August 14, 2018
For millions of retirees, Social Security is more than just a part of their retirement income ? in many cases it is the vast majority of their retirement income. Social Security benefits account for half of family income for roughly 50 percent of those over the age of 65. The percentage of families in which Social Security benefits account for at least half of their retirement income rises to 70 percent for blacks and Hispanics. Benefits are roughly 90 percent of income for one in four of those 65 or older.
In particular, people of color, women, and people at the lower ends of the income distribution rely on Social Security more than other groups. In this piece, we use data from the 2015 March CPS to take a look at the share of Social Security income for groups 65 and older by race, gender, and marital status.[1]
In aggregate, Social Security benefits are the largest share of income for those in the first and second quintiles (80.8 and 79.4 . respectively). For the third quintile, Social Security is 59.9 percent of income, while for the fourth and fifth quintiles, benefits are 38.4 and 15.1 percent of income, respectively.[2]
Overall and among married couples, Social Security is 28.1 percent of total income. For the first and second quintiles, however, Social Security was well over 50 percent of retirement income for all groups, other than Asians in the first quintile (45.8 percent). Though most married couples in the third quintile rely on Social Security benefits for between 41.9 and 49.0 percent of their income, the share rises to 63 percent for Hispanic couples.
Race |
Quintile 1 |
Quintile 2 |
Quintile 3 |
Quintile 4 |
Quintile 5 |
---|---|---|---|---|---|
Hispanic |
64.8 |
81.9 |
62.6 |
36.4 |
14.7 |
White alone |
82.7 |
63.7 |
41.9 |
26.4 |
11.0 |
Black alone |
77.8 |
77.3 |
49.0 |
32.7 |
13.3 |
Asian alone |
45.8 |
75.3 |
48.3 |
29.1 |
12.2 |
Other non-Hispanic |
76.5 |
62.9 |
46.8 |
27.2 |
10.2 |
Source: Author’s calculations using 2015 March CPS
Among single men in the first quintile, Social Security was 92.6 percent of retirement income for those in the “Other, Non-Hispanic” category. For all other men in the first and second quintiles, Social Security was between 71 and 85 percent of income.
Race |
Quintile 1 |
Quintile 2 |
Quintile 3 |
Quintile 4 |
Quintile 5 |
---|---|---|---|---|---|
Hispanic |
71.5 |
82.0 |
70.4 |
44.1 |
16.6 |
White alone |
85.2 |
83.0 |
57.2 |
36.1 |
14.2 |
Black alone |
75.2 |
85.6 |
77.1 |
43.6 |
16.3 |
Asian alone |
75.7 |
82.3 |
69.4 |
40.8 |
13.6 |
Other non-Hispanic |
92.6 |
80.9 |
74.3 |
40.4 |
14.9 |
Source: Author’s calculations using 2015 March CPS
While the variation among quintiles was not as substantial for single women, Social Security benefits were a significantly larger share of income at the third, fourth, and fifth quintiles than for men and married couples. Despite this, the average Social Security benefit for women is $14,270 a year, compared to $18,375 a year for men.
Race |
Quintile 1 |
Quintile 2 |
Quintile 3 |
Quintile 4 |
Quintile 5 |
---|---|---|---|---|---|
Hispanic |
56.3 |
71.8 |
82.5 |
73.5 |
31.0 |
White alone |
83.2 |
89.9 |
75.2 |
51.1 |
21.7 |
Black alone |
71.1 |
84.7 |
80.2 |
56.5 |
24.6 |
Asian alone |
64.3 |
65.8 |
74.2 |
67.1 |
22.3 |
Other non-Hispanic |
61.0 |
84.9 |
86.8 |
61.1 |
25.6 |
Source: Author’s calculations using 2015 March CPS
Social Security continues to be one of our most successful programs, both in terms of administrative costs and by keeping millions of people out of poverty. As illustrated here, the program is particularly important to people who have earned the least over the years and have fewer alternative sources of income. Strengthening and expanding benefits would go far in providing even more retirement security for those who need it the most.
[1] Specifically, the units used by the Social Security Administration are individuals 65 years of age or older, married couples in which the husband is 65 or older regardless of the wife’s age and married couples in which the wife is 65 or older and the husband is 54 or younger.
[2] Author’s calculations