It should have been no surprise that the Department of Housing and Urban Development's proposed 2017 budget, which Secretary Julian Castro unveiled last week, included yet another push for a sizeable increase in the Housing Choice Voucher program, the rent subsidy program for low-income tenants.

The proposed $1.2 billion addition would bring the share of HUD's $48 billion budget allocated for the voucher program to $20 billion. Devoting such a significant amount to rent subsidies overlooks a range of problems and complications that should give pause to those who would expand it.

HUD's increasing focus on rent subsidies represents a historic departure from the original concept of the agency — which was predicated on the belief that low-income neighborhoods, where the majority of those in poverty live, could be transformed into better neighborhoods.

Castro's proposal to increase funding to the Housing Choice Voucher Program, also known as Section 8, was couched in the context of recent enthusiasm for using vouchers to "deconcentrate" poverty — to provide the means for poor households to move to more affluent areas where their opportunities might be better. A voucher budget increase, said Castro, "would provide 2.2 million families with the chance to move into neighborhoods with better schools, safer streets and more jobs — and stay there for the long term."

There is no doubt that the voucher program can lead to such an outcome — and that some families will be better off as a result. Indeed, that was the finding of 2015 social science research published by the Harvard economists Raj Chetty, Nathan Hendren and Lawrence Katz, who found that "low-income children are most likely to succeed in counties that have less concentrated poverty."

Notwithstanding Secretary Castro's emphasis on "moving to opportunity," the structure of the program itself creates incentives for voucher holders to cluster, not scatter. For instance, a 2014 study by Eva Rosen of Harvard's Joint Center for Housing Studies found that landlords whose units are in neighborhoods in which it's hard to find non-subsidized  tenants are those most likely to turn to voucher tenants.

The result, the study found, is a "balkanization of the rental housing market that retains voucher holders where they can be most profitable to landlords — in the very neighborhoods policymakers would like to provide them with the opportunity to leave. Landlord tactics serve as a powerful mechanism in the reconcentration of poverty."

Put another way, property owners in better-off neighborhoods lack an incentive to rent to voucher holders — thus creating the possibility of replacing public housing towers with what Sen. Barbara Mikulski once called "horizontal ghettoes."

Moreover, rather than being a path up from poverty, the program Castro would expand may well encourage dependency, rather than upward mobility. HUD data shows that the typical voucher tenant has lived for nearly nine years in her unit. (Women, whether young single mothers or the elderly, predominate in the program. Two-parent families comprise but 4% of voucher tenants.)

The lack of any limit on how long a voucher holder can remain in the program means that waiting lists for the limited number of rent subsidies will inevitably remain long. It's worth noting the Census Bureau reports that of the 45 million Americans in poverty, more than half lived in "poverty areas" in 2010. The current $19 billion voucher appropriation supports a small proportion of such families — only 2.38 million voucher households, composed of 5.3 million individuals.

Secretary Castro would be well advised, if he is to convince a Republican Congress to increase the HUD budget, to be open to rules changes in the voucher program that would address its problems. Current rules, for instance, meant to protect tenants, actually discourage them from increasing their income. That's because a tenant's share of the rent is fixed at 30% of income; if their income goes up, so does a tenant's rent. It would be far better for voucher tenants to enjoy the same fixed-rate rent that other tenants do — and to keep their own money if they start to earn more.

Public housing authorities administer the voucher program and a small handful of them have moved to fixed-rate rent, while also setting a time limit for new voucher tenants. Both of these strategies are permitted through a small HUD regulation-waiver program called Moving to Work. Republican members of Congress would like to expand that program. It's just the sort of compromise Secretary Castro should consider if he's serious about pushing to increase the HUD budget.

Finally, HUD should not abandon its founding mission: to help ensure that poor neighborhoods are good neighborhoods. Notwithstanding the current fashion for "deconcentrating" the poor into higher-income neighborhoods, there are some cold hard facts that must be kept in mind. Convincing landlords in more affluent ZIP codes to accept voucher tenants would mean increasing the rent the government is willing to pay. That would require greater subsidies per housing unit, as rents are typically higher in higher-income neighborhoods. In turn, that would mean that fewer households could be supported with the same appropriation level.

Expanding the program to serve all those of low income currently living in neighborhoods of concentrated poverty would require an approximate five-fold increase in program size. Such a combination of program expansion and "deconcentration" would, moreover, risk backlash in higher-income neighborhoods, which would be asked, for instance, to bear increased education costs.

Low-income neighborhoods can have good schools — as the charter school movement is demonstrating. They can also provide good parks and recreation programs — and, with the right policing methods, can be safe. HUD was founded to help save America's cities. Now is not the time for it to give up that fight.

As founding HUD Secretary Robert Weaver (who was also the first African-American Cabinet secretary) said in 1965: "Our most critical domestic problem is improving the quality of urban life for all Americans. It is our goal to reconstruct the physical and social fabric of the American urban environment." In effect, the emphasis on Moving to Opportunity is an admission of defeat in the battle to improve high-poverty neighborhoods themselves.

Howard Husock is a senior fellow at the Manhattan Institute.