The Montana Budget and Policy Center’s Affordable Housing Map is a county level index that shows demographic information and rental housing affordability. The Housing Index is composed of 12 indicators. The primary data sources are the American Community Survey 2012-2016 five-year estimates and the National Low Income Housing Coalition’s 2017 Out of Reach report. The remaining data is sourced from Policy Map, an online database that publishes administrative data from the Department of Housing and Urban Development.
To determine the number of renters living at or below the poverty level, we used existing variables in the American Community Survey dataset. We counted the number of all renter households reporting income of $20,000 or less in the ACS dataset and divided that number by the total count of renters in the county. To determine the number of renters living at or below the poverty level with rent burden, we counted the total number of renter households reporting income of $20,000 or less who also reported paying more than 30 percent of income for rent.
Although the U.S Census Bureau issues poverty thresholds annually, thresholds vary by family size and composition. Due to data limitations, we could not break down households by multiple variables of housing tenure, family size, income, and rent burden on a county-level without losing data reliability. Therefore, we considered all renters under the $20,000 income threshold to be living in poverty and did not differentiate between different household compositions, although family size determines the poverty income threshold.
American Community Survey estimates are derived from a sample of the total population, and the sample for a subset within that county, such as cost-burdened low-income renter households, is even smaller. To provide a more accurate range of data values for indicators sourced from American Community Survey data, we included the confidence intervals at 90 percent margin of error the survey publishes alongside each data point.
For two of our indicators, Renters living in poverty and Housing cost burden for renters living in poverty, a number of counties had small sample sizes so low that its confidence interval was too large to be realistic. To account for this, we used a coefficient of variation (CV) to determine the accuracy of each county estimate. When data for one or more counties reported a CV score at or greater than 30 percent, counties were grouped and data aggregated to provide a more reliable estimate. When determining county groups, we combined contiguous counties when possible to avoid combinations with dissimilar population size and housing affordability estimates. In some cases, a neighboring county with a CV below 30 percent was included in groups in order to create a more reliable estimate.
Source: Table S2503. U.S Census Bureau, 2012-2016 American Community Survey five-year estimates.
Source: Table S0101. U.S Census Bureau, 2012-2016 American Community Survey five-year estimates.
Source: Table S1903. U.S Census Bureau, 2012-2016 American Community Survey five-year estimates.
Source: Table S1101. U.S Census Bureau, 2012-2016 American Community Survey five-year estimates.
Source: Table B25118. U.S Census Bureau, 2012-2016 American Community Survey five-year estimates.
Source: Table S2503. U.S Census Bureau, 2012-2016 American Community Survey five-year estimates.