even absent the analysed policy - the distribution would not be smooth
Threshold serves as reference point - which people target in their estimation

Estimates are thus confounded - making it difficult to uncover causal effect of studied policy

Solution: obtain additional bunching observations - e.g. from variation in size of the discontinuity or over time that do not affect the confounding factors

Question 27

Bunching estimator: shape of the counterfactual

Answer 27

We estimate a specific shape of the counterfactual distribution (i.e. specific functional form)

Approximation might be very well locally (within tight windows around the threshold) but less so for policies leading to large behavioural responses (where individuals even far off from the threshold still respond)

Solution: sensitivity analysis - vary order of polynomial and the excluded range and assess robustness of your estimates