The state income data from the American Community Survey (ACS), released this morning by the Census Bureau, showed that in 2018, household incomes across the country rose—albeit more slowly, and in fewer states, than in the previous year. From 2017 to 2018, inflation-adjusted median household incomes grew in 33 states and the District of Columbia (14 of these changes were statistically significant.) This marks a decline from the broader growth seen between 2016 and 2017 when median household incomes grew in 40 states and the District of Columbia, with 24 of those changes being statistically significant.
The ACS data showed an increase of 0.2% in the inflation-adjusted median household income for the country as a whole—an increase of just $130 for a typical U.S. household and a slowdown in growth compared to the past three years: household incomes increased by 3.8% in 2015, 2.0% in 2016, and 2.5% in 2017. [i] Despite these increases, households in 23 states still had inflation-adjusted median incomes in 2018 below their 2007 pre-recession values, which makes this year’s slowdown particularly disappointing.
From 2017 to 2018, the largest percentage gains in household income occurred in Idaho, where the typical household experienced an increase of $2,085 in their annual income—an increase of 3.9%. Maryland remains the state with the highest median household income at $83,242, having experienced a slight increase (0.6%) from 2017 to 2018. The District of Columbia has the highest median household income in the country at $85,203—though comparing D.C. to states is problematic, since D.C. is a city, not a state.
From 2017 to 2018, there were 17 states in which the median household income declined or was unchanged: Alaska (-0.8%), Iowa (-0.1%), Maine (-3.6%), Missouri (-0.7%), Nebraska (-3.0%), Nevada (-1.3%), New Hampshire (-0.2%), New Jersey (-0.4%), New Mexico (-1.5%), North Carolina (-0.3%), South Dakota (-2.8%), Rhode Island (-1.7%), Tennessee (-0.4%), Virginia (-1.0%), West Virginia (-1.0%), and Wyoming (-0.5%). Only one of these—Maine, where incomes declined by 3.6% after increasing by 3.8% in 2017—had a statistically significant drop. 31 states and the District of Columbia saw either a slowdown in their growth compared to last year, or experienced even steeper declines, reflecting the disappointing growth seen at the national level.
The 0.2% increase in median household incomes reported in the ACS is an even weaker increase than the 0.9% that the Census Bureau reported earlier this month in their annual release of data from the Current Population Survey (CPS). It is not unusual for the two surveys to show slightly different values, although the weakness in growth reported in both surveys affirms that household income growth for middle-income households is slowing.[ii] This slowdown in growth not only affects middle-class households who, in many states, still have not recovered to their pre-recession income levels, but it also stalls progress in lifting the lowest-income families out of poverty.
[i] The ACS and CPS have different samples and cover slightly different timeframes, which can explain why the two surveys will differ. The CPS surveys all of its respondents in March of each year, and asks them to describe their income in the preceding calendar year. The ACS is a rolling 12-month survey—i.e., households are surveyed on an ongoing basis throughout the year and when surveyed, respondents are asked to report their income over the preceding 12 months. Thus, a significant portion of 2018 ACS respondents (those surveyed in the first half of 2018) were describing income mostly from 2017. Similarly, the 2017 ACS partially describes income in 2016. Therefore, the change in ACS incomes from 2017 to 2018 is describing changes occurring over a somewhat broader period that includes portions of 2016.
[ii] The increase we calculate from 2017 to 2018 is smaller than the 0.8 percent growth reported by the Census Bureau because they inflation-adjust the underlying microdata to calculate the real median household income for 2017. Because we do not have access to the microdata, we inflation-adjust the published 2017 median and all previous years’ medians. We do not use their measure of real median household incomes since it is only available back to 2014.