The RAND HRS Tax Calculations 2014 (V2), supported by NIA and SSA

The RAND HRS Tax Calculations 2014 (V2) contain information about federal, state, and FICA taxes for Respondents to the HRS 2000 – 2014 surveys. The NBER Internet TAXSIM1 calculator is used to perform these calculations (https://users.nber.org/~taxsim/taxsim-calc9/index.html).

Background

HRS Respondents are generally asked to report their income before taxes and any other deductions. However, for a number of research purposes, after-tax income would be the most appropriate concept to use in analyses. For example, when studying saving behavior, the question of interest is how much a household saved out of the household’s disposable income (i.e., the household’s after-tax income). Similarly, if one is interested in economic well-being, disposable (i.e., after-tax) income would presumably be a more accurate proxy than pre-tax income.

Since the amount of taxes paid is not observed in the HRS, we use the TAXSIM calculator to compute taxes for HRS households. We construct the necessary input variables from self-reported information recorded in the HRS surveys.

Input Variables

TAXSIM requires the user to provide 22 input variables for each case where taxes are to be computed. We create these variables, and submit them to the TAXSIM calculator for each Respondent in the household. Those whose marital status is defined as "married" or "married, spouse absent" are treated as filing jointly, and as single otherwise. The input variables were constructed using information from the RAND HRS Detailed Imputations File, the RAND HRS Fat Files, and the RAND HRS Longitudinal File.

In some cases, the HRS survey does not have sufficient information to construct a particular input field in the best way. In other cases, the HRS asks about the relevant information, but the Respondent did not answer. In those situations, we made certain assumptions in order to fill in the missing information. A description of some of the assumptions that were made while developing these data can be found in the "General Notes on File and Variable Construction" section of the RAND HRS Tax Calculations 2014 (V2) codebook. The codebook also includes detailed information about how each of the 22 input variables was constructed, as well as descriptions of the TAXSIM output variables.

Output Variables

The files include information about federal, state, and FICA taxes for respondents to the HRS 2000 – 2014 surveys. Some researchers may also be interested in marginal tax rates (i.e., how much would the household pay in taxes if its income increased by a given amount), which we have also included in the files. Because it is difficult to anticipate what amount of additional income is a meaningful margin to study, we performed multiple computations assuming increments of $1k, $5k, $10k, and $50k. The output variables for these computations are included in the files, and are labeled as “finite difference.”

Supplemental Data

A large number of variables were used in deriving the HRS input variables. Some users may be interested in changing the derivation of one or more of the input fields. For example, users may wish to:

  • replicate the calculations
  • see what went into the calculations (i.e., transparency)
  • rerun the calculations with alternative assumptions about tax year, tax rates, etc.
  • modify the input variables, either to obtain alternative scenarios, or to adjust any assumptions we may have made to make them more appropriate for a particular research context.

To facilitate this, we have included an additional set of files that contain all the variables used in the derivations.

File Structure

The RAND HRS Tax Calculations 2014 (V2) come in eight separate files for the years 2000, 2002, 2004, 2006, 2008, 2010, 2012, and 2014. For confidentiality reasons, the calculations are performed 51 times for each Respondent, once for each state. Therefore, a Respondent who participated in a specific wave of the HRS survey will have 51 records in the associated file. Users of the restricted geocodes can select the record that corresponds to the state the Respondent lived in at the time of the interview.

Each of the files (including those that contain the supplemental data described above) can easily be merged with the RAND HRS data products (by HHIDPN), or with the Core HRS data (by HHID and PN). These identifiers are included in the files.

Data Access

  • All RAND HRS data products are available at: https://hrsdata.isr.umich.edu/data-products/rand.
  • The RAND HRS Tax Calculations 2014 (V2) can be found in the "Longitudinal and Cross-Wave Data Products" section on the right-hand side of the page.
  • You will be able to see the data descriptions and documentation without an account, but you will need to register with HRS to access the data.
  • User note: This is a new data downloads system as of September 1st, 2020. If you have an account on the old system but have not created a new account since September 1st, 2020, you will need to create a new account. You can use the same email address as your previous account. (You will not need to create a new account if you have used this system to apply for or manage an HRS Restricted Data Agreement since November 2019.)
  • Once your account is active and you have logged in, you will be able to see the data download links for each RAND HRS data product (and other public HRS data products).

For more detailed information, please see the documents:

For more information, questions or comments please see our FAQ or email us at RANDHRSHelp@rand.org.

Before doing so, however, we kindly request that users first consult the documentation that accompanies our data products, as we have found that our responses often point users to specific sections of the documentation that provide further detail on the variables mentioned in the users’ queries.

In addition, we recommend that users become familiar with some of the information provided on the HRS website, such as the HRS questionnaires and codebooks for the key variables under study.

We have also found the concordance tool extremely useful to help find available variables across waves:

Footnote

  1. Feenberg, D.R., and Coutts, E. (1993), An Introduction to the TAXSIM Model, Journal of Policy Analysis and Management, 12(1), 189-194. (Return to text)