the way prisoners were accounted for in legislative districts. Instead of being counted at their

Transcription

1 No Representation without Population : Maryland s Attempt to Fix Prison- Based Gerrymandering Introduction In 2010, Maryland passed the No Representation Without Population Act, which changed the way prisoners were accounted for in legislative districts. Instead of being counted at their prisons, they will now be counted as living at their last-known home address. The motivation for this was nominally to fix the prison-based gerrymandering present by miscounting a portion of Maryland s population, along with curtailing all of the problems in representation and democracy that accompanied these practices. However, Maryland failed to follow these changes with nonpartisan redistricting and ended up with a partisan gerrymander that technically fixed the prisonbased gerrymandering, but appeared to leave all of the problems and side effects. In order to fix prison-based gerrymandering for good, Maryland should continue its new plan of redistributing prisoners but should then develop a non-partisan redistricting plan to work towards a fair electoral map that allows equal representation and democracy for everyone. Prison-Based Gerrymandering and its Effect on Representation Across the country, political districts are drawn using the results of the US Census, which happens every ten years. The Census Bureau apportions representatives per state based on changes in state population, but the drawing of individual districts is left up to state and local governments. 1 But one major flaw in this method of redistricting is the way prisoners are accounted for. In the US Census, prisoners are counted as living in the prison that they are currently incarcerated in. When states then draw legislative districts based on these census results, they include prisoners in the 1 U.S. Census Bureau. Apportionment Population and Number of Representatives, U.S. Department of Commerce Found at 1

2 districts that their prisons are in. Even though they are forced to live there and are not members of the greater community, they are still counted as living in that district and included in population totals. The method of counting prisoners in their prison districts becomes even more problematic when considered along with the fact that prisoners are not eligible to vote. In Maryland, convicted felons are only eligible to vote once they have served their full incarceration sentence and completed any parole or probation once they are released. 2 Assuming they complete these things, they can regain the right to vote again. But while they are in prison, even though they are disenfranchised, they are still included in the population totals for the district. Essentially, prisons can be used as population buffers. Because their residents cannot vote but they are still counted as residents of the district, the votes of everyone else in the district become more heavily weighted. This method of districting prisoners in their prisons has come to be known as prison-based gerrymandering, as it arguably gerrymanders districts in an unfair way by strategically incorporating prisons (large blocks of citizens who cannot vote) into certain districts. Prison-based gerrymandering creates two major problems that could affect representation: malapportionment and overrepresentation of certain groups. Both of these could lead to non-majoritarian policy outcomes and could impede democracy in Maryland. Malapportionment, or creating electoral districts with unequal ratios of constituents to elected officials, was a major problem in the United States in the 1960s. Districts were commonly not redrawn after population shifts, which meant that fast-growing urban areas had significantly more constituents in them than surrounding rural areas, even though they had the same 2 Maryland State Senate. Voter Registration Protection Act. Senate Bill 488. March 22, Found at 2

3 representation. 3 The issue was brought to the Supreme Court in the 1960s in Baker v. Carr and Reynolds v. Sims. The Supreme Court ruled that malapportionment was unconstitutional, and states were mandated to draw districts for the US House of Representatives and state legislatures that contained equal (or approximately equal) populations. States then began drawing new legislative districts decennially after each Census, in order to account for population shifts. This nominally made districts equal, as they were being updated for population changes and were mandated to hold approximately equal populations in compliance with the one person, one vote principle, which holds that all constituents votes should be approximately equal in a democracy. This method of districting prisoners in their prisons arguably does not violate the one person, one vote principle, as prisoners have lost the right to vote after being convicted, do legally reside in their prisons, and need to still be counted somehow for population purposes in the Census. At first glance, it seems like the counting of prisoners could be similar to the counting of children: children also cannot vote, but are still counted in the Census and included in population totals for legislative districting purposes. Likewise, the counting of prisoners also resembles the way the census counts college students, as both are counted as living at their institutions and families are explicitly told not to include them as family members. 4 But counting prisoners in political districts where their prisons are creates malapportionment problems that are not present when children are counted as living in their family homes. Prisoners do not belong to the communities where their prisons are. When prisoners are released from prisons, they will likely return to the communities they came from, instead of staying around their prison. This is not the case when children are counted as living with their families or when college students 3 Gary Cox and Jonathan Katz. Elbridge Gerry s Salamander: The Electoral Consequences of the Reapportionment Revolution. New York: Cambridge University Press Nancy Scola. Making Prisoners Count. The American Prospect. October 22, Found at 3

4 are counted as living at their colleges. Children grow up as part of the community and feel connected to it, even though they are too young to vote. Likewise, while college students may not get heavily invested in local politics, they tend to engage and interact with the community they reside in while at school. 5 In other words, both children and college students have other ways to have their voice heard and feel like they are a part of the community. Prisoners do not. By forcing them into a political district with a rural community they have no ties to and cannot participate in, they functionally become a phantom population. Elected officials have no reason to represent the interest of the prisoner group, since they have no attachment to the district, cannot vote, will most likely not be around for future elections, and most likely will not ever willingly reside in the district, unlike the rest of the constituency. The non-incarcerated remainder of the constituency is the only section that is listened to and also the only section that can vote; this remainder of the constituency thus holds the same amount of power as the entirety of constituencies without prisons, which is the essence of malapportionment. One of the most extreme examples of this is from a ward in eastern Iowa, where out of the 1,400 constituents, 1,300 are in prison and only 58 are non-prisoners. 6 The three other city wards have relatively equal populations around 1,400, but none of them contains a prison. Effectively, this means that the 58 non-prison residents each have thirty times as much political power as residents in the other three wards. When being interviewed by the New York Times, the city councilor from the previously mentioned eastern Iowa district that was primarily prisoners remarked, Do I 5 Ibid. 6 Sam Roberts. Census Bureau s Counting of Prisoners Benefits Some Rural Voting Districts. The New York Times. October 23, Found at 4

5 consider [the prisoners] my constituents? They don t vote, so, I guess, not really. 7, highlighting the problem quite nicely. This malapportionment could lead to non-majoritarian outcomes, particularly in Maryland when the demographics of the involved populations are considered. Prisons tend to be in rural areas but prisoners are overwhelmingly from urban areas. 8 The 2010 Census found that, of Maryland s nearly 17,000 prisoners, 7,797 of them were from Baltimore City alone (out of a total population of 626,664 for the city). On the other hand, Western Maryland (comprised of Allegany, Garrett, and Washington Counties) had a total population of 252,614 and housed prisons that held 8,030 prisoners, even though only 665 prisoners in the state of Maryland came from those counties. By counting prisoners as residing in their prisons for purposes of Maryland districting, Maryland s urban population is systematically undercounted and its rural population is systematically overcounted. This systematic undercounting of urban centers and overcounting of rural populations is strikingly similar to the malapportionment problems present before one person, one vote, and we could expect to see similar problems with representation here. 9 Biasing the representation in favor of rural constituents could lead to non-majoritarian policy outcomes driven by these rural constituents. For instance, a rural constituency the size of one district could turn into two rural districts padded with disenfranchised urban prisoners. This could result in the rural constituency being able to elect two representatives who would argue for their interests, instead of the one they fairly deserve. In even more extreme cases, like the Iowa constituency mentioned above, there are 7 Ibid. 8 Maryland Department of Planning Adjusted Census Population for Maryland by County and Region Found at 9 Gary Cox and Jonathan Katz. Elbridge Gerry s Salamander: The Electoral Consequences of the Reapportionment Revolution. 5

6 clear non-majoritarian decisions being made when less than 5% of the population are the only ones who can vote and are the only ones who are represented. Prison-based gerrymandering in Maryland also seems like it could lead to unequal representation of certain groups (namely differentiated by class and race), which could lead to nonmajoritarian outcomes. Bartels and Gilens both discuss the prevalence of legislators favoring higher-income constituents over lower-income ones Notably, they did find that the preferences between higher and lower income constituents did not differ considerably on many issues, but that representatives were considerably more likely to listen to their richer constituents when the opinions between classes differed. Even after for controlling for participation, voter turnout, and other methods of political activism, lower class constituents were still less likely to be listened to. Brunner, Ross, and Washington argue that the effects found by Bartels and Gillens are less extreme than they argue, as Brunner et al found that representatives responsiveness to their constituents when classes clashed depended on party: they found that Democrats were more likely to follow the views of lower class constituents, and Republicans were more likely to follow the views of upper class constituents. 12 All three articles discuss the potential negative impact that such findings could have on democracy. If certain class groups are responded to and represented more than others, a social and political hierarchy is created. Policies are more likely to be enacted to meet these groups interests and thus ignore lower-class interests. 10 Larry Bartels. Economic Inequality and Political Representation Martin Gilens. The Preference/Policy Link. Affluence and Influence. Princeton University Press. July 22, Eric Brunner, Stephen Ross, Ebonya Washington. Does Less Income Mean Less Representation?

7 Prison-based gerrymandering could exacerbate both of these inequalities. Maryland s prison population is overwhelmingly black and poor. Statewide, Maryland s population is around 28% black. But on average, 76% of the prison population is black. 13 These numbers match with the demographics of the areas where prisoners are most likely to be from (e.g., the population of Baltimore City is 63.6% black) and where prisons commonly are (e.g., the population of Washington County is only 10% black). 14 Likewise, these prisoners are overwhelmingly likely to be from lower social classes, many of who are living below the poverty line and are homeless and unemployed. 15 By counting prisoners, who are predominantly black and poor, as members of the predominately middle or upper class white communities where prisons are located, the electoral process becomes biased against these lower-class black communities. This can only add to the discrimination and lack of responsiveness they already encounter. The original articles do not discuss race, but it seems likely, given the correlation between race and social class, that black communities in Maryland are also underrepresented. By undercounting lower-class black communities due to prisoners living in prisons in other districts, Maryland has taken a political group that would already be underrepresented and underresponded to and counted them as being smaller than they are. The problems presented by counting prisoners as being residents of their prisons for districting purposes can be summarized as presenting an unfair and inaccurate view of populations. 13 Vincent Schiraldi and Jason Ziedenberg. Race and Incarceration in Maryland. Justice Policy Institute. October 23, Found at 14 U.S. Census Bureau. Maryland. U.S. Department of Commerce Found at 15 Vincent Schiraldi and Jason Ziedenberg. Race and Incarceration in Maryland. Justice Policy Institute. October 23, Found at 7

8 Because prisoners cannot vote, counting them as living in their prisons can create phantom populations that can simply be used to pad districts, causing some the votes of some to be worth more than the votes of others. Given the skewed demographics of prisoners in Maryland, this, in turn, means that some communities and populations can hold more political power than others. Prison-based gerrymandering thus seems to cause one old problem (malapportionment) that has already been considered to be undemocratic enough that the Supreme Court had to rule on it, and exacerbate another problem (unequal representation of populations) that has been shown to be present and problematic to democracy. Maryland s Attempted Solution Maryland s proposed solution to prison-based gerrymandering came in the form of the No Representation Without Population Act of The act required that incarcerated prisoners be counted as living at their last-known home address instead of living at the prison, for the purpose of congressional, legislative, and local districting. The federal prison in Cumberland, Maryland and its 1400 inmates are excluded from the redistricting plans and will continue to be counted as residents of the prison, as federal prisons do not release last-known addresses. 16 Logistically, the act requires a decent amount of work from Maryland. The Maryland Department of Corrections must work with the Department of Planning to verify the last-known addresses of prisoners and add them into the correct districts. Any prisoners who were homeless or who had unknown last addresses would be counted as living at the prisons. Any prisoners who previously lived outside of Maryland immediately before they were incarcerated are not counted in any districts. 16 Maryland Department of Redistricting. No Representation Without Population Act. Maryland Department of Planning Found at 8

9 Following the 2010 Census, 18,309 prisoners were shifted based on previous home addresses. 16,988 of them were moved between districts and 1,321 were removed. The full table, provided by the US Census Bureau, is attached as Appendix 1. As one would expect, urban areas tended to gain population. Baltimore City, Baltimore County, and Prince George s County each gained a net increase of over a thousand prisoners. Areas on the Eastern Shore where there are no prisons also gained net prisoners (as it was impossible for them to lose any), but no county s population gained more than 200 prisoners, presumably due to their low starting populations. On the other hand, rural counties such as Somerset County, Washington County, and Allegany County lost significant numbers of prisoners. In particular, Somerset County lost 10.3% of its population, due to a nearly 3,000-person prison whose inmates were redistributed. Maryland redistricted following the results of the 2010 Census and the redistribution of prisoners to their last-known home addresses. On October 3, 2011, the Governor s Redistricting Advisory Committee released its recommendations for redistricting. 17 The plan was met with criticism from Maryland Republicans, who claimed the Democrat-led committee had intentionally gerrymandered the districts in favor of Democrats. But the bill passed through the state legislature and was signed into law by Governor O Malley on October 19, By March 5, opponents to the new redistricting plan had obtained enough signatures to add the approval of the plan as a ballot initiative in the 2012 elections, believing that Maryland voters may vote against the redistricting plan. But the ballot initiative passed with 61% of the vote, effectively ending debate on the new districts Maryland Department of Redistricting. Congressional and Legislative Districts. Maryland Department of Planning Found at General Election Results. Maryland State Board of Elections Found at 9

10 Recommended Solution: Non-Partisan Redistricting with Redistributed Prisoners In order to fix these institutional problems, Maryland should adopt non-partisan redistricting in addition to having prisoners counted as living at their last-known home addresses. Counting prisoners at their real address should give a more accurate picture of Maryland s population and its distribution throughout the state. When districts are drawn according to these populations, districts should be fairer and more accurately represent the real constituencies. But non-partisan redistricting should be adopted because it will prevent other types of gerrymandering, ensuring that districts remain as compact and cohesive as possible. Without non-partisan gerrymandering, Maryland could potentially ignore the larger problem by redistributing prisoners and then drawing districts in a way that still favor the same communities. This would have technically solved the problem of prisonbased gerrymandering, but would not have solved the larger problems in representation and democracy. Maryland s proposed solution, in fact, comes close to doing exactly this, as is discussed in the next section. Evaluating the effectiveness of redistributing prisoners is hard to do, because Maryland was the first state to adopt such a plan and has only had one election cycle (2012) with the new districting. New York, Delaware, and California have since passed similar laws, but not enough data is available to be able to see any changes (or lack thereof) in voting, policy, representation, or responsiveness. Results of the prison redistribution indicate that the population shift was not particularly large in both Maryland and New York. 19 The redrawn New York districting plans showed a slight increase in population in urban areas and a decrease in rural areas, but the districts themselves did not change shape or size in a meaningful way. Although Maryland s districts did 19 Maggie Clark. Could a Recount of Prisoners Affect Elections?. Stateline: The Pew Charitable Trusts. October 12, Found at 10

11 change, it seems like the change was due more to partisan redistricting and less population shift from prisoners, since the number of prisoners shifted was so low. 20 From looking at the shifted numbers, it seems that redistributing prisoners did succeed in giving more accurate population totals. Both New York and Maryland showed increases in urban centers and decreases in rural areas, as is expected. The magnitude of the change is also somewhat expected, as prisoners are only around 1% of the population, and thus should not significantly impact districts. Jason P. Kelly s article, The Strategic Use of Prisons in Partisan Gerrymandering, provides one of the few quantitative views of prison-based gerrymandering and its effects. He looked at the changes in districts containing and around prisons over time and found that the prison-based gerrymandering was incorporated into most plans: On average, we can expect a party that has recently taken control of the redistricting process to draw more than 5,000 prisoners from districts controlled by the other party or marginal districts into their safest districts. Further, as shown with the Texas and Florida cases, states that aggressively pursue such a strategy will shift tens of thousands of inmates into the districts that they are the least likely to lose over the following 10 years. 21 He found that both Democratic and Republican legislators were likely to draw plans that incorporated these large blocs of non-voters into districts that they would win. This strategy creates safe districts for their party, but uses the minimal number of their voters, allowing them to redistribute others into more competitive districts to give them a greater advantage there. Importantly, he found that the change in prison populations among districts was only statistically significant after the party in control of districting switched. These results indicate that legislators only move prison populations around when it helps their party (presumably, if there is no change in 20 Maryland Department of Planning Adjusted Census Population for Maryland by County and Region Found at 21 Jason P. Kelly. The Strategic Use of Prisons in Partisan Gerrymandering. Legislative Studies Quarterly. Vol 37, Issue 1. February Found at 11

12 power, the same party continues to control redistricting and has already shifted prison populations to their advantage). These results highlight the need for non-partisan redistricting to fully fix the problems surrounding prisons. Even if prisoners are properly redistributed, it seems likely that partisan redistricting could still take advantage of the high number of disenfranchised prisoners in urban areas and create similarly gerrymandered districts that take advantage of this non-voting bloc. The empirical evidence on non-partisan redistricting, however, is somewhat mixed. Michael McDonald argues in favor of them, saying that, while non-partisan redistricting is not perfect, appointing an independent committee to develop redistricting plans is one of the key factors in an ideal redistricting process. 22 He argues that states with partisan committees (or committees whose members are appointed on a partisan basis) are prone to creating gerrymandered maps, and that commissioners should be removed from politics as much as possible. They should be focused on drawing districts that conform to objective criteria (such as compactness and cohesiveness) and they should be accountable to the public, not to partisan legislators. The main problem with non-partisan redistricting seems to be that it is hard to make it truly non-partisan. It is nearly impossible to create districts that are compact, meaningful, cohesive, competitive, and allow for minority views to be heard; many of these factors are frequently at odds. Thus, some factors have to be chosen over others, which inevitably causes districting plans to favor one party over the other. 23 The article details how Iowa s non-partisan committee is often lauded as a success, but argues that this is more likely due to Iowa s successful districting comes from its naturally competitive and mostly even geography. It seems that outcome-based regulations, where 22 Michael P. McDonald. Legislative Redistricting. Democracy in the States. Brookings Institution Press. June 1, Michael P. McDonald. Regulating Redistricting. PS: Political Science and Politics, Vol. 40, No. 4 (Oct., 2007), pp Found at 12

13 the political outcomes (whether it be competitive districts or minority-majority districts) are decided beforehand and the redistricting board must then meet these goals. This may give less room for gerrymandering regardless of the commission s composition, as they have a set of goals that need to be achieved. Redistributing prisoners and non-partisan redistricting may not be a panacea for the problems in representation surrounding the districting of prisons, but both seem to be at least somewhat effective and would certainly be better than Maryland s current solution. Effectiveness of Changes in Maryland Maryland redistributed prisoners and redistricted based on those results in As described previously, the prison redistribution appears to have been successful. Population was added to the major urban centers in Maryland and was subtracted from rural areas. The problem of padding rural populations with non-voting blocs of prisoners seems to have disappeared, as the two populations have been separated before districting occurs. However, the redistricting that followed was partisan and heavily driven by Maryland Democrats. Unsurprisingly, this resulted in a heavily gerrymandered map. A geospatial analysis firm calculated the average compactness of each state s districts, and Maryland ranked last. 24 Much of this can be attributed Maryland s pinwheel-shaped third district, which contains separated parts of Anne Arundel, Baltimore, Howard, and Montgomery Counties. The district weaves together wealthy voters from DC suburbs with poor Baltimore residents, creating a constituency that 24 The Washington Post Editorial Board. Vote Against Maryland Redistricting. The Washington Post. October 19, Found at 13

14 manages to lean Democratic while have few demographics or interests in common. 25 The 2010 redistricting also split Maryland s 6 th district, which had consisted of the rural, western counties that were historically Republican, in half. Half of it remained in the 6 th district and absorbed a heavily Democratic chunk of Montgomery County, while the other half was given to the 8 th district, whose remaining constituents were also heavily Democratic. Unsurprisingly, this resulted in Democratic victories in both the 6 th and 8 th districts, netting the Democrats another member in the House of Representatives. 26 So while the redistricting did technically take prison readjustments into consideration, it failed to fix the representation problems. Most of the districts continue to consist of politically favorable constituencies, instead of meaningful communities or even compact geographic areas. Paired with the earlier discussion on unequal representation, it seems likely that the underrepresented constituencies are still underrepresented. In some districts, like the sprawling 3 rd district, inner-city areas that likely contain large amounts of prisoners are still grouped with rural areas, indicating that perhaps the representation problem has not been solved at all. Non-partisan redistricting has the potential to at least partially fix these problems. Given that Maryland is considered to be the most gerrymandered state at the moment, it seems unlikely that a non-partisan committee could do anything besides help. An outcome-based redistricting plan has potential. If they started with the idea of what districts should look like and what political outcomes should be, perhaps the districts would be more meaningful. If one of the goals were a compact minority-majority district, Baltimore City would probably be mostly contained in such a district. 25 Jeff Guo. Welcome to America s Most Gerrymandered District. The New Republic. November 8, Found at 26 Gerrymandered? Maryland voters to Decide. The Washington Post. Updated September 27, Accessed at 14

15 Much of Maryland s disenfranchised prison population would be included in this district, but their real community would be, as well. They would be able to participate once they finished serving their sentence, so representatives would have a bigger incentive to listen to and respond to them. Likewise, their community would be heavily represented for the size that it truly is, which would hopefully lend to an appropriately sized amount of political power. Without this non-partisan redistricting, Maryland s redistribution of prisoners seems somewhat futile. Perhaps they stopped prison-based gerrymandering in practice, but the same communities seem like they would still be underrepresented and ignored, and gerrymandering clearly still exists. Thus, the fundamental democratic problems found in prison-based gerrymandering continue to exist and seem like they will until a non-partisan redistricting effort is implemented in Maryland. Recommendations/Conclusions Maryland should adopt a non-partisan redistricting plan in conjunction with its policy of redistributing prisoners to their homes. Prison-based gerrymandering is certainly a problem that exists in Maryland, and it does seem likely to have a negative impact on democracy. But simply redistributing prisoners and continuing to redistrict on a partisan basis cannot solve these problems. The same underrepresented communities will continue to be ignored, and the overrepresented communities will continue to hold more political power than they should. Non-partisan redistricting may not produce a perfect electoral map, but it will certainly be an improvement to the current one and help to fix the problems of representation found in Maryland today. 15

16 Works Cited Bartels, Larry. Economic Inequality and Political Representation Brunner, Eric, Stephen Ross, Ebonya Washington. Does Less Income Mean Less Representation? Clark, Maggie. Could a Recount of Prisoners Affect Elections?. Stateline: The Pew Charitable Trusts. October 12, Found at Cox, Gary and Jonathan Katz. Elbridge Gerry s Salamander: The Electoral Consequences of the Reapportionment Revolution. New York: Cambridge University Press Gilens, Martin. The Preference/Policy Link. Affluence and Influence. Princeton University Press. July 22, Guo, Jeff. Welcome to America s Most Gerrymandered District. The New Republic. November 8, Found at Kelly, Jason P. The Strategic Use of Prisons in Partisan Gerrymandering. Legislative Studies Quarterly. Vol 37, Issue 1. February Found at Maryland Department of Planning Adjusted Census Population for Maryland by County and Region Found at Maryland Department of Redistricting. Congressional and Legislative Districts. Maryland Department of Planning Found at Maryland Department of Redistricting. No Representation Without Population Act. Maryland Department of Planning Found at Maryland State Board of Elections General Election Results Found at Maryland State Senate. Voter Registration Protection Act. Senate Bill 488. March 22, Found at McDonald, Michael P. Legislative Redistricting. Democracy in the States. Brookings Institution Press. June 1,

17 McDonald, Michael P. Regulating Redistricting. PS: Political Science and Politics, Vol. 40, No. 4 (Oct., 2007), pp Found at Roberts, Sam. Census Bureau s Counting of Prisoners Benefits Some Rural Voting Districts. The New York Times. October 23, Found at Schiraldi, Vincent and Jason Ziedenberg. Race and Incarceration in Maryland. Justice Policy Institute. October 23, Found at Scola, Nancy. Making Prisoners Count. The American Prospect. October 22, Found at U.S. Census Bureau. Apportionment Population and Number of Representatives, U.S. Department of Commerce Found at U.S. Census Bureau. Maryland. U.S. Department of Commerce Found at The Washington Post. Gerrymandered? Maryland voters to Decide. Updated September 27, Accessed at The Washington Post Editorial Board. Vote Against Maryland Redistricting. The Washington Post. October 19, Found at 17

18 Appendix 1: Census Bureau Chart 18