Phantom Voters, Thanks to the Census — New York Times Editorial

by New York Times, December 27, 2005 Link

New York Times Editorial Board, December 27, 2005

The first Constitution took for granted that enslaved people could not vote, but counted each slave as three-fifths of a person for the purpose of apportioning representation in Congress. This inflated the voting power of slaveholders and gave them much more influence in legislative matters than their actual numbers warranted. No American would knowingly tolerate such an arrangement today. But a glitch in the census that inflates the populations of some state legislative districts - thus exaggerating their voting power - has led to a contemporary version of that problem. It involves counting prison inmates in the district where they are confined rather than where they actually live. The Census Bureau could fix this problem in a heartbeat, so it needs to get a move on.

The culprit is a provision in the census that counts prison inmates as “residents” of the institutions where they are held, often for relatively short periods of time. Denied the right to vote in all but 2 of the 50 states, the inmates are nonetheless treated as voters when the State Legislatures draw up legislative districts. This practice mattered little 30 years ago, when the prison population was tiny. But with about 1.4 million people in prison today, it can be used to shift political power from one part of the state to another.

A startling analysis by Peter Wagner of the Prison Policy Initiative found seven upstate New York Senate districts meeting the population requirements only because inmates were included in the count. The Republican Party in New York relies on its large upstate delegation for its majority in the State Senate - and for its political power statewide. New York is not alone. The Prison Policy Initiative’s researchers found 21 counties nationally where at least 21 percent of so-called residents lived behind bars.

By counting these nonvoting inmates as residents, the prison counties offend the principle of one person one vote, while siphoning off political power from the home districts to which the inmates will return as soon as they are released. Since inmates are jobless, their presence also allows prison districts to lower their per capita incomes, unfairly increasing their share of federal funds earmarked for the poor. Congress, which has just caught on to this, recently gave the Census Bureau 90 days to file a report on the feasibility of counting inmates at their homes of record rather than in prison. At the same time, a committee overseen by the National Academy of Sciences has been studying the residency issue and is expected to make its final report this spring. But why does the bureau need another study to decide whether it wants to uphold the one-person-one-vote principle? The bureau should get to work immediately on procedures that would allow it to count inmates where they actually live - and get those procedures locked in place by the 2010 census.

Financial burden of how prisoners counted in Census falls on rural — not urban — communities

by Peter Wagner, December 8, 2005 Link

Census 2000 found one out of every 200 residents of New York City in an upstate prison and counted them as if that was their actual residence. I have written extensively on this site about how this relatively small population transfer is magnified through the redistricting process to radically change the balance of power in New York in violation of the state and federal constitutions.

The impact of Census counts of incarcerated people on funding streams for local governments is far smaller than the political impact, but it is worth exploring how funding is affected and who gains and who loses funding from the practice. In contrast to the political effects, the transfer of 0.5% of New York City’s population upstate is, when dropped in the giant ocean that is the budget processes of the federal, state and local governments, a tiny ripple that is difficult to see long before it reaches the shore.

These budget processes are not commonly understood, and the resulting confusion impairs an honest debate about reforming the Census Bureau’s method of counting incarcerated people.

This article will make two critical points:

1. The way the Census counts prisoners does not significantly reduce the funding available to the urban communities where most prisoners come from.
2. The financial Census benefit to prison towns comes not from the places that prisoners come from but at the expense of other rural communities without prisons.

As discussed in the article Eric Lotke and I wrote in the Pace Law Review, [PDF] while Census data does play a role in the distribution of more than $1.5 trillion each decade in federal aid, the vast majority of this aid is completely unaffected by where prisoners are counted within a state. The two largest programs, Medicaid and Highways, amount for 74% of the total and are completely unaffected. The majority of the remainder are highly tailored programs that target the program to the need. Simply put, most government programs are smart enough to base a program for poor school age children on a formula that includes not the total population in a county but the number of poor school age children.

There are funding streams that are less sophisticated, but the losers aren’t urban citizens. For example, the U.S. Department of Agriculture distributes $60 million annually to impoverished Appalachian communities via a formula that includes the total population. The completely unanticipated result is to reward communities that build prisons by giving them a larger share of the $60 million Appalachian aid pie.

Stop and think who loses here: The victims are not the urban communities that supply the prisoners because they are not eligible to apply for this money. The real victims are the poor Appalachian communities that see the prison communities more quickly draw down the very limited $60 million fund.

As we described in the Pace Law Review article, most of the money redirected by prison census counts is raised in specialty taxes (liquor taxes, cigarette taxes, recreational park usage fees, hunting-fishing licenses, etc.) and county sales taxes. Not all states have these revenue sources, and in the big picture this is small change, but it is important to see who pays for the windfall received by some.

Dutchess County, NY, can provide a detailed example. In 2003, the town of Fishkill and the small City of Beacon argued over whether the prison counted in one town was really in the other because $85,000 in county sales tax revenues was at stake. Although the prisoners were from New York City, neither the prisoners nor New York City had a valid claim on these funds.

This was not a state sales tax being distributed within the state on the basis of population, but a county sales tax being distributed on the basis of population within the county. The county sets the tax rate — about 3% of each purchase — and keeps that money locally. As a result of their “population” based formula, towns with elevated populations due to prisoners get an extra share. So if that money doesn’t belong to New York City or to the prison towns, to whom does it belong?

That money belongs to every other town in the county that does not have a prison. The towns with prisons get a windfall, and every community without a prison is deprived of about 1.7% of the tax receipts it would otherwise receive. (See below table).

It is in fact true that if the Census started to count prisoners as residents of their legal and pre-incarceration addresses, there would be an impact on funding for rural communities. There would be little change in the money that urban communities receive, and towns with prisons would see some decline in their revenue. But most critically, a fairer count of the population would result in an increase in the funds received by the many rural communities that do not have prisons.

As I’ll be describing in more detail in the coming weeks, most urban and most rural people have something to gain from changing how prisoners are counted in the Census. But the biggest difference is in political clout, not funding.

Note: The two cities in Dutchess County receive a fixed share of county sales tax revenue and are therefore unaffected by a decision to change how prisoners are treated for tax distribution purposes. All figures in this table are based not on Census 2000’s official results, but rather the results of the Count Question Resolution program. The Census counted many prisons in this county in the wrong spot, and those corrected numbers are used in the county for planning purposes. However, for purposes of redistricting, the prisoners were simply excluded rather than corrected in any way. For more on tax distribution in Dutchess County, see Anthony Farmer, “Plans slice local sales tax share”, Poughkeepsie Journal, Feb 22, 2005, p. 1A and Ginny’s Fair Tax Page. Tax receipt figures for 2004 were the basis for all calculations.