Simulation of Kearney and Turner's Secondary Earner Proposals

This page is available to simulate various parameter combinations for a secondary earner deduction proposed by Melissa Kearney and Lesley Turner in their paper "Giving Secondary Earners a Break: A Proposal to Help Low and Middle INcome Families", done for the Hamilton Project December, 2013.. The simulation methodology is similar to Feldstein and Feenberg "The Taxation of Two-Earner Families".

   

Secondary Earner Deduction Percentage
Secondary Earner Limit

Exclude SED from list of AMT preferences
Include SED in list of AMT preferences

% Point(s) to reduce deduction rate for each $1,000 of AGI above threshold

AGI threshold for start of clawback

Haircut on secondary personal exemption

Elasticity of earnings wrt after-tax wage
For Earnings of Taxpayer
For Earnings of Spouse
For Participation of Spouse

Income Effect (de/dy)

Use data from file:

Notes on the form:

  1. The default calculation includes a SED with a phaseout. Set either percentage to zero to see suppress that feature.
  2. The default sample is an aged 2% subsample of the 2007 PUF. To get the full sample, change the "s" to "x" in the file box. Earlier years are also available - just change the digits.

About the imputation of secondary earnings

Take 3 random numbers, r1, r2, and r3 uniform on 0,1 and some taxpayer specific parameters based on kids, assets and total wages:
pr_secpart025: conditional on working, probability secondary earner earns less than 25% of total wages for the couple.
pr_secpart2550: conditional on working, probability secondary earner earns 25% or more of total wages for the couple
two025: probability the lower earning spouse earns less than 25% of total wages
two2550: probability the lower earning spouse earns more than 25% of total wages

The "pr_secpart" values add up to 100%, the "two" variables add up to total participation. The secondary earner imputation is without regard to the actual credit for child care, which is (in law) dependent on spouse earnings.

Pseudo-code for imputation

if r1 < pr_secpart025 then use r2 to assign a random wage less than a quarter of total wages else if r1 <pr_secpart2550 then use r2 to assign a random wage between a quarter and a half of total wages. else assign a random wage of zero if r2 < two025 assign a random potential wage less than a quarter of total wages else assign a random potential wage between a quarter and a half of total wages endif endif

The effect is to make the imputation in two steps. In the first step we assign a probability of working to each secondary worker and a wage to each secondary worker, including zero wages for non-workers. This wage is used to caculate the base tax revenue and the change in revenue from the change in effort by current primary and secondary workers.

In the second stage we assign a potential wage to non-workers. This is always non-zero. This potential wage is used to calculate the additional tax from secondary workers entering the labor force.

The change in wages for current workers is based on the formula:

delta_wages = elasticity*wages*fciats - .15*static_tax_change where fciats is fractional_change_in_after_tax_share. This is applied separately to primary and secondary workers, and the wages and static tax changes are allocated between them according to the imputation above.

For the change in participation,

delta_participation = elasticity*probability_participation*fciats

If this change exceeds r3, then the imputed potential wage is used to calculate the post reform earning and tax revenue. There is an option to use the finite difference over projected secondary earnings to calculate the fciats.

Last update March 11, 2013 by feenberg@nber.org