While the goal of SAIPE is providing estimates of numbers of people in poverty in various groups, for many purposes poverty rates are more readily interpreted. We provide these rates but regard them as inferior to the estimates of numbers of people in poverty because of the unknown quality of our estimates of the required denominators. We provide what we regard as "illustrative" confidence intervals around the poverty rates, which are computed as if the poverty universe estimates were "true", i.e., without error.

The state models estimate ratios of number of people in poverty to population, as measured in the American Community Survey (ACS), for the groups of interest. We convert these ratios to estimates of numbers of people in poverty by multiplying by demographic estimates of the population, as covered by the ACS, for these groups. The county models directly estimate logarithms of numbers of people in poverty for the groups of interest. The computation of poverty rates corresponding to the model-based state and county estimates of numbers of people in poverty requires estimates of the number of people in the relevant poverty universes. Because the poverty numbers are consistent with the ACS definitions, the poverty universes must also be. The ACS universe for 2005 does not include group quarters populations, such as residents of nursing homes, college dormitories, correctional institutions, or other group quarters populations. Also, children under the age of 15 who are not related to the reference person within the household by birth, marriage or adoption (for example, foster children) are not included in the poverty universe, and so are neither in poverty nor not in poverty. Procedures for computing the 2005 poverty universe estimates at the state and county levels are described below.

For information on denominators used before 2005 see Denominators for State and County Poverty Rates: 1993 - 2004.

We derive state level estimates of the poverty universes in four steps:

1. We begin with state-level postcensal household population estimates provided by the Census Bureau Population Estimates Program. Household population estimates differ from resident population estimates in that they do not include group quarters populations.
2. Second, the postcensal household population estimates are raked to the national ACS population, by age group, through a simple ratio adjustment. For age groups that have a direct population control in the ACS, the ratio equals exactly one.
3. Third, we compute the ratio of the ACS poverty universe to the ACS population directly from the ACS survey estimates for each state. We apply these ratios to the demographic-based ACS population estimates calculated in step two.
4. Finally, we rake the resulting state estimates of the poverty universe, by age group, to the national ACS poverty universe through a simple ratio adjustment.

We derive county level estimates of the poverty universes in four steps:

1. We begin with county-level postcensal household population estimates provided by the Census Bureau Population Estimates Program. Household population estimates differ from resident population estimates in that they do not include group quarters populations.
2. Second, the postcensal household population estimates are raked to the state-level household population estimates, by age group, through a simple ratio adjustment. The state-level household population estimates used for raking are those produced in Step 2 from "State Level Estimates" above.
3. Third, we tabulate the ratio of the poverty universe to the household population for each age group in each county from the most recent decennial census. We apply these ratios to the demographic-based ACS population estimates calculated in step two.
4. Finally, we rake the resulting county estimates of the poverty universe, by age group, to the state-level ACS poverty universe estimates through a simple ratio adjustment. The state-level poverty universe estimates used for raking are those produced in Step 4 from "State Level Estimates" above.