Despite the fact that 2004 represented the third full year of economic recovery, the Census data released today show that poverty increased again last year and median income failed to rise.  The new data are particularly troubling for workers, showing backward movement for American workers on several fronts[1]:

      Real median earnings of full-time year-round male workers fell by nearly $1,000 (from $41,761 to $40,798), a decline of 2.3 percent.

• Real median earnings of full-time year-round female workers fell by over $300, or 1 percent (from $31,550 to $31,223), marking the second consecutive year of decline.  This is the first time since 1995 that the median earnings of full-time year-round female workers have dropped for two years in a row.
• Real median income among the working-age population — households headed by adults under 65 — fell by $600 (from $51,559 to $50,923), a decline of 1.2 percent.  (Overall median income for all households was unchanged.)
• The number of people who work but live in poverty increased by 563,000.  The poverty rate increased among this group from 5.8 percent to 6.1 percent.
• Among people age 18 to 64 who work, the number who were uninsured climbed by 772,000, and the percentage without insurance rose from 18.6 percent to 19.0 percent.
• The percentage of people with employer-based health insurance coverage fell below 60 percent — to 59.8 percent — for the first time since 1993.

The data show that the current recovery has been slow to help low- and middle-income families.  Four million more people were poor in 2004 than in 2001, when the economy hit bottom, and 4.6 million more were uninsured.  This three-year poverty trend is worse than in all previous recoveries over the last 45 years, except for downturn of the early 1990s.[2] Median income is about equal to its level in 2001, but remains significantly below pre-recession levels.  Income growth between 2001 and 2004 was slower than during all previous recessions over the last 45 years, again with the exception of the downturn of the early 1990s.  

In short, the Census data provide new evidence that, as in 2002 and 2003, the recovery in 2004 was neither robust nor broad-based.  Developments in early 2005 point toward continued problems, with real wages failing to improve according to the data currently available.

Robert Greenstein, executive director of the Center on Budget and Policy Priorities, noted:  "These data are the latest evidence that the economic recovery is neither strong nor broad-based and is leaving many Americans behind.  2004 was the third full year of economic recovery.  Yet among working-age Americans — adults under 65 — poverty increased, income fell, and both the percentage and number of people who were uninsured climbed further.'

Greenstein added that, "In September, Congress intends to cut taxes again primarily for high-income Americans, and to cover a portion of the tax-cut costs by reducing benefits and services, like Medicaid and food stamps, for low-income working families who are not faring well in this economy.  Congress should not be pursuing policies that take these adverse trends in poverty, income, and health insurance and make them worse.'

This short analysis contains seven observations about the new Census data.

1.  The poverty rate in 2004 remained significantly higher than in 2001, the year of the recession.  The number of people in poverty increased from 32.9 million in 2001 and 35.9 million in 2003 to 37 million in 2004.  The poverty rate rose from 11.7 percent in 2001 (and 12.5 percent in 2003) to 12.7 percent in 2004.  The rise in poverty in 2004 is particularly disturbing because 2004 represented the third full year of the economic recovery.

Contrary to the impression left by a Census official today, this three-year poverty trend is not typical for recoveries.

• In no other downturn over the past 45 years did poverty increase between the second and third full years of the recovery.
• In all other downturns except that of the early 1990s, the poverty rate by the third year of the recovery was at or below the poverty rate in the recession year itself.  In 2004, by contrast, the poverty rate was a full percentage point higher than in 2001, the recession year. [3],[4]

This recovery stands out (as does that in the early 1990s). Three years of economic growth did not improve the circumstances of low-income Americans.

2.   Overall median income in 2004 was unchanged as compared to 2003, but the income of the typical non-elderly household fell in 2004 as compared to 2003.  The real median income of non-elderly households was $50,923 in 2004, almost $600 below the 2003 level and $1,600 below the level in 2001, the recession year.

3.  Income growth during this recovery has been worse than all prior recoveries of the last 45 years, with the exception of the recession of the early 1990s.  In each of these prior recoveries (except the 1990s), median income in the third year of the recovery exceeded income in the recession year. In each of the four recessions prior to the 1990s, median income grew by between 1.8 percent and 8.3 percent from the recession year to the third year of the recovery.  In this recession, median income has fallen by 1.5 percent since the recession year.[5]

In addition, in all of these prior recoveries (even including the recovery of the 1990s) overall median income grew between the second and third full years of the recovery.

4.  The number of Americans who are uninsured climbed again in 2004.  The percentage of people who are uninsured remained unchanged, but at a level well above the percentage in 2000, prior to the recession.   Some 45.8 million people were uninsured in 2004, an increase of 800,000 people over 2003.  The percentage of the population without insurance remained unchanged at 15.7 percent.  Since 2000, the number of uninsured has increased by 6 million, and the percentage of Americans without insurance has risen from 14.2 percent to 15.7 percent.  The Census data show that the proportion of Americans lacking health insurance would have been even higher if it were not for the increase in the number of people covered by public health insurance, most notably Medicaid.

5.  &Some groups experienced losses in 2004.  Some groups and some parts of the country experienced increased poverty and deteriorating living standards in 2004.

• Among regions, the Midwest fared worst.  The poverty rate in that region increased from 10.7 percent to 11.6 percent, and the number of people living in poverty climbed by 600,000.  In addition, real median household income in the Midwest fell by $1,277 (from $45,934 to $44,657), or 2.8 percent.  Poverty increased or median income fell in some other regions, but those changes were not statistically significant.
• Among age groups, people aged 18-64 fared worst.  The number of people aged 18-64 who live in poverty rose by 1.1 million, while their poverty rate climbed from 10.8 percent to 11.3 percent.  (For children under 18, the poverty rate was unchanged at 17.8 percent; for those 65 and over, the poverty rate declined, from 10.2 percent to 9.8 percent.)  The number of people aged 18-64 who were uninsured increased by 754,000, and the percentage of these people who lack insurance increased from 20.2 percent to 20.5 percent.  Finally, as noted earlier, median household income declined by about $600, or 1.2 percent, for households headed by a person under 65.

6.  Inequality, and the severity of poverty, remained at record levels.  The Census Bureau's summary measure of income inequality shows that household income gaps, while not statistically different from the gaps in 2003, matched the previous record high set in 2001.  (These data go back to 1967.)  Similarly, the average amount per person by which people who are poor fall below the poverty line stood at $3,100.  This, too, matched the highest on record (with data available back to 1975).

7.   Picture for 2005 is mixed.  Early indications show that employment and some measures of income growth have been faster in the first half of 2005 than in the first half of 2004, but real wages (that is, wages adjusted for inflation) have fallen so far in 2005.

End Notes:

[1] Unless otherwise noted, all increases or decreases in poverty, health insurance, or income discussed in this report represent statistically significant changes.

[2] We compared trends for the economic recovery periods beginning in 1961, 1970, 1975, 1982, 1991, and 2001.  We omitted the brief recovery that began in 1980 because it lasted less than three years.

[3]  In conducting this comparison to previous recoveries, we measured growth rates in poverty and income over the first three years of each recovery, from the low point in the business cycle.  As noted previously, we omitted the brief recovery that began in 1980.  For recoveries that began part way through a calendar year — that is, before the fourth quarter of the year —  we used a weighted average approach that incorporates data from the previous year.  Thus, for the recovery that began in March 1991, we treated the starting year as the twelve months ending in March 1991, which we calculated as [.75 times the poverty rate in 1990] plus [.25 times the poverty rate in 1991].

[4] Because economic recoveries differ in their severity, some prefer to examine economic cycles by comparing the current year to the previous peak, that is, the year before the last recession began.  Using this approach and comparing the third year of the economic recovery to the previous peak year, the poverty rate grew faster than the average growth in previous economic cycles and grew faster than any other prior economic cycle, except that of the early 1990s.

[5] For details of how this analysis was done, see footnote 2.

Commentary: