Dr. Larry J. Sabato University of Virginia Center for Politics 465 Crestwood Drive Post Office Box 400806 Charlottesville, Virginia 22904

FEATURE CLE: SABATO'S CRYSTAL BALL

CLE Credit: 1.0 Wednesday, June 18, 2014 1:15 p.m. - 2:15 p.m. Exhibit Hall I Northern Kentucky Convention Center Covington, Kentucky

A NOTE CONCERNING THE PROGRAM MATERIALS

The materials included in this Kentucky Bar Association Continuing Legal Education handbook are intended to provide current and accurate information about the subject matter covered. No representation or warranty is made concerning the application of the legal or other principles discussed by the instructors to any specific fact situation, nor is any prediction made concerning how any particular judge or jury will interpret or apply such principles. The proper interpretation or application of the principles discussed is a matter for the considered judgment of the individual legal practitioner. The faculty and staff of this Kentucky Bar Association CLE program disclaim liability therefore. Attorneys using these materials, or information otherwise conveyed during the program, in dealing with a specific legal matter have a duty to research original and current sources of authority.

Printed by: Evolution Creative Solutions 7107 Shona Drive Cincinnati, Ohio 45237

Kentucky Bar Association TABLE OF CONTENTS

The Presenter ...... i

The End of Amendments? Why State Legislative Polarization Makes Constitutional Amendments Increasingly Unlikely ...... 1

It's Time to Increase the Size of the House ...... 5

The 2020 Reapportionment and the Voting Rights Act ...... 11

14 for '14: Some Bold Predictions for the New Year ...... 15

THE PRESENTER

Dr. Larry J. Sabato University of Virginia Center for Politics 465 Crestwood Drive Post Office Box 400806 Charlottesville, Virginia 22904

DR. LARRY J. SABATO is the founder and director of the University of Virginia Center for Politics. He is also the University Professor of Politics at the University of Virginia, and has had visiting appointments at Oxford University and Cambridge University. A Rhodes Scholar, he received his doctorate from Oxford, and he is the author or editor of two dozen books on American politics.

In 2013 Professor Sabato won an Emmy award for the television documentary Out of Order, which he produced to highlight the dysfunctional U.S. Senate.

Professor Sabato directs the Center for Politics' Crystal Ball website, a leader in accurately predicting elections since its inception. In 2004, the Crystal Ball achieved a 99 percent accuracy rate in predicting all races for House, Senate, Governor and each state's Electoral College outcome. In 2006, the Pew Research Center and the Pew Charitable Trusts' Project for Excellence in Journalism recognized the Crystal Ball as the leader in the field of political predictors, noting that the site "came closer than any other of the top ten potential predictors this cycle."

In 2008, the Crystal Ball came within one electoral vote of the exact tally in the Electoral College, while also correctly picking the result of every single gubernatorial and Senate race across the country. In 2010, the Crystal Ball was the first to forecast a solid Republican takeover of the House. While others were predicting a Romney victory in 2012, the Crystal Ball forecast a substantial Obama margin in the Electoral College, and ultimately missed just two states. The Crystal Ball had a combined 97 percent accuracy rate in forecasting the Electoral College, Senate, House and gubernatorial contests.

Earlier this year, the Crystal Ball won a "Beast Best" award from The Daily Beast as one of the top political sites on the web.

In October 2013, Prof. Sabato and the Center for Politics unveiled the Kennedy Half Century project. The project consisted of a New York Times bestselling book, The Kennedy Half-Century, a PBS documentary, a Massive Open Online Course (MOOC) available on Coursera and iTunes U, an app with the complete recordings and transcripts from Dealey Plaza on 11/22/63, and a website (www.thekennedyhalfcentury.com).

ii THE END OF AMENDMENTS? WHY STATE LEGISLATIVE POLARIZATION MAKES CONSTITUTIONAL AMENDMENTS INCREASINGLY UNLIKELY Thomas F. Schaller, Guest Columnist Reprinted with permission from Sabato's Crystal Ball, March 27, 2014

This year marks the centennial anniversary of the first class of popularly-elected U.S. Senators, as mandated by adoption of the 17th Amendment. A hundred years later, several current or former Republican members of Congress, including Todd Akin (MO), Paul Broun (GA), Pete Hoekstra (MI) and Jeff Flake (AZ), have indicated their support for returning the selection of U.S. Senators to state elites.

Although the movement to repeal the 17th Amendment is likely to fizzle, the fact is plans to amend the Constitution are mostly a waste of time because, other than a widely popular and highly-unifying suggested change, it is probably almost impossible to ratify or even propose amendments in our highly-polarized nation and divided national government. Holding aside the ten amendments of the Bill of Rights – and an 11th regulating congressional compensation that, proposed more than two centuries ago as one of twelve originally proposed amendments, was belatedly ratified in 1992 as the 27th Amendment – only sixteen amendments that were not part of the constitutional bargain struck in 1787 have been proposed and ratified in the 226 years since the founders met in Philadelphia.

That works out to one amendment about every fourteen years. After setting aside the unusual case of the 27th Amendment, that means almost forty-three years have passed – three times the average interval – since the last normal, proposed-then-ratified-soon- thereafter amendment was adopted: The 26th Amendment's extension of voting rights to those aged eighteen-to-twenty years old. The record between amendments is sixty-one years period between the 12th Amendment, which in 1804 corrected flaws in the presidential election process, and the landmark 13th Amendment, which abolished almost all forms of slavery in 1865. That sixty-one-year standard may soon fall.

Indeed, is it politically possible to adopt a constitutional amendment in today's polarized America? Barring an extremely unlikely second constitutional convention, two-thirds of members of both chambers of Congress are required to propose any amendment, which seems virtually unattainable. But even if congressional supermajorities could agree, the bigger hurdle might be ratification by state legislatures, which have become almost as polarized as the national government.

The number of split-party legislatures, for example, has steadily declined over the past three decades. Because most state legislators and governors are elected in presidential midterm cycles, it's instructive to look at state results following midterms. As shown in Table 1 below, the number of split-party state legislatures hovered around eleven or twelve from 1998 to 2006 – presidential and odd-year elections move the numbers a bit in between – but dipped to eight both prior to 2008 (not depicted) and as a result of the 2010 elections.

1 Table 1: State legislative chamber control, following recent midterm elections

Sources: Washington Post (1998), Polidata.org (2002, 2006), NCSL (2010)

Following the 2012 cycle (also not depicted), the number of split state legislatures fell further to five (now six after the 2013 statewide elections in Virginia gave Democrats the tiebreaker to control the state's senate), which is about half the number from a decade- and-a-half ago. The 1998 midterms may have been the tipping point at which control over state legislatures began to stabilize. At the time, National Conference of State Legislatures analyst William Pound declared 1998 a "status quo election" in which fewer state legislatures had changed hands in any election going back to 1988.

The fact that just six of forty-nine partisan legislatures – as usual, holding Nebraska's unicameral and non-partisan legislature aside – are split is relevant to the amendment ratification process for two, related reasons.

First, and most obviously, the rise of unified state legislatures reflects the rising blue/red polarization evident throughout American electoral politics. In the past five presidential elections, forty states have voted all five times for either the Republican nominee or the Democratic nominee, and the number of divided U.S. Senate delegations today is almost half of what it was four decades ago. And as Table 2 shows below, state election results increasingly mirror national ones.

Table 2: Unified state party control, following recent midterm elections

Sources: Washington Post (1998), Polidata.org (2002, 2006), NCSL (2010)

Table 2 reports the number of unified state governments, including control of the governorship: That is, those in which Republicans hold the governorship and both legislative chambers, or Democrats have unified control. Again using results following midterm cycles, notice that the number of unified state governments reached thirty-one following the 2010 midterms. Although this figure dropped slightly as a result of the 2012 presidential and 2013 off-year elections, the number of unified governments right now is higher than at any time going back to the start of the Eisenhower Administration. "A lot of it is the top of the ticket influence," NCSL's long-time elections expert Tim Storey remarked on the declining number of split-state governments. "Some states clearly cast votes for one party or the other."

2 Second and more relevant to the calculus for amendment ratification is the fact that the cleaving of states into reliably blue and red subsets makes it highly unlikely that anything other than an amendment idea supported by overwhelming national majorities could attract the necessary three-fourths of the states required by the Constitution's Article V. Even if there were a dominant party that boasted regular control of both legislative chambers in a solid majority of states, it would still be difficult. Although Republicans have made recent and steady gains – thanks in large part to their now near-hegemonic control over southern state legislatures – neither party has been able to maintain reliable control over more than about fifteen legislatures each.

The 2008 and 2010 election results are instructive: Following the former, a strong Democratic cycle, the Democrats controlled both chambers in twenty-seven states, with Republicans dropping to fourteen; following the latter, a strong Republican cycle, the Republicans controlled both chambers in twenty-five states, with Democrats falling to sixteen. If polarization and the declining number of split-state legislatures allows each party to consistently lock down somewhere between eighteen to twenty-two state legislatures each, the parties have a long way to go to amass the thirty-eight states necessary to meet Article V's three-quarters requirement to ratify an amendment.

For the sake of argument, consider the Democrats' recent high-water mark of twenty- seven states as a starting point for either party and throw in Nebraska to make matters interesting. That's still only twenty-eight states, and at best the window of control over this high a number might last two or perhaps four years. So in very short order a party would still need to persuade a sufficient number of opposition party legislators to support ratification in ten states where the opposition controls at least one if not both chambers. Suffice it to say that in a polarized, red/blue America, building that sort of supermajority support would be extraordinarily difficult.

Of course, barring a second constitutional convention, any constitutional change would have to first be proposed by two-thirds votes in both chambers of Congress – a standard which by itself almost certainly prevents an amendment from being sent to the states for their ratification consideration in the first place. The bottom line? America hasn't adopted an amendment in the traditional, proposed-then-ratified-soon-thereafter fashion in more than forty years, and it may be a long time before it happens again.

Thomas F. Schaller is a professor of political science at the University of Maryland, Baltimore County. He is the author of Whistling Past Dixie: How Democrats Can Win Without the South and national political columnist for the Baltimore Sun.

4 IT'S TIME TO INCREASE THE SIZE OF THE HOUSE Sean Trende, Senior Columnist Reprinted with permission from Sabato's Crystal Ball March 6, 2014

"Our founders put the first amendment first for a reason. It protects all Americans' right to free speech, regardless of political affiliation or views." This statement was made by former Senator Norm Coleman (R-MN) in 2007, but it expresses a commonly cited view among lawyers, judges, politicians and pundits.

In fact, what we today know as the first amendment wasn't originally intended to be the first amendment. Examine closely this copy of the original Bill of Rights, as submitted to the states. Today's first amendment was originally the third article. The original second article, which prevented Congress from giving itself a raise without an intervening election, was ratified in 1992 as the 27th Amendment.

The actual first article – what Congress thought should be the first amendment – dealt with congressional apportionment. It read:

After the first enumeration required by the first article of the Constitution, there shall be one Representative for every thirty thousand, until the number shall amount to one hundred, after which the proportion shall be so regulated by Congress, that there shall be not less than one hundred Representatives, nor less than one Representative for every forty thousand persons, until the number of Representatives shall amount to two hundred; after which the proportion shall be so regulated by Congress, that there shall not be less than two hundred Representatives, nor more than one Representative for every fifty thousand persons.

In simple English, this amendment, properly interpreted (most agree that there's a scrivener's error in the final line), would have fixed the maximum size of a congressional district at 50,000 people. It should technically be part of the Constitution: It was ratified by the requisite number of states in June 1792, but for whatever reason, Connecticut's vote to ratify the article was not recorded and was only later rediscovered. Indeed, there has been (unsuccessful) litigation to force the Archivist of the United States to do so.

With that said, it is probably just as well that it has never been recorded. We would be forced to choose between two bad options. First, we could allow the size of Congress to grow to over 6,100 members. When Congress was in session it would qualify as a top- twenty city in eight states; California would have more than 700 representatives. The other option would have been to amend expressly our original Bill of Rights, a precedent we've thus far avoided.

Nevertheless, this article really was placed first for a reason. The debates over the ratification of the Constitution, as collected in the Federalist Papers and less-well- organized Anti-Federalist Papers, contain a surprising amount of discussion over the size of congressional districts (the Federalists argued that large districts were beneficial to avoid elections from turning into personality contests), and there was an implied promise contained in the ratification of the Constitution to pass an amendment regulating district size.

5 Even though Article the First was thought to have gone unratified, the early Congresses seemingly tried to follow the animating principle behind the Amendment. It kept the size of districts roughly stable until the 1820 reapportionment, and district size didn't increase beyond 100,000 constituents until the 1860s. When Congress last increased the size of Congress in 1912 (setting aside temporary expansions for the admission of states), each representative had just over 200,000 constituents.

Today a representative answers to over 700,000 constituents, well over ten times the number of constituents deemed appropriate by the First Congress. While it seems unwise to adhere to the strict letter of Article the First, the time has likely come to abide by its spirit and increase the size the House.

The United States is supposed to be the world's premier representative democracy, yet India's Lok Sabha is the only lower chamber on the globe where representatives have more constituents. Indeed, Pakistan's National Assembly and Indonesia's People's Representative Council are the only other lower chambers with a population-per-seat ratio exceeding even 400,000.

Even setting aside abstract ideals, increasing the size of the House could bring numerous benefits:

First, a bigger House would diminish the impact of malapportionment that comes with single-member states. A certain amount of malapportionment is inherent in House redistricting, but as the number of single-member states grows, this effect becomes more and more pronounced. The largest single-member state, Montana, requires a representative to take on over 1 million constituents. On the other hand, the smallest state, Wyoming, only has 582,000 constituents.

This will become more pressing in 2020, when Rhode Island is expected to lose one of its two House seats. Unless Montana gains one back, there will be more single-member states than at any time in our country's history. Roughly one in six states will be single member; the only time that ratio was higher was immediately following the 1820 apportionment, and then only because three single-member states had been admitted to the Union shortly before the reapportionment.

As a side note, and as a result of the next apportionment after the 2020 census, New Jersey will likely have its smallest congressional delegation since the first decade of the 1900s, New York since the first decade of the 1800s, Ohio since the 1820s and Pennsylvania since the 1790s. As a result of the last apportionment, Illinois has its smallest delegation since the 1860s, Indiana since the 1830s, Iowa and Missouri since the 1850s and Massachusetts since the original apportionment. This is all despite massive increases in population size.

Second, increasing the size of the House could diminish the impact of gerrymandering. While the commonly voiced concerns about the evils of gerrymandering are overstated, they aren't entirely without basis. Moreover, the perception of impropriety is important, and there's something inherently untoward surrounding legislators drawing districts that resemble Rorschach inkblots rather than compact units that draw together communities of interest.

6 Larger legislatures make it more difficult to gerrymander effectively. Think of it this way: If there are 100 residents in a state with 100 congressional districts, there is no gerrymandering possible. If there are fifty congressional districts, it isn't impossible, but it is still difficult. If, however, there are only five districts, a legislator will probably be able to draw the lines to cluster a disproportionate number of Republicans in a single district, leaving Democrats overrepresented in the remaining districts.

In fact, we see this to some extent in state legislatures. Let's assume, somewhat controversially, that the partisan composition of state legislatures ought to roughly correlate with presidential performance in the state. We can look at all ninety-eight state chambers that have partisan representation, and look at the difference between the Republican vote share and Republican presidential performance in the state to get a sense of how gerrymandered the state legislatures are (we'll take the absolute value, since we don't care if Republicans are over- or under-represented). We can then perform a regression analysis, using the size of the legislative chamber as our independent variable.

The coefficient points the right direction (as chamber size increases the disparity between presidential performance and the partisan orientation of the chamber decreases). The result isn't statistically significant at 95 percent confidence, but from a policy standpoint, we're still awfully confident that the relationship is "real." (p=.229). If we add a further control for whether the lines were drawn by a commission or partisan legislators, the relationship becomes stronger (p=.17). I suspect that with a proper control for partisan concentration in a given state, the relationship would reach full significance.

Third, increasing the size of the House would make majority-minority districts easier to maintain. As I noted in my last piece for the Crystal Ball, as states shed districts it becomes increasingly difficult to draw compact groupings with populations that are made up predominately of members of minority groups.

A number of Northern states are likely to find themselves forced to draw fewer majority- minority districts in the upcoming round of redistricting. At the same time, many Southern states have congressional delegations that underrepresent the minority share of the population. Mississippi is 37 percent African-American, but only one of its four-seat delegation is majority black. Louisiana is 32 percent African-American, but only one of six is majority black. Alabama is 27 percent African-American, but only one of seven is majority black. Florida is 16 percent African-American, but if Democrats are successful in pressing a redistricting lawsuit against the state, it may be the case that only one of its twenty-seven members is African American.

With more districts, however, relatively small, discrete populations can suddenly support their own preferred members. For example, if you doubled the size of the House, drawing three African-American majority districts in Alabama is a piece of cake, and drawing a fourth district – rendering 29 percent of the state's seats majority black – is not difficult. Drawing nine African-American majority districts in Florida is a bridge too far, but four of fifty-four – easily done – is closer to true proportionality than one of twenty- seven.

Fourth, increasing the size of the House would improve rural representation. It's striking to go through early Almanacs of American Politics and observe how many districts were

7 based in small towns or even entirely rural areas. But as America has become more urban and as district size has increased, that has gradually eroded.

In 1970, five of Virginia's ten districts contained little-to-no territory from the major population centers of Richmond, Hampton Roads and Northern Virginia. Today, that's unambiguously true of only two districts. Now-extinct districts like Louisiana's Eighth (no major population center) and Alabama's old Third (Dothan, Phenix City) have been necessarily eliminated. Even the few remaining rural/small town districts are endangered; each decade, Nebraska's Third District marches inexorably eastward toward the outskirts of Lincoln and Omaha.

Rural and small-town districts will inevitably comprise smaller and smaller fractions of Congress as long as our nation grows and urbanizes. At the same time, they shouldn't be allowed to go extinct, as rural and small-town America have unique needs and problems. Increasing the size of the House would help address this.

Fifth, though this is difficult to measure and hence more speculative, increasing the size of the House could help mend some of the detachment that is felt between Washington and the states. Campaigns would be less expensive, so politicians would have to spend less time fundraising, and representatives would have fewer constituents to answer to, allowing for more personalized representation. It may even be that smaller constituencies allow for the election of more ideologically diverse members. Indeed, there is some correlation between the size of state legislative chambers and the number of third-party/independent candidates elected.

What would the downsides be? Perhaps it would be more difficult to get anything accomplished, although that rings false given the glacial pace with which the current Congress operates. It would be more expensive, although congressional staffs could be smaller, offsetting some of the cost. It would still be only a small fraction of the federal budget.

At bottom, there's nothing magic about the number 435. While we probably wouldn't want 6,000 representatives, there are still twenty-seven legislative bodies that are larger than the House of Representatives. The British House of Commons operates effectively with 646 members (each representing about 98,000 constituents). Adopting that size would translate to one representative per 485,000 constituents, bringing us back roughly to where Congress was in the 1960s.

Even adopting the so-called "Wyoming Rule," which would set the size of a Congressional district at the population of the smallest state, would result in a House of only 547 members, tying for the thirteenth largest in the world. This would still decrease the number of constituents per district by 20 percent, and would help minimize the distortions from malapportionment. A variant of this would guarantee that every state receives at least two representatives; this would increase the House to around 1,100 members, about 40 percent larger than the British House of Lords (and still half the size of the Chinese National People's Congress).

Of course, this is unlikely to happen. For one thing, none of the interest groups in Washington would want it, as it would make their jobs harder and diminish their clout. More importantly, increasing the size of the House of Representatives diminishes the power of each individual member. It isn't in their interests to pass further increases in

8 House size. But regardless of whether such a proposal is likely to pass, it seems an unabashedly good idea that citizens and legislators should nevertheless press for.

Sean Trende is the senior elections analyst for RealClearPolitics and a senior columnist for the Crystal Ball. He is the author of The Lost Majority: Why the Future of Government Is Up for Grabs and Who Will Take It, and co-author of the Almanac of American Politics 2014. Follow Sean on Twitter @SeanTrende.

10 THE 2020 REAPPORTIONMENT AND THE VOTING RIGHTS ACT Sean Trende, Senior Columnist January 30th, 2014 Reprinted with permission from Sabato's Crystal Ball, http://www.centerforpolitics.org/crystalball/articles/the-2020-reapportionment-and-the-voting- rights-act/

We're pleased to announce that Sean Trende, the shrewd senior elections analyst for RealClearPolitics, will be contributing regularly to the Crystal Ball. His first piece looks at the future of majority-minority House districts in northern states, and how in some instances it will be difficult to maintain these districts as some of these states lose House seats after the next census. Follow Sean on Twitter at @SeanTrende.

In late December, psephologists received a belated Christmas present in the form of the 2013 population estimates from the United States Census Bureau. While relatively uninteresting in and of themselves, the figures do allow us to further estimate what the next congressional apportionment will look like. In the first few years after the census, analyses along these lines are mostly parlor games that should be clearly marked "for entertainment purposes only."

But with each passing year population numbers get locked in, and it becomes increasingly difficult to significantly alter the trajectory of population growth. By this point in the 2000s, our estimates for 2010 were a decent approximation of what actually occurred.

The Crystal Ball's Kyle Kondik recently looked at the states that would be expected to gain and lose seats as a result of the census, and concluded that although seats are moving from bluer states to redder states, Republicans will likely be unable to use redistricting to press their edge in Congress much further.

I want to look at this from a slightly different angle, and lay down a marker for 2020: While the major fight will, as always, be over the partisan makeup of the districts being drawn, the fight over the future of majority-minority districts will rival it. But while we have plenty of experience with partisan fights over redistricting and know how to deal with them, the coming fight over the death of some majority-minority districts is basically without precedent.

We first need an understanding of the basics of the Voting Rights Act (VRA) as it relates to redistricting. Before we undertake that inquiry, we need an important caveat about the scope of this piece: The Voting Rights Act as it relates to redistricting is an extremely complicated subject, and one that is constantly in flux. So what follows is necessarily a rough outline; use it as the basis for an answer on your law school exam at your own risk.

The Supreme Court of the United States has interpreted Section 2 of the VRA to require states to draw districts that will tend to elect a candidate of a minority group's choice if (a) the minority group is compact and will comprise a majority of the inhabitants of the

11 district; (b) the group is politically cohesive; and (c) white voters will vote in blocs to defeat the minority group's preferred candidate.

We will look at these factors in turn. Most people only focus on the first prong, and only half of that prong to boot: that the minority group must form a majority of the district's inhabitants. But the district must be geographically compact as well, else the Gingles factors (named for the preconditions under which a district can be challenged under Section 2 of the VRA) won't be triggered. In fact, drawing non-compact majority-minority districts may trigger court scrutiny and risk having the district declared an unconstitutional racial gerrymander (although that, too, is a very complicated issue).

This compactness requirement has significant implications for drawing majority-minority districts in shrinking states. A state that loses in reapportionment necessarily sees the population of its remaining districts increase significantly. So if a state has majority- minority districts that are on the verge of losing their majority-minority status, losing an additional district could be fatal.

Let's look at the states likely to be affected by redistricting, starting with New York. The Empire State presently has a number of majority-minority districts, but I want to examine the African-American majority districts. There are three: The Queens-based Fifth, which is presently 51 percent African-American, and the Brooklyn-based Eighth and Ninth, which are, respectively, 56 percent and 53 percent African-American.

Unless there is a surge in immigration from the Caribbean Basin or Africa itself, it is difficult to see how all three of these districts can be maintained. Each of those districts will have to add somewhere in the neighborhood of 30,000 residents. But only one of the adjacent districts has a population that is more than 10 percent African-American. In other words, it will be difficult to avoid adding substantial nonblack populations to these districts. This could drop them all below the 50 percent African-American threshold.

This presents two challenges. First, the Queens-based Fifth actually can be maintained as a majority-minority district. It abuts the other district that contains an African-American population greater than 10 percent: the Nassau County-based Fourth. About the only option for map-drawers would be to drop minority-plurality precincts in central Queens, and to extend the district over toward Hempstead in Nassau County. This was once former Sen. Al D'Amato's (R) political base, but today is both heavily Democratic and minority.

But removing some 40,000 voters from the Fourth District, 90 percent of whom voted for Barack Obama, would leave the Fourth leaning Republican, with no obvious source of additional Democratic votes nearby to offset the loss. (The seat is currently held by retiring Democratic Rep. Carolyn McCarthy and gave 56 percent of its votes to President Obama in 2012.) The fight that would erupt among Long Island Democrats in this event would be fierce.

The other two black-majority districts will pose a challenge of a different sort. Both could probably be maintained as plurality-minority districts that would tend to elect the candidate of African-Americans' choice (although continued upper-middle class white immigration to northern Brooklyn could alter this). Remember, however, that the Voting Rights Act requires majority-minority districts. Would redistricters be required to keep one of the two districts with an African-American majority, at the expense of tilting the

12 other district toward not electing a candidate of African-Americans' choice? It seems like a strange outcome, but it's plausible under current interpretations of the act.

A similar conundrum presents itself in Ohio. To keep the Eleventh District an African- American majority seat, redistricters extended it out of Cleveland and down into Akron. It is now 54 percent African-American. Despite its increasingly odd shape, it is less heavily minority than it was in 1980 (58 percent black) or 1990/2000 (55 percent black).

Once again, the problem is that there just simply are not enough African-American adults to keep it majority-minority without further distorting the district. Under the 2010 census numbers, Ohio districts would have to have 770,000 residents if the state had fifteen districts overall (Ohio has sixteen seats now but could lose another district after 2020). According to census data, there are only 436,000 black residents of Cuyahoga (Cleveland) and Summit (Akron) counties. So a line drawer would have to incorporate virtually all of the African-American population of these counties to keep the district majority-minority, without adding too many whites. This is nearly impossible, given that only a handful of precincts bordering the Eleventh have even African-American pluralities.

If Republicans control redistricting, they could try to extend the district eastward toward Youngstown, trying to keep it majority-minority, which Democrats might challenge, credibly, as an unconstitutional racial gerrymander. If Democrats control redistricting, they would likely resist these efforts, hoping to disperse the African-American population and damage GOP chances in neighboring districts. This could create an odd coalition of Republicans and blacks versus Democrats on redistricting, similar to what we saw emerge this cycle in Florida, Missouri and in central Ohio.

Additional problems could arise for Democrats in Illinois. There are three African- American districts in Chicago, including the First District, which has elected an African- American representative since the 1920s. None has an African-American population exceeding 55.5 percent, and the First District – which was extended far into the suburbs in redistricting – is down to a 51 percent African-American population. None of the neighboring districts has a substantial African-American population, so again, at least one of these districts will be in jeopardy.

Michigan is our last state, and it illustrates the perils of dropping the African-American share of the electorate by too much. It has two majority-minority districts, and both have a large enough minority population that they can probably be maintained as majority- minority districts, although doing so requires drawing bizarrely-shaped districts that could make Rep. Sander Levin's (D, MI-9) white working class district into a swing district.

But in 2012, the African-American share in both districts was necessarily lowered when Michigan lost a district, and in a low-turnout primary, a white Democrat from the suburbs won a seat that had been held by African-Americans since the 1950s (Rep. Gary Peters, now running for the U.S. Senate). So the threshold for a functional VRA seat may be above 50 percent in Michigan.

Further complicating this is the fact that many Democrats would prefer to weaken majority-minority districts. Part of the Democrats' challenge in winning the House is that the VRA forces them to place their most loyal supporters into districts with one another. If Democrats could weaken these districts, they could dilute Republican strength in the

13 suburbs and create more Democratic districts. We see this with the ongoing Florida litigation, where Democrats are urging the dismantling of two African-American plurality districts in the hopes of weakening neighboring districts currently held by Republicans.

Remember also that there are two other requirements for a majority-minority district: racially polarized white voting and minority voting cohesion. The former has been breaking down, especially in northern states. We've seen minority candidates elected to Congress in white-majority districts in places like Indiana, Minnesota, Ohio and Wisconsin. An African-American Republican has a good chance at winning an 87 percent white district in Utah. Of course, we also have an African-American president who didn't win the white vote, but against whom whites did not bloc vote in most northern states.

These examples are insufficient evidence to demonstrate that white candidates no longer bloc vote to defeat candidates of minority voters' choice as a general matter. But what if this trend continues throughout the decade? At some point in the future, racial bloc voting could weaken enough that the VRA no longer requires majority-minority districts in some Northern states.

At the same time, we may find ourselves presented with increasing questions about the cohesiveness of the minority vote. It's often overlooked, but Republican congressional candidates won at least 35 percent of the Hispanic vote in 1994, 1998, 2000, 2002, 2004 and 2010. The Republican presidential candidate performed well in 2000 and 2004. We assume that years like 1996, 2006, 2008 and 2012 represent the rule rather than the exception, and there are good arguments for this view. But what if Republicans perform well with Hispanics in 2014 (the president's job approval is weak with this group right now), and a candidate with a broader Hispanic appeal than Mitt Romney wins the GOP nomination in 2016? At some point, the assumption of Hispanic cohesiveness may need to be questioned.

We know the template for fights over partisan redistricting: We've been having them for almost as long as we've had parties. But the template for shrinking VRA districts is heretofore unknown, and presents fascinating potential coalitions between minority groups and Republicans. While I wouldn't predict any sort of realignment on the basis of these coalitions, I would suggest that, because of these upcoming fights, the post-2020 redistricting will be unlike anything we've seen.

14 14 FOR '14: SOME BOLD PREDICTIONS FOR THE NEW YEAR Larry J. Sabato and Kyle Kondik Reprinted with permission from Sabato's Crystal Ball, December 19th, 2013

As we wrap up our publishing schedule for 2013, we looked into the Crystal Ball to offer some predictions for next year. While the picture is still a little hazy for the 2014 midterms, we've got some other prognostications related to the political year to come and some hints about what you should expect.

1. At least one additional U.S. senator will announce his or her retirement in 2014. We're giving ourselves some wide latitude here, because it's possible the retirement will come after the 2014 midterms. But we suspect at some point next year another senator will decide to quit; the question is, will it be an incumbent who determines he or she cannot win in 2014 – similar to ex-Senator Ben Nelson (D-NE), who retired in late 2011 after realizing his perilous political position – or will it be someone who retires for another reason – like ex-Senator Olympia Snowe (R-ME), who left the game in early 2012 because of frustration despite having an easy road to another term.

While it's not technically a retirement, another important development shook the Senate Wednesday evening: Senator Max Baucus (D-MT) is reportedly going to be nominated as ambassador to China. Baucus had previously announced his retirement, but this would open up his seat well in advance of the midterm election. After Baucus would resign, Governor Steve Bullock (D) would then reportedly appoint Lt. Governor John Walsh (D) to fill out the rest of Baucus' term. Walsh was already running for the seat, and being an incumbent would make him a stronger candidate both against his main primary opponent, ex-Lt. Governor John Bohlinger (D), and Representative Steve Daines (R), the likely Republican nominee and current general election favorite. Obviously, there are a lot of moving parts here, so as of this writing – Wednesday evening – we're not going to change our rating in this race yet. But provided this scenario unfolds as follows – Baucus resigns, and Bullock appoints Walsh to the seat – we'll be very tempted to move this race from Leans Republican to Toss-up.

2. There will be at least one surprising result in a Republican Senate primary. Several Republican senators are facing primary challenges next year, and there are also Tea Party insurgents running in other Senate contests who are hoping to win the Republican nomination over more establishment-oriented challengers. Frankly, we think there is a little more smoke than fire in many of these races, but it's also clear that the battle for the soul of the GOP continues, and that there are several outside groups (like the Club for Growth and the Senate Conservatives Fund) that will fund outsider candidates. So expect at least one upset in a Republican Senate primary this year in which either an incumbent senator goes down (like Richard Mourdock's primary win over then-Senator Richard Lugar in Indiana last cycle) or an establishment favorite is knocked off before he or she can advance to the general election (such as Sue Lowden's loss to Sharron Angle in Nevada's 2010 Senate primary).

15 3. Fewer than a half-dozen incumbent governors will lose reelection (including re- nomination defeats). We've noted throughout the year that despite many competitive gubernatorial races, there might not be the kind of turnover in the statehouses that we saw in 2010 because there are so many incumbent governors running: Of thirty-six races coming next November, incumbents are running for reelection in twenty-nine of them. In 2010, just thirteen incumbents ran in the general election, which contributed to a great deal of turnover even though only two incumbents actually lost reelection bids. In the thirteen midterms conducted over the past half-century (the years when most gubernatorial races are held), an average of 4.5 incumbent governors lost in each general election. Given that no incumbent governor seems to be in serious danger of losing a primary next year at this point, it appears likely that only a relative handful of incumbents will lose.

4. The result in the special U.S. House election in FL-13 will get far too much attention. If ever there was an apparent bellwether special election, the one coming up in FL-13 this March would seem to be it. The district went 50 percent- 49 percent for President Obama in 2012, quite similar to his 51 percent-47 percent national edge over Mitt Romney. Not only that, but the district is located entirely in Pinellas County, which can fairly be described as one of the key presidential swing counties in the entire country (Obama won this Tampa-area county with 52 percent of the vote in 2012). The seat, held for decades by the late Representative Bill Young (R), is probably a necessary part of any future Democratic House majority. It is particularly valuable to the party of Jackson's long shot hopes of winning the lower chamber, but the result won't necessarily be predictive of anything. The local issues emerging in the race will be unique to it, and the national environment guiding the results will not necessarily be replicated in the full midterm, which will be held about eight months later (several eternities in politics). But that won't stop a lot of breathless projection from a single House result. We'll try to keep a level head about it.

5. There will be more House retirements. The race for the House was rocked by three important retirements Tuesday. First, Representative Frank Wolf (R, VA- 10), who was initially elected in 1980, called it quits, giving Democrats an opportunity to win a Northern Virginia seat that's very balanced politically. Then, Representative Jim Matheson (D, UT-4) became the first Democrat this cycle to retire (although some others are leaving the House to run for other offices). That was quickly followed by the exit of Representative Tom Latham (R, IA-3), who, like Wolf, holds a swingy House district (Latham's seat backed Obama in 2012, while Wolf's backed Romney).

These retirements prompted some dramatic ratings changes: VA-10 and IA-3 go from Likely Republican to Toss-up, and UT-4 – the most Republican-leaning House district held by any Democrat – goes from Leans Democratic to Likely Republican.

16 Chart 1: Crystal Ball House ratings changes

Expect more retirements. Over the past forty years, there has been an average of twenty-two House retirements each cycle (this is according to Roll Call's helpful congressional casualty list). So far, there have only been nine House retirements this cycle; if no more members retire – unlikely – it would tie for the smallest total over the past four decades (the 1983-1984 cycle). However, given the historical average, it's a safe bet that more are coming.

For more on what we're calling "Getaway Day" in the House, check out this report on the Crystal Ball website.

6. Scandal will cost several House members their seats, through defeat at the polls or resignation. It's hard to beat House incumbents, who are reelected in general elections at a greater than 90 percent clip. Those who do end up losing are typically either on the wrong side of a party wave or have crippling personal or ethical problems. Expect some representatives to end up in the latter camp this year. For instance, Representatives Scott DesJarlais (R, TN-4) and Trey Radel (R, FL-19) are both suffering from terrible publicity, and Representative Charlie Rangel's (D, NY-13) various ethical problems surely contributed to his near- defeat in a primary last cycle. There are others, on both sides of the aisle, with ethical questions that could derail them next year. The lower chamber of Congress is called the "People's House" because it's supposed to be representative of the average American, capturing both the towering heights and depth-defying lows of the human experience. Its members, at the very least, often do a very good job of epitomizing the latter.

7. An anti-Hillary will begin to emerge in the Democratic Party for 2016. Yes, Hillary Clinton is a big favorite to win her party's nomination for president in 2016. But there's too much lingering resentment among Democrats for her and her husband, and too much baggage – like Clinton's 2002 Iraq War vote, which remains a liability more than a decade later – for her not to at least receive at least a minimal challenge from someone for the nomination. One possibility to watch: ex-Governor Brian Schweitzer of Montana.

8. One or more new disputes or issues will emerge that the news media will label game-changers for the midterm, but actually won't change much of anything. There will almost certainly be something that happens nationally, or in a given race, that will affect the outcome of the midterm. There also will be several developments that will seem big at the time but will probably end up being pretty meaningless. Consider President Obama's "you didn't build that" or Mitt Romney's "corporations are people" remarks from the last cycle: Yes, they made

17 headlines and provided fodder for the opposition. But it's a sizable stretch to say that either remark moved many votes.

9. As President Obama slips fully into lame-duck status, an ex-Obama staffer will publish a highly critical book or major magazine piece on his flaws. And it might well reflect worse on Valerie Jarrett, the president and first lady's close adviser, than it will on the first family (Jarrett has been absolutely hammered in recent post-election books and articles).

10. Trial balloons for wannabe presidential candidates will endanger commercial air traffic over Iowa. So fly safe.

11. The national GOP will restructure its presidential nominating process, following the successful precedents set by national Democrats in the 1970s and 1980s. Oh, wait…

12. Third-party candidates will make few inroads in the 2014 midterms despite widespread dissatisfaction with Congress. Hate it or love it, the United States is locked into a two-party system, with very few exceptions. There are numerous reasons for why this is this case, but we'll defer to The Simpsons for a concise explanation.

13. People who follow Washington reporters and analysts through social media will learn far more than they need or want to know about DC's sports teams and weather. Be ready for the next (non)-apocalyptic snowstorm, coming to a place (maybe not) near you.

14. Finally, we believe that elections for at least thirty-five Senate seats, 435 House seats, and thirty-six governorships will occur in the new year. In collaboration with the Farmers' Almanac, we have picked the likeliest date for these elections: November 4, 2014. Call us crazy, but we feel really good about this one.