TAKING VIRTUAL REPRESENTATION SERIOUSLY.

• Author: Fishkin, Joseph
• Position: Special Issue on Gerrymandering

TABLE OF CONTENTS INTRODUCTION 1682 I. BAD VIRTUAL REPRESENTATION--AND ALTERNATIVES 1690 TO IT A. Which Interests Matter? 1697 1. All Politics Is Partisan 1697 2. All Politics Is Racial 1699 3. All Politics Is Local 1701 B. The American Approach 1702 II. GOOD VIRTUAL REPRESENTATION? 1709 A. Piggybacking on Geography 1710 B. Prison Gerrymandering and Baker v. Carr 1712 C. The Limits of Proportional Representation 1719 III. VIRTUAL REPRESENTATION IN THE CONSTITUTION 1722 CONCLUSION 1727 INTRODUCTION

"If I am not for myself, who will be for me? But if I am only for myself, what am I?" (1) Nobody likes virtual representation. Even the suggestion of it carries a taint of illegitimacy. There are good reasons for this. The history of democratic political development, both in this country and elsewhere, has been a history of the incremental, halting, painfully slow, sometimes reversed, always contested replacement of virtual representation, in which people do not get to vote for their representatives, with actual representation, in which they do. Over time, democracies have determined that various groups of people such as women, racial minorities, and the poor, are capable of choosing their representatives at the ballot box.

In the United States in particular, this is not just any history. It is the spine of our dominant democratic narrative. Our major moments of enfranchisement, many of them memorialized in Article V Amendments to the Constitution, (2) link together into a constitutional story, and the story has a moral: we are capable of choosing for ourselves, rather than relying on the wisdom and beneficence of others--for definitions of "we" that include not just well-heeled white men but also women, minorities, and the poor. To be sure, our actual constitutional history is considerably less Whiggish than this narrative. Just ask the women of Revolutionary-era New Jersey, who won the right to vote in 1776 and then lost it in 1807 for more than a hundred years, or the African American men of the former Confederacy, who won the right to vote during Reconstruction and then lost it for the first two-thirds of the twentieth century. (3) The tragedy of these reversals does not diminish the moral story, but rather tends to augment it. The right to vote--that is, the right to actual, direct representation--is a moral beacon by whose light we now retrospectively view much of our democratic history.

In this great story, virtual representation is cast as a villain. It played an especially explicit and conspicuous role as a villain in the fight over women's suffrage. "[T]he virtual representation argument," Reva Siegel explains, was "the core of the antisuffrage case." (4) For more than half a century, suffrage opponents pressed various related arguments to the effect that women were better off with virtual representation, with their husbands and fathers voting in ways that would take their interests into account. (5) There was a right side and a wrong side in that debate. When the right side eventually won, (6) our polity crossed a line that later democratic theorists would view as conceptually significant: for the first time ever, most of the representation in our system of representative government was actual rather than virtual. (7)

This long trajectory, and the legitimate sense of enlightened democratic accomplishment that accompanies it, leaves virtual representation today in a very awkward place, like an unwanted guest from less democratic times that has greatly overstayed its welcome. The reasons why we have virtual representation in the first place--its conceptual foundations and justifications, the normative universe in which it made sense--have largely been lost.

This is what led the plaintiffs in Evenwel v. Abbott, the 2016 Supreme Court case, (8) to see an opening for a very bold and modern claim, one that had the potential to completely remake our system of representation in a way not seen since Baker v. Carr. (9) Essentially their claim, about which I will say more below, was that virtual representation is over, and from now on we ought to draw our maps in a way that takes into account only actual representation. (10) The present and long-standing practice in the United States is to draw district lines so that each district contains the same number of people. (11) The Evenwel plaintiffs argued that instead, we should ignore the nonvoting people and draw districts with equal numbers of voters. (12) The plaintiffs' key move was to urge the Court to view districting exclusively in terms of the voting power of voters seeking actual representation, and therefore to disregard--as irrelevant to the project of districting--the representation of those who cannot vote.

The plaintiffs in Evenwel surely knew their odds were long. A victory would have upended every districting map in every jurisdiction in the country. Courts are rightly loath to do that. Furthermore there was a serious problem of constitutional text looming behind the case. (13) But the plaintiffs also knew they had a shot. The reason they had a shot was that they were aiming at the soft underbelly of our present system: its extensive yet undertheorized reliance on virtual representation. (14) At oral argument, Justice Stephen Breyer brought this up. He was clearly uncomfortable with endorsing this fundamental aspect of our current system of representation--the fact that some of those represented cannot vote, and therefore rely on other people's votes to choose the people who will represent them. Breyer said:

That sounds an awful lot [like] what they had in 1750 or something, where the British Parliament said, well, don't worry, America, you're represented by the people in England because after all, they represent everybody in the British Empire.... I mean, that people are being represented through somebody else is a little--possible, but tough. (15) In the end the Evenwel plaintiffs lost decisively. The Court held unanimously that total population is a perfectly fine basis for drawing district lines. (16) Justice Ruth Bader Ginsburg, writing for the Court, went one step further, elevating the use of total population (the current practice) to a sort of default baseline, and suggesting that departures from it require some justification. (17) However, Justice Ginsburg never tried to provide a defense of virtual representation, even though her opinion means that our Constitution--now more explicitly than before--favors an approach to districting that relies on it.

This Article offers a qualified defense of virtual representation. I do not claim that virtual representation is superior to actual representation. It is not. My main claims instead are: (1) that virtual representation is an inevitable feature of any democratic system, including ours; (2) that it has real value, even though it is second-best to actual representation; and (3) that rather than run away from virtual representation, we ought to take it seriously--and try to do it better.

Let us begin with the first and simplest point. Virtual representation is inevitable. It is an inescapable component not only of the current American system but of every democratic system, past, present, or future. In any society that has some children too young to vote--and here it does not matter whether the voting age is set at eighteen or at any other number greater than zero that one might choose--those children are necessarily represented only virtually in the halls of government. In any society with immigrants--which is to say, in nearly any conceivable society in a world with human migration--as long as immigrants are not immediately granted the right to vote on the day they arrive, there will be virtual representation of those immigrants for some period of time. (18) In the American system, there are also several groups of adult citizens who are represented only virtually. In every state, it is possible to lose one's right to vote by reason of mental incapacity. (19) In every state but two, people who are in prison lose their right to vote, while of course in many states, people who were convicted of crimes many years ago cannot vote. (20) All of these people are virtually represented.

It is possible to deny this. It is possible to claim that really none of these groups is virtually represented. On this view, the job of an elected representative is exclusively to represent the people who vote, or perhaps, those eligible to vote. If we take this view seriously, it would follow that children--the largest group of people who cannot vote--are not actually the constituents of any representative. If we accept this, it follows that when a seventeen-year-old contacts the representative of his local area for constituent services, such as help navigating a government bureaucracy, or perhaps a nomination to a military service academy, the representative should, in theory, politely decline, to the same extent that she would decline a request from a person whose home address fell outside her district. (21) It would also follow on this view that, when making policy decisions, a representative ought not view the interests of children in her district as having any more independent weight in her decision calculus than the interests of people on the other side of the country. (22) Only voters, on this view, count as the constituents whose interests an elected official has a special obligation to serve.

This is an implausible view. One indication of its implausibility is that nobody ever seems to argue for it. Even the plaintiffs in Evenwel repeatedly claimed that should they win their case, children would still be constituents of whoever was the representative for the place they lived--that is, the children would be virtually represented--even though they would not count for district line-drawing purposes. (23) Similarly, a century earlier, opponents of women's suffrage did not claim that women were not...