Increased attention has been focused on the decline in poverty among older persons in recent years. Comparing the poverty rate in 1997 with 1989, older persons were the only age group whose poverty rate decline (from 11.4 percent in 1989 to 10.5 percent in 1997) was statistically significant. Since then, it has declined still further to 9.7 percent in 1999. The decline in the poverty rate for older persons, however, does not present a complete picture of the poverty experience of older persons. The official poverty measure in the U. S. is based on the March Current Population Survey (CPS), an annual cross-sectional survey. However, one cannot tell from cross-sectional data whether the povertythis population contains the same individuals from year to year or whether it consists of a much higher number of of many morindividuals, each of whom spends only a relatively short period (i.e., only one year) living in poverty.
This paper presents the findings from a study of the poverty experience of individuals from a long-term perspective, focusing particularly on older persons. The study usedexamines a longitudinal survey, the Panel Study of Income Dynamics (PSID), to track long-term poverty status for the same individuals over the 1981-1992 period, (1982-1993) when the annual official poverty rate for older persons was lower than for the entire population. We applied a survivor analysis method to estimate the probability of exiting poverty and the distribution of completed poverty spells for older and younger persons in the 12-year period.
Exit Probabilities From Poverty Years in Poverty For Older Persons and Younger Persons
Table 1 compares a distribution of poverty exit probabilities by length of poverty spell for persons age 65 and above and non-older persons under age 65. The results indicate that escaping from poverty was more difficult for older than for younger age groups, regardless of the length of poverty spell. For example, for an older person who was in a poverty spell of one year, the probability of exiting from poverty was 35.2 percent compared to 40.3 percent for a person under age 65.
(Table 1. Probability of Leaving Poverty Status by Giving Different Length of Year Spent in Poverty for Older Persons and Persons Under Age 65, 1981-1992.)
Although poverty exit probabilities for older persons were lower than for younger persons, the pattern of poverty exit probabilities in the first three years of a spell is somewhat similar for older and younger persons in that the probabilities drop at comparable rates as the length of poverty spells increases.
After the first three years in poverty, the exit rate for an older person was substantially lower than that for a younger person (7.3 percent for the older persons, compared with 21.3 for younger persons). For an older person who spent more than four consecutive years in poverty, the average of exit probabilities for the following five years was less than 5 percent per year. In contrast, for a person under age 65 who spent the same number of years in poverty, the average of exit probabilities for the following five years was about 15 percent (see Table 1).
In other words, the majority of the older persons who spent over three consecutive years in poverty will stay in poverty for a long time, and some of them may never escape. This is a major distinguishing characteristic of the poverty experience for individuals in their older years as compared with that of younger individuals (Coe, 1988).
Distributions of Completed Poverty Spells for Older and Younger Persons
Table 2 presents percent distributions of completed poverty spells for those ever poor and for those poor at a point in time for older and non-older younger persons. Columns 2 and 3 show that over one-third (35.2 percent) of older persons who were ever poor completed their poverty spell in one year, and about 58 percent of older persons who were ever poor remained in poverty three years or fewer. On the other hand, nearly 31 percent of completed poverty spells of older persons who were ever poor can be expected to last ten or more years.
(Table 2. Percentages Distribution of the Length of Completed Poverty Spells for Older Persons and Persons Under Age 65, 1981-1992.)
By comparison, over 40 percent of the younger persons who were ever poor completed their poverty spells in one year. Nearly 69 percent of younger persons' completed poverty spells lasted three years or fewer. Only 11.2 percent of poverty spells completed by those under age 65 can be expected to last ten or more years (see Table 2).
These differences again reveal the major distinguishing characteristic of the poverty experience of older persons. For both older and younger persons, a substantial fraction of poverty spells last three years or fewer. After three years, however, the situation is significantly different. Among older persons, 42 percent remain poor after three years and 31 percent remain poor for ten years or more. Thus, older persons who remain poor for more than three years only have one chance in four of getting out of poverty in less than ten years (i.e., 11/42,) while younger persons who stay in poverty for more than three years have a two in three chance of getting out of poverty (i.e., 20/31).
The longer-term poor tend to accumulate on the poverty rolls because they are unable to escape from poverty. Thus, at any point in time, the longer-term poor will comprise a disproportionately large fraction of the current poverty population. Over time, people having long-term poverty spells make up a larger and larger proportion of the poverty population. Columns 4 and 5 in Table 2 report the distribution of completed poverty spells for persons who are poor at a point in time. The results show that among older people who are identified as poor in a given year, 75 percent are in the midst of a poverty spell that can be expected to last ten or more years and 5.7 percent of the older poor will be in the midst of a poverty spell that will last just one year. In contrast, among the younger poor in a given year, only 43 percent are in the midst of a poverty spell that will last ten years or more, and 10.4 percent of the younger persons will be in the midst of a poverty spell that will last one year.
The Difference Between Poverty Experiences of Older Men and Older Women
Older women have different poverty experiences from older men.
Table 3 shows that the poverty spell exit probability for older women was lower than that for older men, regardless of length of spell experienceds. For example, the exit probability was 44.6 percent for older men but only 32.1 percent for older women who had spent one year in poverty. (see Table 3). For both older men and older women, a relatively high exit probability existed for the first three consecutive years in poverty. After these three consecutive years, however, the exit probability for an older woman fell substantially below that for an older man. For example, after five consecutive years in poverty, the exit probability for an older man was 16.1 percent, but only 3.3 percent for an older woman.
(Table 3. Probability of Leveaing Poverty Status by Giving Different Length of Year Spent in Poverty by Sex for Persons Age 65 and Over, 1981-1992.)
Even though the poverty situation for older persons looks better in the cross-sectional survey (the March CPS) in recent years than it did in the 1970s, older people still experience higher rates of long-term poverty than younger age groups do. The characteristic long-term poverty experience soffor older persons found in this research is similar to that found in Coe's study (1988), which examined 1970s PSID longitudinal data, when the annual official poverty rate for older persons was higher than that of the entire population. For older persons, especially older women, poverty spells are either relatively short or extremely long. The majority of the older persons who spent over four consecutive years in poverty will stay in poverty for a long time, and some of them will remain poor until death. This research shows that, despite lower poverty rates among older persons, they are more likely than younger persons to stay poor for a long time.
- The PSID is a longitudinal survey of a representative sample of U. S. individuals and the family members with whom they reside, which has been ongoing since 1968. The data files contain the full span of information collected annually over the course of the study.
- Like the March CPS, the PSID defines an individual's poverty status as annual family income below an income/need ratio of 1.0, which is essentially equivalent to the poverty threshold used by the Census Bureau to calculate the annual official poverty rate.
- A PSID longitudinal data file (from 1981 to 1992, 12-year period) was constructed for this study. The unit of analysis was the individual. The sample included 11,412 individuals. 1,340 of whom were age 65 and over in 1992.
- The survivor analysis method was used by Bane and Ellwood (1984) and Coe (1988) to estimate the exit probability from poverty and completed poverty spells.
- An exit probability from poverty is the probability that a person will escape from poverty after having a poverty spell. Spells are measured here in units of one or more consecutive years, but two spells must be separated by at least one year out of poverty. Since individuals might experience two or more distinct poverty spells during the 12-year period, each spell would be included in the calculation of exit probabilities.
- Poverty status for older persons measures only the individuals' poverty experience in their older years (age 65 and above). That is, no cases of poverty spells overlap the two age groups.
- These calculations are based on exit probabilities for length of poverty spells 5 to 9 years in Table 1.
- The distribution of completed poverty spells is generated from exit probabilities (Table 1).
- A point in time here refers to any given year.
Bane, MaryJo and David Ellwood. 1986. “Slipping Into and Out of Poverty: The Dynamics of Spells.” The Journal of Human Resources, 21, No. 1: 1-23.
Coe, Richard D. 1988. “A Longitudinal Examination of Poverty in the Elderly Years.” The Gerontologist, 28, No. 4: 540-544.
U.S. Bureau of the Census. 1998. Poverty in the United States: 1997. Current Population Reports, P60-201, Washington, D.C.
Written by Ke Bin Wu, AARP Public Policy Institute
May be copied only for noncommercial purposes and with attribution; permission required for all other purposes.
Public Policy Institute, AARP, 601 E Street, NW, Washington, DC 20049