National Housing Law Project Housing Law Bulletin

Proposed Changes to the Brooke Amendment Will Not Work

HUD calculates area median incomes for 3,177 counties throughout the country.² In no way does 30 percent of the median incomes in those localities accurately separate out those households who can afford to give up the protection of the Brooke Amendment. The 30-percent-of-median cut-off leaves far too many poor people without protection because that level in many places is well below the poverty level. Setting the cut-off at 50 percent of area median would inflict hardship on fewer families, but still not without unfairness.

The median incomes that HUD calculates range from $13,000 in Starr County, Texas,³ to $82,900 in the Stamford-Norwalk, Connecticut, Metropolitan Statistical Area (MSA). That range reflects not only cost-of-living differences among the localities, but also the fact that some parts of the country are much poorer than others. In those poor localities, the median income is far lower than elsewhere, usually less than half the national median income and sometimes even less than the national poverty level. Therefore, setting the Brooke Amendment cut-off at 30 percent of those low median incomes would leave many poor people unprotected, because they would not have enough income to afford expected increases in their rents.

In 95 percent (3,021) of the localities analyzed by HUD, the Brooke Amendment cut-off for a family of four is lower than the national poverty level of $15,600. In 75 percent (2,400) of the localities, the Brooke cut-off would eliminate households with incomes above $12,000; in 48 percent (1,515) of the localities, the Brooke cut-off would be for incomes above $10,000; and in 25 percent (831) of the localities, spread throughout 39 of the 50 states,⁴ the cut-off would be for incomes over $8,840. This last amount represents the annual earnings of a person working full-time at the current federal minimum wage of $4.25 per hour.⁵ In an additional 60 localities that have a higher minimum wage, 30 percent of local median income is less than the full-time minimum-wage earnings. In 14 percent of the localities (448), spread throughout 33 states,⁶ 30 percent of area median income is at or less than $8,000. In 6 percent of the localities (181), spread throughout 24 states,⁷ 30 percent of area median income is at or less than $7,000. In 50 localities located in 12 states,⁸ 30 percent of area median income is at or less than $6,000; in 13 of the country's poorest localities located in five states,⁹ 30 percent of area median income is at or less than $5,000.

The following table might more clearly illustrate the extent of the adverse impact of the proposed 30-percent cut-off.

What does this mean?

Eliminating Brooke Amendment protections to all but households with incomes below 30 percent of area median income would adversely affect people with incomes below the poverty level in all but the wealthiest areas of this country. It would hit particularly hard those families who are underemployed or work at minimum-wage jobs, including many current public and assisted housing tenants and those transitioning from welfare to work. It would disproportionately hit more people living in poorer areas, especially in the Dakotas, Missouri, Oklahoma, Arkansas and many other Southern states.

Adverse impact on working poor and families in transition. In the 891 localities where the Brooke cut-off is less than the earnings of a minimum wage, full-time worker, anyone who moves from welfare to full-time work, even if they earn only the minimum wage, would lose the protection of the Brooke Amendment cap and face unlimited rent increases.

In almost half the localities, working poor households trying to survive with family incomes at or above $10,000 per year would lose their Brooke Amendment protection.

In 75 percent of the localities, working poor households with incomes at or above $12,000 — the average income of working public and assisted housing tenants — would lose their Brooke Amendment protection and face rent increases.

Many current tenants would no longer be eligible. In virtually all of the localities, there would be public and assisted housing tenants with incomes below the national poverty level who will no longer be protected by the Brooke Amendment.

Poorest states would feel widespread adverse impacts. The picture is even more grim in several parts of the country. For example, in the Deep South states of Alabama, Mississippi, and Louisiana where the incidence of poverty is much higher than in the rest of the country, the Brooke Amendment cut-off would be lower than the poverty level in every county. It would be less than the annual earnings of a full-time, minimum-wage earner in almost half the counties of Alabama, almost three-fifths of the counties in Louisiana, and 83 percent of the counties in Mississippi. In Texas, the situation would be nearly as bad. The Brooke cut-off would be below the poverty level in all of the state, less than $12,000 in 90 percent of the state, and below full-time, minimum-wage earnings in over 40 percent of the state.

The situation is similar in West Virginia, Tennessee and Kentucky, each of which covers much of Appalachia. In all of those states, there would be people with incomes above the Brooke cut-off who are still below the national poverty level. The cut-off would be below $10,000 in 90 percent of the counties in West Virginia, almost 80 percent of those in Kentucky and 70 percent of those in Tennessee. Full-time, minimum-wage workers would be unprotected by the Brooke Amendment in 40 percent of Tennessee, and nearly 60 percent of Kentucky and West Virginia.

In the neighboring states of Arkansas, Missouri and Oklahoma, a 30-percent-of-local-median-income cut-off would leave out people who cannot afford rent increases in similarly large portions of the states. Again, in every county in these states, the cut-off would be below the poverty level. Thirty percent of median is below $12,000 in all but one of the counties in Oklahoma, in 99 percent of the counties in Arkansas, and 85 percent of of the counties in Missouri. It would be less than full-time minimum wage earnings in 67 percent of the counties in Arkansas, nearly 60 percent of the counties in Oklahoma and over 40 percent of the counties in Missouri.

In the Dakotas the picture is just as bleak. In every county in those two states, 30 percent of median income is less than the poverty level. In 62 percent of the counties of North Dakota and 68 percent of the counties of South Dakota, the cut-off would be below $10,000. In more than one-third of the counties in both states, people working full-time at the minimum wage would have incomes above the Brooke Amendment cut-off.

In all of these 12 states, people in well more than half and in some cases nearly all of the counties in the state would lose the Brooke Amendment's protection against unaffordable rents, even though their incomes are far below the poverty line. In other parts of the country, the Brooke Amendment will be ineffective in protecting poor people in some areas of each state and in a good portion of many of them, though not quite as drastically as in the 12 states described above.

What Is Wrong with a Cut-off Based on Area Median Income?

Since area median incomes have historically been used in the federal housing programs, one could ask the question: Why doesn't a cut-off at 30 percent of area median income work? Obviously, one reason is that the percentage chosen (30 percent) is too low. In that regard, the percentage used in the Senate bill for the cut-off (50 percent of area median) is somewhat better, but even that percentage will leave unprotected some people who will be hard-pressed to pay increased rents in addition to the other necessities of life. In 19 counties in the country, 50 percent of local median income is less that the earnings of a full-time, minimum-wage worker. In 47 counties, it is below $10,000; and in 215 others it is less than $12,000, the average earnings of working families currently living in public housing. In 1,100 counties, about one third of the total, 50 percent of median is less than the national poverty level.

In fact, basing a cut-off on any percentage of local median income will leave out some people who need protection, because local variations in median incomes do not mean that a family is more or less able to afford higher rents. People with incomes at 30 percent of the $13,000 median in Starr County, Texas, are not equally able to pay market rents as people in Stamford, Connecticut, with incomes at 30 percent of that community's $82,900 median. Nor are they as able to pay their prevailing local rents as people with incomes at the 30 percent level in places like Wichita, Cleveland, Austin, Las Vegas, Grand Rapids, Albany (Georgia), and Charlotte, where the median incomes are about $45,000. The $367 fair market rent for a two-bedroom home in Starr County, Texas, is 113 percent of 30 percent of the median income there. In Stamford, the two-bedroom FMR is 49 percent of 30 percent of the local median income. The FMR for a two-bedroom unit is 44 percent of 30 percent of local median income in Wichita, 47 percent in Grand Rapids, 50 percent in Phoenix, 55 percent in Las Vegas, and 58 percent in Austin. These rent-to-income ratios — especially the 113 percent figure from Starr County — and the variations among them quite clearly demonstrate that 30 percent of local median income is hardly related to local costs of living, local housing costs or ability to secure an affordable rent without protection under the Brooke Amendment. It also illustrates the tremendous inequity that would result from use of such an arbitrary standard of measure to determine who loses that protection.

What it comes down to is that places with low median incomes are primarily places where there are proportionately more people who are poor, but not necessarily places where living costs are lower. Costs may be somewhat lower, but incomes tend to be a lot lower. When incomes are a lot lower and living costs are only a little bit lower, 30 percent of the lower median income does not have the same purchasing power as 30 percent of the median income in the higher income areas.

No Need for Income-Based Cut-offs

In fact, there is no need for an income-based cut-off point for deciding who gets Brooke Amendment protection. Under the House bill, H.R. 2406, the public housing authorities would be allowed to set whatever rent they consider appropriate for an apartment. If Brooke Amendment protection were not cut off at 30 percent of median income, cut-off of protection would automatically occur at the point when 30 percent of the family's income rose above the rent set by the PHA for the apartment. Then the family would pay the rent set for the apartment, which would be less than 30 percent of its income.

Using median-income standard would complicate administration. With a cut-off based on 30 percent of median income, the PHA first would have to determine whether or not the family is protected by the Brooke Amendment; if it is, the PHA would proceed to calculate 30 percent of adjusted income and then decide which of the two is lower, that figure or the standard rent for the apartment. Eliminating an income-based cut-off would also simplify administration of the program. Without it, the PHA would have only to calculate 30 percent of the family's income and charge the family the lower of that figure or the apartment's rent. Every additional step in the calculation creates additional opportunities for error.

Proposed changes would create disincentive for work. Using a 30-percent-of-area-median-income cut-off for Brooke protection also creates an enormous disincentive to going to work, the very thing that the sponsors of H.R. 2406 are trying to eliminate. With the cut-off, families who are not working at all or only part time will get the protection of the Brooke Amendment and their rents will not exceed 30 percent of their adjusted incomes. If they go to work full time and their income exceeds the 30-percent-of-median-income figure, as it will in over 25 percent of the localities in the country, Brooke will no longer cap their rent and the PHA will be free to raise the rent far above what they could afford. With the existing Brooke protection, the rent increase can never exceed 30 percent of the added income the tenant gets from working. Without the existing Brooke protection, the increase could go as high as 100 percent of the added income, or even higher. Given the rent levels prevailing in the local markets and the flat rents the PHAs are talking about, rent increases of this dimension are likely and the work disincentives that they will create will be enormous.

A real-life example brings the point home. A four-person family living in Johnstown, Pennsylvania, would be entitled to a $514 AFDC grant, of which, assuming three dependent children, $118 would go toward the rent (i.e., 30 percent of adjusted income). If the parent went to work full time at the minimum wage, the family's income would be $737 a month, which would exceed the cut-off for Brooke Amendment protection, that is, $727 in Johnstown. The PHA would be free to charge the family a rent it considered appropriate for the apartment. Given the $412 FMR for two-bedroom units in Johnstown, a rent of at least $300 or $350 would be entirely possible. If the PHA set the rent at $300, the family's increased rent would consume almost all of the extra $213 it got by going to work. If the PHA went to $350, the $232 rent increase would be greater than the extra income secured by going to work. In contrast, without the Brooke Amendment cut-off, the family's rent, even after going to work, could not exceed $185 (i.e., 30 percent of adjusted income). Cutting off the Brooke protection at 30 percent of area median allows the PHA to increase the family's rent by $200 or more, instead of $70. A $200 rent increase that eats up virtually all of the additional income is a lot more of a work disincentive than a $70 one.

If Brooke Amendment protections were not cut off at 30 percent of area median, then this family would pay no more than $185 for rent. If the working member got a pay raise or a better job, and the family income increased to $950 a month, the PHA could raise the rent to $250. If the family's income subsequently went up to $1,200 per month, the PHA could charge the tenant the $300 flat rent for the unit, which would amount to 28 percent of its adjusted income. If, at any of these stages in the family's improving financial status, the PHA were to consider that increasing the rent to the maximum allowed under the Brooke Amendment was too much of a disincentive, it would be free to charge the family less than 30 percent of adjusted income. The Brooke Amendment without a 30-percent-of-median-income cut-off is just a maximum percentage of household income, not the minimum that the PHA could charge.

Conclusion

It is clear that reforming the rent provisions for public housing and Section 8 does not require the creation of a median-based cut-off for the protection of the Brooke Amendment cap. Such a cut-off is not only unnecessary; it would cause injustice in many localities, involve extra administrative steps and create an enormous work disincentive. For all of these reasons, the Conference Committee scheduled to reconcile H.R. 2406 and S. 1260 in July would be wise to keep the Brooke Amendment cap that both bills have retained for some tenants, and eliminate the cut-offs based on 30 and 50 percent of area median income.

1. See The Brooke Amendment Should Not Be Repealed, 26 HOUS. L. BULL. 17 (Feb. 1996).
2. This includes 849 in 340 Metropolitan Statistical Areas and 2,328 non-metropolitan counties. The data were released by HUD as an attachment to HUD Notice PDR-95-04 (Dec. 14, 1995).
3. This analysis does not include Puerto Rico, where the median incomes in six of the seven areas are below $13,000.
4. The states are the 33 listed in notes 6-9, infra, plus California, Iowa, Idaho, Indiana, Pennsylvania and Vermont.
5. Six states have their own minimum wage that is higher than the federal minimum. They are Alaska ($4.75), California ($5.15), Iowa ($4.65), Oregon ($4.75), Vermont ($4.75), and Washington ($4.90).
6. The states are the 24 in notes 7-9, infra, plus Alaska, Illinois, Kansas, Maine, Minnesota, Nebraska, Ohio, South Carolina and Washington.
7. The states are the 12 listed in notes 8-9, infra, plus Arizona, Colorado, Florida, Georgia, Michigan, Missouri, Montana, North Carolina, Oklahoma, Oregon, Utah and Virginia.
8. The states are the five in note 9, infra, plus Alabama, Arkansas, North Dakota, New Mexico, Tennessee, West Virginia and Wisconsin.
9. The states are Kentucky, Louisiana, South Dakota, Mississippi and Texas.