As would be expected in a state with the nation's highest rate of unemployment in 2007, Michigan's poverty rate increased and its median income decreased from 2006 to 2007. However, the median income for full-time year-round workers continues to be significantly higher in Michigan than in the U.S. as a whole.

Median Earnings of Full-Time Year-Round Workers. The median income of workers employed full-time year-round in Michigan ranked 10th highest among the states for males and 18th highest for females in 2007. Although these rankings remain relatively high, they have decreased somewhat since 2000. For males, Michigan's rank has declined from #5 in the 2000 Census, #7 in 2005, and #9 in 2006. For females, Michigan's previous rankings have been #15 in the 2000 Census, #17 in 2005, and #18 in 2006. (See tables R2001 and R2002.) High levels of employment in the auto industry have contributed to Michigan's high rankings, and contractions of that employment have contributed to recent declines.

Overall Median Income. Despite relatively high medians for full-time year-round workers, overall median household income was 5.5% lower for Michigan than for the U.S. as a whole. Michigan was also the only state to have a statistically significant decrease in its median household income from 2006 to 2007, after adjustment for inflation. Michigan's median household income ranked 27th and its median family income ranked 25th among the states. (See tables R1901 and R1902.) Michigan's relatively low median income, despite high medians for full-time year round workers, reflects a tendency for people who are unemployed or underemployed for part of the year to bring down median levels of income. It also reflects Michigan's relatively high number of retirees, which is discussed below.

Poverty Rates. Michigan's poverty rate of 14% was the 17th highest rate in the nation in 2007, and Michigan was the only state to have a statistically significant increase in its overall poverty rate from 2006 to 2007. Michigan also ranks 17th in its percentage of related children below the poverty line. Michigan's poverty rate for people over age 65 is relatively low, however, ranking 37th highest among the states. (See tables R1701, R1702, and R1703.) A relatively high percentage of Michigan retirees have private pensions that supplement their Social Security benefits.

Geographic Distribution of Poverty. Nationwide, poverty rates are lowest in metropolitan counties (12.4%) and micropolitan counties (15.4%), and highest in the remaining non-metropolitan counties (16.5%). Michigan follows a similar pattern, with 13.9% poverty rates in its metropolitan and micropolitan counties and a 15.3% rate in its other non-metropolitan counties.

The poverty rates cited above for Michigan's micropolitan and other non-metropolitan areas are significantly lower than the corresponding national figures. The suburban and outlying portions of Michigan's metropolitan areas have a poverty rate of only 9.4%, which is equal to the corresponding national rate. However, the poverty rate for principal metropolitan cities is much higher for Michigan than for the U.S. as a whole--23.4% vs. 17.2%. (See Table B-1 in Income, Earnings, and Poverty in the United States: 2007.)

For example, Detroit has the nation's 8th highest poverty rate among the nation's cities with 65,000 or more residents, and it ranks first among cities with 250,000 or more residents. Kalamazoo, Flint, and Pontiac rank 5th, 6th, and 11th respectively among the nation's cities with 65,000 or more residents. (See Table 11 and Table 12 in Income, Earnings, and Poverty in the United States: 2007.)

In addition to reflecting high rates of unemployment, these high poverty rates in Michigan's principal cities reflect the fact that Michigan cities have been less able than cities in many other states to annex surrounding land when it becomes developed. Some cities in other states, such as Los Angeles, Houston, Phoenix, San Antonio, Indianapolis, Jacksonville, and Louisville, have two to five times as much land area as Detroit and exceed the size of several Michigan counties. Many smaller cities in other states have also been able to annex outlying areas as they become developed. Michigan's principal cities would have much lower poverty rates if they had also been able to annex their suburbs.

Thus, the high rankings of Michigan cities relative to those in other states are not necessarily very meaningful. Nevertheless, the fact remains that many of Michigan's principal cities have very high rates of poverty.

Impact of Military Service on Income and Poverty Statistics. Like past censuses, the American Community Survey inquires about the incomes of people who currently reside in each household. The incomes of family members who are absent for military service or other reasons are not included, even though such incomes may be the family's primary source of support. The deployment of military personnel away from their families helps to explain why national statistics on median income and poverty have not improved during a period of low national unemployment.

Public Assistance and Retirement Income. Public assistance income and retirement income are reflected in calculations of income and poverty.

Michigan ranks 5th in the nation with respect to its percentage of households with cash public assistance, surpassed only by Alaska, Maine, the District of Columbia, and Oklahoma. Michigan ranked only 15th in the 2000 Census. Nevertheless, it is important to note that the estimated number of Michigan households with public assistance income actually decreased from 3.6% of the total in 2000 to 3.2% in 2007. The change in the rankings reflects an improvement in the national economy between 2000 and 2007 that was not shared by Michigan. (See Table R1904.)

Michigan ranks 4th in the nation with respect to its percentage of households with pension income. As has been the case since 2005, Michigan is surpassed only by West Virginia, Delaware, and Hawaii. (See Table R1903.)  Michigan ranked 6th in the 2000 Census, and its high rankings throughout this period reflect a substantial number of people with private pensions who have taken early retirement.

Perhaps the most notable aspect of the retirement data is that Florida's rank dropped from 4th to 12th between the 2000 Census and the 2007 ACS, while Arizona dropped from 9th to 15th and Nevada dropped from 17th to 33rd. These changes may reflect a difference in the way the data is collected rather than a change in the actual residence patterns of retirees. Censuses are intended to count seasonal migrants at the location where they spend the largest part of the year, but the 2000 Census did not include instructions to that effect. Thus, the 2000 Census probably tended to count people where they happened to be when they received census forms in March or April, rather than at their usual residences. The American Community Survey also counts people where they are, but it is conducted every month throughout the year. Thus, the ACS reflects each state's average percentage of households with retirees rather than the percentage in late winter or early spring.

Any tendency for censuses to count seasonal migrants at their winter residence rather than (or in addition to) their usual residence has serious implications for the allocation of federal funds, apportionment, and the calculation of appropriate population controls for ongoing surveys such as the ACS and the CPS. The ACS has great potential for producing information that would be useful for calculating seasonal population estimates and estimates of average population levels and characteristics over the course of each year.

Created August 27, 2008