Experimental Poverty Measures: Summary of a Workshop (2005)

Session 3: Incorporating Medical Out-of-Pocket Expenses (MOOP)

Recent reports that have published experimental poverty measures, P60-219 and P60-222, have used three approaches to incorporating MOOP. The first subtracts MOOP from income, the second incorporates MOOP into the thresholds, and the third combines the first two, adjusting both income and the threshold.

The Census Bureau appears to have adopted a three-parameter equivalence scale to adjust thresholds for differences in family size. A quick overview of why the three-parameter scale was chosen and a discussion of related research priorities for the future may be helpful. Units of analysis for the measure will be discussed in this session, including the question of how to treat cohabitators, foster children, and roomers and boarders.

Session 5: Geographic Adjustments

The Census Bureau has been producing estimates that are adjusted for geographic differences based on differential housing costs and that are not adjusted for geographic differences. Is the Fair Market Rents method the most appropriate? Are there viable alternatives to Fair Market Rents as a basis for the adjustment? Are the methods used to adjust for geographic differences technically sound?

Session 6: Incorporating the Value of Housing

(a) Imputing Rent for Owner-Occupied Housing

(b) Valuing Housing Subsidies

The NAS report stressed the importance of accounting for the flow of services homeowners obtain from their home in counting resources, but noted limitations in data and measurement that made this impractical for poverty measurement. What new data and methods are available to impute rent or otherwise account for home ownership?

The second experimental poverty measure report, P60-216, includes measures using two alternative approaches to valuing housing subsidies, one based on Fair Market Rents and the

other based on the 1999 American Housing Survey. What are the strengths and limitations of these approaches?

Continental Breakfast

Session 7: Issues for the Poverty Thresholds

What are the different methodological options for setting the thresholds? What are the technical implications for each option? How do alternative methods to account for other components of the poverty measure interact with these options for setting the threshold? What are the implications of these options for the different uses of poverty measures (e.g., for administrative and policy purposes or for statistical purposes)?

Given these different options, how can the thresholds be updated? Should CPI or CE or other surveys be used for updating the poverty thresholds? How often should the thresholds be updated?

Session 8: Data Issues

(a) What adjustments ought to be considered for SIPP, if we were to consider moving away from CPS as the source of official poverty statistics in the future?

(b) What are the strengths and weaknesses of applying CE data to poverty measurement?

(c) How often do other data sources used in alternative poverty measures (the CE and the MEPS) need to be updated?

(d) What are the data needs for non-income surveys that also collect data relevant for poverty measurement research?

(e) How can the American Community Survey be used to estimate poverty at the state and local levels and what are the complications involved with using the ACS?

(f) What are the issues raised by the need for state-level estimates?

(g) How does the problem of underreporting of income interplay with alternative methods?

Session 9: Leftover topics

This session will be devoted to the discussion of topics that have arisen during the workshop and that are important overall, but were not precisely relevant to the sessions in which they arose.