Losing to Win How Partisan Candidates Help Parties Win in the Future Kai Steverson April 8, 2012 Abstract In the U.S. presidential election of 1964 the republicans nominated a candidate, Barry Goldwater, who was seen as too ideological extreme to be viable. And as predicted, Goldwater lost in a landslide. Yet conservatives in America argue that the election in 1964 played a crucial role in securing future victories by energizing and building the political base. This paper formalizes this logic in an in nite horizon model of two party competition. In the unique subgame perfect equilibrium parties concede elections by picking a partisan candidate in order to help them win in future periods. The key mechanism is motivated by two pieces of empirical evidence: 1) that partisan candidates increase turnout from the base, and 2) that voting is a persistent behavior. Thus a partisan candidate running today can raise future turnout from the base. The model also generates three testable empirical predictions. 1) Parties newly out of power nominate more partisan candidates. 2) The longer a party is in power the more likely they are to lose. 3) Parties converge in terms of policies but not in terms of candidates nominated. 1 Introduction In the U.S. presidential election of 1964 the Republican party nominated a candidate, named Barry Goldwater, who was considered too ideologically extreme to have any chance of winning. In a editorial written the day after the nomination the New York Times called Goldwater "a disaster for the Republican party" who threatened to turn the GOP into a "right-wing splinter group" and who had "minimal" chance of winning ("The Goldwater Nomination", July 16, 1964, p. 30). During the campaign, the Democrats successfully played o Goldwater s extremism with the slogan "in your guts, you know he s nuts". As predicted, Goldwater lost in a landslide. He got less than 40% of the vote and only won six states. 1
Given this large loss, and its predictable nature, how can Goldwater s nomination be explained? Conservatives now argue that Goldwater s nomination energized the conservative movement, which played a crucial role in future victories. William Middendorf, who worked on the Goldwater campaign and later served as Secretary of the Navy, argued that the campaign of 1964 "left behind a cadre of millions of true believers" without which "there would have been no Reagan or Bush administrations" (Middendorf (2006), p. xii). Conservative columnist George Will wrote that "[Goldwater] lost 44 states but won the future...he catalyzed conservatism s breakthrough" ("A Man Who Won the Future", May 27, 1994, p. C7). Segal (1968) reports evidence that in 1964 the republican base turned out a higher rate than the democratic base, supporting the idea that Goldwater energized the base. Goldwater himself viewed the primary purpose of his candidacy as energizing the base. Goldwater didn t really believe he could win, yet decided it was important to rally the base, to give voice to the conservative cause. In his autobiography (Goldwater (1988)) he re ects on a conversation he had in December of 1963 with his wife: I told Peggy something I ve never to this moment told anyone else in my life. I said I didn t want to run. In my gut, there was never a burning desire to be President. I just wanted the conservatives to have a real voice in the country. Many of us were damn tired of the Democrats. They d eventually wreck the economy and bring the whole country down with them. Someone had to rally the conservatives, take over the Republican Party and turn the direction of the GOP around. There was no one to do it but me. We d lose the election but win the party. (p. 154) The case of Goldwater exempli es the thesis of this work, which can be stated as follows. Partisan candidates energize the base, but are less electorally viable. Energizing the base is valuable for future elections. This creates a trade-o between winning today and winning tomorrow. The goal of this paper is to formalize this argument. To do so I develop an in nite horizon model of political competition in which parties nominate candidates to compete in an election. In the unique subgame perfect equilibrium a moderate candidate wins in every period and yet partisan candidates are chosen in order to invest in the future. In particular parties choose partisan candidates when they face electoral adversity, which in the model will be created through an incumbency advantage. Out of power parties, that did not recently energize their base, face a less than 50% chance of winning even with a moderate candidate. Given this situation they decide it is better to concede the current election and build a stronger future with a partisan candidate. 2
The key mechanism where partisan candidates help the party win in the future has two ingredients: that partisan candidate energize the base and that this lasts at least until the next election. Energizing the base has potentially many dimensions, for the purposes of this work I speci cally mean increasing turnout. That partisan candidates increase turnout among their base nds empirical support in the literature on "abstention due to alienation". The idea that increasing turnout today impacts turnout tomorrow is founded in the literature on "habitual voting". Section 2 discusses these two ingredients and uses them to motivate the key mechanism of the model. The formal literature on candidate selection has largely centered around reconciling the following two facts: that moderate candidate perform better in elections and yet partisan candidates are still chosen. This literature has only employed models that take place in period, and in a one period parties are restricted to care only about two things: winning today and policy today. Thus a one period cannot explain nominating a sure loser, such as Goldwater. Because my model is multi period parties can care about something new: winning and policy in the future. This allows for a new reason to nominate partisans, that they help parties win future elections. The model developed also delivers three testable predictions. 1) The longer a party is in power the more likely they are to lose. 2) Parties converge in terms of policies but not in terms of candidates nominated. 3) Parties newly out of power nominate more partisan candidates. All three predictions arise from the tension that parties face between winning today and winning tomorrow and di erentiate my model from previous models of candidate selection. These predictions as well as evidence supporting them will be discussed in section 6. Original empirical work is beyond the scope of this paper, but evidence taken from the literature will be presented. The rest of the paper is organized as follows. Section 2 motivates and discusses the mechanism whereby partisan candidates help parties win in the future. Section 3 gives a brief review of previous formal models of candidate selection, and compares those models with this work. Section 4 gives the formal description of the model. Section 5 characterizes the equilibrium and gives a sketch of the proof. Section 6 discusses the three testable predictions of the model. Sections 7 and 8 relax the main assumptions of the model by looking at purely o ce motivated and impatient parties respectively. Section 9 concludes. 2 Motivating the Model: Evidence and Theory In this section I discuss and motivate the two key ingredients of the model: that partisan candidates increase the turnout of the base and that this e ect lasts until the next election. 3
The rst ingredient links to the literature on "abstention due to alienation" and the second to "habitual voting". These are discussed in 2.1 and 2.2 respectively. Crucially these literatures connect in that people who did not vote previously are more likely to have their turnout decision a ected by the ideology of the candidate. In other words "habitual voters" are less likely to "abstain due to alienation". This connection leads to the key mechanism of the model: that partisan candidates can in uence whether voters from their base acquire the habit of voting. This mechanism is fully discussed in 2.3. 2.1 Abstention due to Alienation Zipp (1985) examined how the ideological position of the candidates relative to the voter in uences turnout. Zipp scored the ideological of position of each individual and each candidate on a range of policy issues. He found that individuals who were far from either candidate voted less, an e ect referred to as "abstention due to alienation". Of course the ip side of this e ect is that people who are ideologically close to at least one candidate vote more. This captures the idea that a partisan candidate is capable of energizing the base. More recently Plane and Gershtenson (2004) and Adams et al. (2006) have conducted similar studies and also found a signi cant "abstention due to alienation" e ect. Adams et. al. assumes the e ect works through a threshold, that is an individual votes only if the closest candidate is close enough. Using maximum likelihood they estimated the determinants of this threshold. They nd that having voted in the past election signi cantly reduces how close the candidate has to be. In other words non-voters are more likely to require a candidate with similar ideology to decide to vote. 2.2 Habitual Voting Plutzer (2002) notes: "[v]irtually all major works on turnout have concluded that voting behavior is, in part, a gradually acquired habit". The observation that underlies this conclusion is that voting is a highly stable behavior. This can be observed from both cross sectional and panel data. Miller and Shanks (1996) look at cross sectional cohort data and nds cohort turnout rises for the rst two or three election before hitting a long "turnout plateau" and then tailing o at old age. Similarly Plutzer (2002) looks at panel data from U.S. presidential elections and notes that of the 516 respondents who voted in both 1968 and 1972 97% voted again in 1976. More generally numerous studies show previous voting is a robust predicator of current voting in multivariate analysis (e.g. Franklin (2004); Brody (1977)).The interpretation given to these results is that the population can be divided into two groups: voters and non-voters. During young adulthood individual citizen 4
sorts themselves into these groups and, once settled, are high unlikely to shift from one to the other. For the purposes of this work what matters is that there is a causal like between voting today and voting tomorrow. That such a link exists has been con rmed by both the instrumental variable and experimental approach. Green and Shachar (2000) instrumented for past voting using aggregate variables from the past elections known to in uence voting, such as perceived closeness of the election or the ideological gap between candidates. They concluded that past voting causes current voting. Gerber et al. (2003) conducted an experiment where the treatment groups was urged to vote through direct mail or face to face canvassing. The treatment was more likely to vote in the current election and this di erence persisted in subsequent elections. In fact the those who voted in the treatment group became indistinguishable from those who voted in the control group. These two studies point to something intrinsic to voting that makes future voting more likely. 2.3 The Mechanism The common explanation given in the literature for habitual voting is that voting involves a one time xed cost. Plutzer (2002) puts it as follows: As young citizens confront their rst election, all of the costs of voting are magni- ed: they have never gone through the process of registration, may not know the location of their polling place, and may not have yet developed an understanding of party di erences and key issues. Moreover, their peer group consists almost entirely of other nonvoters: their friends cannot assure them that voting has been easy, enjoyable, or satisfying. Of course there is variation, some people enjoy engaging in the political process, even in their rst election. The crucial point is, all else equal, having experience with voting makes it easier, and hence more likely. This explanation of habitual voting connects in a crucial way with the evidence mentioned above that non-voters have a much higher "abstention due to alienation" threshold. In other words those who still face the xed cost to voting are more likely to require a candidate with a similar ideology to be convinced to vote. This evidence isn t perfect in that nonvoters are de ned as someone who didn t vote last period, not someone who has never voted. Nevertheless this highly suggests that a partisan candidate brings rst time voters from the political base to the polls, overcoming the initial cost to voting, and pushing them onto the path of habitual voting. More habitual voters in the base leads to higher turnout in the 5
future, enhancing the party s future election prospects. This is the key mechanism of this paper by which partisan candidates help parties win in the future. 3 Previous Models of Candidate Selection The formal literature on candidate selection began with the insights of the median voter theorem, which goes back at least far as Downs (1957). In that model whichever party is closer to the position of the median voter wins with certainty, which causes both parties converge to the median voter in equilibrium. This creates two testable implications: that moderate candidates perform better in election and that both parties should converge to the same position. The empirical evidence strongly supports the rst of these claims and strongly disagrees with the second (see Ansolabehere et al. (2001) and Erikson and Wright (2005)). Much of the subsequent literature has focused on modifying the standard model to explain divergence. To do so it must address the following question: if moderates perform better in elections why are partisans candidates ever nominated? Here I brie y survey some of approaches the literature has taken to this question. The most direct way to explain divergence is assuming that parties have an intrinsic preference for partisan candidates, i.e. a policy motive. This tack was rst taken by Wittman (1983) and later developed by Calvert (1985) and Roemer (1994). This can only work if the basic Downsian model is modi ed to include uncertainty. This makes the candidate closer to the median only more likely to win instead of certain to win. Then policy motivated parties are willing to accept a lower probability of winning in order to get a chance of electing a partisan. This produces divergence in a setting where moderate candidates do better, matching the empirical evidence. Palfrey (1984) took another approach and argued that even if only two parties are observed in an election, the possibility of entry by a third party creates divergence. To see this suppose that both parties have converged to the median position. Then a third party could enter slightly to the right (or left) of the median and capture close to half of the votes, while the original two parties would be forced to split the remaining half. The third party gets a plurality of the votes, making its entry pro table, and so convergence to the median cannot be an equilibrium. This explains candidate divergence even when parties care only about winning. Another approach to generate divergence has been taken by Adams and Merrill III (2003). Their model is similar to mine in that they assume partisan candidates energize the base, increasing turnout. They also motive this with the evidence on "abstention due to alienation". The key di erence is in their model partisan candidates do better in elections than 6
moderate candidates. This happen because the votes a candidate gains from moving to the center is outweighed by the number of citizen in their base who would then abstain. This models generates candidate divergence in equilibrium, but only at the cost of contradicting the strong evidence that moderate candidate perform better in elections. One feature that all of the above models have in common is they are one period. Intuitively in a one period model parties can only care about two things: winning today and policy today. And indeed each of the above models gives one of these two reason as to why partisan candidates are nominated. Because my model is multi-period parties can care about something new: winning and policy in the future. This allows for a new reason why partisan are nominated, that is partisans energize the base and helps parties win in the future. This innovation can explain the nomination of sure losers, such as Goldwater in 1964. My model also produces three other testable implications that di erentiate it from the previous literature discussed above. These will be fully discussed and evidence presented in section 6. 4 The Model 4.1 Basics The model is an overlapping generation model with measure 1 of citizen who each live for 2 periods. The two generations are equally sized with measure :5 each. Everyone discounts the future at 2 (0; 1). Each period there is an election in which each citizen has the right, but not the obligation, to vote. Citizens are split into three equally sized groups based on ideology fl; m; rg 1. The l and r groups are called partisan citizens, and are associated with in nitely lived parties L and R. Type m citizens don t have a party and are referred to as moderate citizens. Each period the parties nominate a citizen candidate from one of the three groups to represent them in the election. For simplicity I suppose the parties will not choose a candidate from the other party s base. Thus parties make a binary choice between a moderate candidate or a partisan candidate from their own base. In order to avoid mixed strategy equilibrium the party that won last period has to choose their candidate rst. The candidate with the larger measure of votes wins the election. If the party that won last period decides to keep the same candidate, that candidates is known as an incumbent. In the event of a tie incumbent candidates win with probability p 2 (:5; 1). If there is no incumbent then ties are decided by a 50-50 coin ip. 1 The assumption that all three groups are the same size is not essential. It is important that the l and r groups are the same size. 7
To summarize, within each period the timing of the game proceeds as follows: 1. The party that won last period chooses a candidate 2. The party that lost chooses a candidate. 3. Each citizen chooses whether to vote, and who to vote for. 4. The candidate with the most votes win. Next I address when citizens vote and who they vote for. 4.2 The Choice to Vote In this model I follow Riker and Ordeshook (1968) and assume people vote because they gain a consumption value for voting. This could take many forms: voters could enjoy expressing their opinion, ful lling their moral or civic duty (as in Feddersen and Sandroni (2006)) or feel social pressure to vote. The alternative approach would be to assume people vote because they might change the outcome. This view is implausible in a large election, and becomes even more implausible if there is a cost to voting. And since the motivation of this model discussed in section 2 involves a xed cost the pivotal voter view is an unpalatable choice. Formally, voting in each period yields utility of D > 0. This payo is net of cost that a voter faces every time they vote such as time spent driving to the polling place, waiting in line etc. On top of this, rst time voters face an additional cost of i in order to vote. A citizen is energized when a candidate from his own ideological group is running, and has his xed cost from voting reduced by > 0 2. It can be easily veri ed that a citizen votes in their rst period of life if and only if they vote in their second period of life. In other words the citizens can be divided into habitual voters who always vote and non-voters who never vote. I restrict i to two values L and H where H > L. The two types are evenly distributed throughout the three ideological groups. The only restriction on how many of each type is that there is strictly positive mass of both. L types nd engaging with the political relatively painless (or even enjoyable) and always vote in both periods of their life. For simplicity I set L = 0. On the other hand H types need to be energized or else they will "abstain due to alienation". Speci cally if there is a candidate from the same ideological 2 Alternatively the citizen utility s from voting could increase by when energized and nothing would change. 8
group in their rst period of life they decide to become habitual voters and vote in both period of life. In all other cases they never vote. Formally this is given by: (1 + )D < H < + (1 + )D 4.3 Preferences over Outcomes The parties are in nitely lived with discount rate and are both policy and o ce motivated. In any period t let w t ; y t be the party and ideology respectively of the winning candidate. Thus w t 2 fl; Rg and y t 2 fl; m; rg. Party L maximizes: 1X t (v L (w t ) + u L (y t )) t=0 Where v L (L) = > 0; v L (R) = 0 and u L (l) = 0; u L (m) = :25; u L (r) = 1. The policy motive is consistent with quadratic loss in a spatial model and represents the degree of o ce motivation. R has symmetric preferences. The citizen s preferences on the ideology of the winning candidate are straightforward. Every citizen s rst choice is a candidate with the same ideology as them. A partisan citizens (l and r types) prefers a moderate to win over a candidate with the opposite ideology. And moderate citizens are indi erent between l and r candidates. Formally let u i (x) be the utility of a citizen with ideology i when a candidate with ideology x wins and then: u l (l) > u l (m) > u l (r) u m (m) > u m (l) = u m (r) u r (r) > u r (m) > u r (l) Citizen are non-pivotal and realize it, and so always vote sincerely according to these preferences. Partisan citizens break indi erence in ideology by siding with their own party, moderate citizens break indi erence by randomizing 50-50. Given this model I can establish the following results about who wins elections. Lemma 1 In every election the following statements hold: 1. A moderate candidate will always defeat a partisan candidate. 2. Suppose both parties choose moderate candidates. Then if party i chose a partisan candidate last period and party j chose a moderate last period than party i s candidate 9
will win. If both parties chose a partisan candidate last period, or neither did, then the election is a tie. Proof: See appendix: Lemma 1 formalizes the central logic of this work: that partisan candidate aren t as electorally viable as moderate candidate but help parties win in the future. Part 1 says a partisan can never defeat a moderate, which makes speci c how they are electorally unviable. Part 2 shows how partisan candidates help parties win in the future. Speci cally if both parties decide to win today and pick moderates then the election is decided by who energized their base last period. A result identical to part 2 would attain in the case when both parties picked a partisan candidate today. The next step is to see how these forces play out in equilibrium, which is the topic of the next section. 5 Equilibrium In this section I analyze the game described above using the solution concept of subgame perfect equilibrium. The model above will have a unique subgame perfect equilibrium, which allows for sharp predictions. This section is split into three parts. 5.1 discusses two key assumptions and introduces some useful notation. 5.2 presents the unique equilibrium in detail and gives a discussion of its important features. 5.3 provides a sketch of the proof. 5.1 Assumptions and Notation Throughout this section the following two assumptions are maintained Assumption 1 : < 1 + 2 2(1 ) ; Assumption 2 : > 1 p p The rst assumption says the parties can t be too highly o ce motivated and the second that they can t be too impatient. If the parties are patient enough (! 1), both conditions are always satis ed. Assumption 1 gives the allowed level of o ce motivation as a function of the patience of the parties. Assumption 1 is satis ed for any level of if is close enough to 1 or close enough to 0. Moreover less than 2:4 works for any. Assumption 2 says the parties have to be patient compared to the advantage incumbents get in the case of ties. As p! 1 any level of works, as p! :5 then needs to approach 1. Assumption 1 is not essential, in section 7 I show purely o ce motivated parties still retain the central behavior of conceding current elections to win later. The main di erence is with purely o ce motivated parties partisan candidates do win in equilibrium. Counterintuitively 10
this implies a policy motive makes it less likely for partisan policies to get enacted. On the other hand, assumption 2 is essential. In section 8 I show if the parties are too impatient then in the unique equilibrium only moderate candidates are ever chosen. It is intuitive that short sighted parties won t invest in future elections. And without this forward looking behavior my model largely resembles a standard one shot spatial model, and the logic of the median voter theorem applies. 5.2 Characterization and Discussion Here I characterize and discuss the unique subgame perfect equilibrium, which I denote by? = (? L ;? R ). The equilibrium ends up being Markovian, so it can be described using a state that includes the party that won and the positions taken in the previous period. To denote these states I will use the notation (I; x; y) where I is the party that won last period, x is that party s last period position, and y is the losing party s previous position. For example (R; m; l) means the R party won last period with a moderate candidate, and the L party lost with a partisan candidate. actions each party takes in each state. In the equilibrium the parties will use symmetric strategies, so only the states where the L party is the incumbent need to be described. This is done in the following table: State L s Position R s Position R s Position if L deviates (L; m; m) m r m (L; m; r) m m r (L; l; r) m m m Note the table doesn t include state (L; l; m) because it is impossible since the moderate candidate picked by party R can t lose to the partisan candidates picked by L. Proposition 1 Both parties employing strategy? is the unique subgame perfect equilibrium. Proof. In appendix By examining the table it can be seen that this equilibrium has the property that after any history the cyclical pattern displayed in gure 1 below results within two periods. Each box in the gure represents a period. The rst line in the box is the state, and the second line describes what happens. The rst thing to notice is a moderate candidate wins in every period, and yet partisan candidates are regularly chosen. To see why consider the lower left hand box with state (R; m; m) which means L is out of power and neither party has the advantage of an energized base. R goes rst and picks a moderate candidate. L could pick a moderate candidate which would result in a tie and give L probability 1 p < 1 2 of winning. Instead L decides to 11
Figure 1: Cycle in Equilibrium (R,m,l) L-m Beats R-m - (L,m,m) L-m Beats R-r 6 (R,m,m) R-m Beats L-l? (L,m,r) R-m Beats L-m concede the current election and pick a partisan. This leads to the upper left hand box (state (R; m; l)) where both parties pick moderates, but L wins for certain because they energized their base last period. Hence L trades o a chance of victory (at state (R; m; m)) for certain victory in the next period (at state (R; m; l)). Notice that 1 p represents how much L is giving up by picking a partisan. This makes clear the role of assumption 2, the parties have to be su ciently patient relative to what they are giving up by conceding the election. In the remaining two boxes, (L; m; m) and (L; m; r), the same story occurs with the role of the two parties switched. One feature that might seem puzzling is that R chooses a moderate candidate in the upper left box in state (R; m; l). R knows L s base is energized, and can foresee defeat, so why doesn t he choose a partisan candidate to build a stronger future? Well if R chooses a partisan then party L can also choose a partisan and win for sure (recall R has to go rst). And since R has a policy motive he prefers to lose to a moderate than lose to a partisan. That R acts in this way is the content of assumption 1. Thus the policy motive of the parties acts to prevent partisan candidates from winning, which is a counterintuitive result. This point is exactly how the equilibrium with purely o ce motivated di ers from the above cycle, and will be discussed in more detail in section 7. Another feature of this cycle is that parties alternate in power in a predictable fashion. This is a result of the fundamental tension between winning today and tomorrow. In equilibrium, only moderates ever win, but they fail to energize the base, and leave the winning party with a weakened future. This creates a testable prediction: that the longer a party is in power the more likely it is to lose the next election. This prediction, as well as two others, will be more fully discussed in section 6. That the model has a unique subgame perfect equilibrium is a striking feature. Uniqueness allows for sharp predictions, which is uncommon for in nite horizon games which are 12
usually plagued by multiplicity as is allowed to go to 1. Dutta (1995) provides a folk theorem that applies to a wide class of games with state dependent payo s, including my model. He shows that, under some regularity conditions, any feasible payo that is above the min max can be achieved with su ciently patient players. As will be discussed in more detail in 5.3, uniqueness holds because the equilibrium is both strongly e cient and pushes both players to their min max payo. Thus the equilibrium is the only arrangement of strategies that is individually rational for both players. The intuition for why? is unique is the model is almost a zero sum game. To see this note that if a moderate always wins then the only variation in payo s comes from who get the o ce motivated payo, which is zero sum. And as seen in Lemma 1 a moderate always defeats a partisan, so a partisan can only win if the other party allows it. And the parties won t allow it as long as they are su ciently policy motivated, which is guaranteed by assumption 1. 5.3 Sketch of Proof Verifying that? is an equilibrium is simply an application of the one shot deviation principle; see appendix for calculations. The calculations are made manageable by the fact the equilibrium is Markovian. Additionally the strict inequalities in the assumptions implies the one shot deviation principle holds strictly at every history, which will be used to prove the equilibrium is unique. To see why? is unique consider any history h that begins a period, so that it is the party in power s turn to move. Let V i? (h) be the continuation payo party i receives at history h if both parties are employing strategy?. The proof starts by establishing the following three statements: 1. At history h, each party can achieve at least V i? (h), no matter the strategy of the other party. 2.? is strongly e cient in the sense that the sum of continuations payo s at history h can never exceed V? L (h) + V? R (h) 3. 1 and 2 imply that in any equilibrium the continuation payo of both parties at history h must be V i? (h). That statement 3 follows from 1 and 2 is obvious. Statement 2 can be seen by noting that if a partisan candidate wins the parties combined payo is candidate wins it is 1 and if a moderate :5 and these are the only two possibilities. Thus combined payo is 13
maximized when a moderate always win, which happens in equilibrium? starting from any history that begins a period. To see statement 1 suppose that party L commits to playing strategy? L, look for the the strategy of R that minimizes L s payo. Notice that, no matter what R does,? L ensures an r type will never win. Thus it is intuitive the worst case for L will have a moderate winning in every period 3. But when a moderate always wins the game is zero sum, and so the strategy that minimizes L s payo is the strategy that maximizes R s payo, which is? R. Thus L s worst case is V L? (h) and hence he can always achieve this payo by playing? L. Given statement 3 the rest of the proof proceed as follows. Take any history h that doesn t start a period, i.e. where it is the out of power party s turn to move. WLOG let it be L s turn to move. Once L moves the next history will be one that starts a period. Using statement 3, for any action a, L 0 s continuation payo must be V L? ((h; a)). This, plus the fact that the one shot deviation principle holds strictly for? L, means that in any equilibrium L must take the same action? L at history h. And this can be used to show the continuation payo s for both parties at h must be V i? (h). And then this same argument can be applied again on histories that start a period, which completes the proof. 6 Three Testable Predictions The model delivers three testable predictions, each of which derives from the central tension between winning today and winning tomorrow. In this section I will discuss these predictions and present some empirical support for them found in the literature. Additionally I will argue that these predictions sharply distinguish my model from the previous models discussed in section 3. Prediction 1: The longer a party is in power the more likely they are to lose the next election The most fundamental consequence of the trade-o parties face between winning today or tomorrow is that the winning party sets itself up for a weaker future. In equilibrium only moderates win, and hence the winning party never energizes its base. A base that isn t energized hurts the party s future turnout leading to its eventual defeat. This dynamic gives rise to prediction 1. In the equilibrium presented this takes the sharp form that a party that has been in power for two periods loses the next election with certainty. This prediction is inter-temporal in nature which means it cannot be generated in a one period model, such as 3 The proof of this requires assumption 1. See Lemma A in the appendix for full details. 14
the previous models of candidate selection discussed in section 3. This make this prediction a novel one in the context of that literature. There is evidence supporting this prediction. Looking at U.S. presidential elections Abramowitz (1988) nds what he calls a "time for a change" e ect. He estimates that a party which holds the Presidency for two terms su ers a four percentage point penalty in the popular vote in the next election. This e ect is robust to including a number of controls including an incumbency e ect. The incumbency control is particularly important because of the two term limit for U.S. Presidents. Along similar lines Lewis-Beck and Nadeau (2004) looks at data from the United Kingdom and nds a 3 percentage penalty for each term a party holds power. Prediction 2: Parties diverge in the ideology of the candidates chosen, but not in policies implemented A key feature of the equilibrium presented above is that moderates always win and yet partisans are still chosen. This runs in sharp contrast to the models in section 3 which had partisans winning in equilibrium. Indeed I argue this prediction cannot be generated in a one period model without contradicting the strong empirical fact (discussed in section 3) that moderate candidates perform better in elections. Intuitively parties should care only about two things: winning and policy. So in a one period model only winning today and policy today can matter. But prediction 2 speci es convergence in policy between the parties, so policy can t be the reason for choosing a partisan. And, as already stated, moderates are better at winning. Thus it becomes impossible to explain why partisan candidates are ever chosen in a one period model. My model solves this by adding something new for parties to care about: winning and policy in future periods. There is moderate evidence to support this prediction. The evidence is limited in that it focuses almost exclusively on the relationship between the ideology of the government in power and government spending as percentage of GDP. Imbeau (2001) surveys 43 such studies of which 29 nd no signi cant relationship. Additionally, many of the studies that found a signi cant relationship found a very small one. For example Blais et al. (1996) nds that a government controlled entirely by the left only spends 0.4% of GDP more than an entirely right controlled government. Prediction 3: Parties newly out of power nominate more partisan candidates Another consequence of parties making inter-temporal choices about who to nominate is there can be predictable time series pattern in that choice. In the equilibrium presented 15
above the parties always (and only) nominate a partisan candidate in the period after losing power. The logic behind this is while in power a party tends to neglect its base, so it needs to reenergize it before being able to reclaim power. Again the inter-temporal nature of this prediction means it could not be generated in any of the one period models of candidate selection discussed in section 3. There is evidence to support this prediction. Cohen et al. (2008) looked at the relationship between the ideology of the candidate of the out of power party and the number of years the party has been out of power. They focus on U.S. presidential elections and rate the ideology of the candidates using the National Election Survey. They nd the longer the party has been out of power the more moderate the candidate. This directly matches up with prediction 3. Their result can be seen graphically in gure 2 below, which is taken directly from their book. Figure 2: Taken from Cohen et al. (2008) pg. 92 16
7 O ce Motivated Candidates In this section I consider the case where parties are purely o ce motivated. I nd and discuss the unique subgame perfect equilibrium. The key feature of the original equilibrium, that parties concede election to invest in the future, is preserved. The main di erence from section 5 is that partisan candidates win in equilibrium. This gives the counterintuitive result that a policy motive enforces a median voter result in the outcomes of elections. The only change to the model is to the parties payo. Party L now maximizes: 1X t (v L (w t )) t=0 Where w t 2 fl; Rg is the party of the winning candidate and v L (L) = 1 and v L (R) = 0. Party R s preferences are symmetric. impatient, so I impose assumption 3: It is still important that the parties are not too Assumption 3 : p > 1 1 2 2 1 + 2 This is more complicated than assumption 2 but serves much the same role, and gives the required level of patience as a function of the incumbency advantage. I describe the strategy pair? o ce =? L;o ce ;? R,o ce in the following matrix. Just as in? the equilibrium is symmetric so I only describe the states where L is in power. State L s Action R s Action R s Action if L deviates (L; m; m) m r m (L; m; r) l r m (L; l; r) m m m Proposition 2? o ce is the unique sub game perfect equilibrium. Proof: In Appendix. By examining? o ce it can be seen that after any history, within 2 periods, the following cycle results. Unlike the previous cycle this one involve ties. Where the ties occur the party in power did not keep their incumbent candidate, so the election is decided by a fair coin ip. The boxes where the ties occur have two arrows coming out to indicate the two possibilities, each of which occur with probability 1. Conceding the current election to invest in the future still 2 occurs at states (R; m; m) and (L; m; m) (the two center boxes), which is the same states where it happened in the original equilibrium. Thus purely o ce motived parties maintain the central feature of the original equilibrium. 17
Figure 3: Cycle in Equilibrium (R,r,l) R-m ties L-m 6 (L,m,r) R-r Beats L-l 1 * 2 H HHj 1 2 (R,m,m) R-m Beats L-l (L,m,m) L-m Beats R-r * 1 2 HY H H 1 2 (R,m,l) L-l Beats R-r? (L,l,r) L-m ties R-m The main di erence from the original equilibrium occurs at states (R; m; l) and (L; m; r), lower left and upper right boxes respectively. Under?, in these states both parties move to the middle and the side with the energized base wins with certainty. In? o ce the side with the energized base still wins, but both parties choose partisan candidates. The reason for this is the party without the energized base knows they are going to lose no matter what. Their only motive for taking a moderate position is to keep a partisan from the other side out of power. Since there is no policy motive they don t care about that and choose instead to pick a partisan candidate to build a stronger future. This provides the counterintuitive result that a policy motive prevents partisan candidates from winning. 8 Short Sighted Parties In this section I consider parties who are short sighted. Formally I take the opposite inequality of assumption 2 and suppose < 1 p p. For simplicity I take the case that parties are completely o ce motivated as in section 7, but the results would generalize to any level of o ce motive. Consider the strategy pair? short =? L;short ; R,short?, which is described in the following matrix. 18
State L s Action R s Action R s Action if L deviates (L; m; m) m m m (L; m; r) l r m (L; l; r) m m m Proposition 3 Suppose < 1 p p is the unique subgame perfect equilibrium. and the parties are completely o ce motivated. Then? short Proof: In appendix. The only di erence between the equilibrium here and the one in section 7 is that at states (L; m; m) and (R; m; m) both parties pick a moderate candidates and the election is a tie. So with probability p the state remains same and with probability 1 p the state switches to either (L; m; m) or (R; m; m), whichever is not the current state. This means once one of these two states are reached the equilibrium will stochastically alternate between them forever. And it can be veri ed that any of the other states will lead to either (L; m; m) or (R; m; m) within two periods. This result could be generalized to policy motivated parties as well. The only change would that at state (L; m; r) the L party would take action m and similarly for the state (R; m; l). So with short sighted parties the equilibrium is one in which both parties converge to the center in every period. This occurs because when parties ignore the future my model strongly resembles the standard one period model with no uncertainty. Myopic parties are not willing to lose today to invest in the future, and thus lose their only incentive to pick partisan candidates. 9 Conclusion In this work I argued that partisan candidates are less electorally viable, but energize the base which is valuable for future elections. I motivated this mechanism with empirical evidence on habitual voting and abstention due to alienation. Previous formal models of candidate selection have largely been one period in which parties can only care about who wins today and what policy gets enacted today. By introducing parties that are forward looking I allow for a new reason to pick partisan candidates: winning in the future. The model also generates three novel testable predictions, which match evidence found in the literature. The core innovation of this work is that parties also care about winning and policy in the future. That is parties are patient enough to consider the rami cation of their actions on future elections. This work has focused speci cally on the mechanism of partisan candidates energizing the base. But this is just one example of a larger principle; exploring other ways 19
forward looking behavior can impact political strategies presents an important direction for future work. References Abramowitz, A. (1988). An improved model for predicting presidential election outcomes. PS: Political science and politics, 21(4):843 847. Adams, J., Dow, J., and Merrill, S. (2006). The political consequences of alienation-based and indi erence-based voter abstention: Applications to Presidential Elections. Political Behavior, 28(1):65 86. Adams, J. and Merrill III, S. (2003). Voter turnout and candidate strategies in American elections. Journal of Politics, 65(1):161 189. Ansolabehere, S., Snyder Jr, J., and Stewart III, C. (2001). Candidate positioning in US House elections. American Journal of Political Science, 45(1):136 159. Blais, A., Blake, D., and Dion, S. (1996). Do Parties Make a Di erence? A Reappraisal. American Journal of Political Science, 40(2):514. Brody, R. (1977). From life space to polling place: The relevance of personal concerns for voting behavior. British Journal of Political, 7(03):337. Calvert, R. (1985). Robustness of the multidimensional voting model: Candidate motivations, uncertainty, and convergence. American Journal of Political Science, 29(1):69 95. Cohen, M., Karol, D., Noel, H., and Zaller, J. (2008). The Party Decides. University of Chicago Press, Chicago. Downs, A. (1957). An economic theory of political action in a democracy. The Journal of Political Economy, 65(2):135 150. Dutta, P. (1995). A folk theorem for stochastic games. Journal of Economic Theory. Erikson, R. and Wright, G. (2005). Voters, Candidates, and Issues in Congressional Elections. In Dodd, L. and Oppenheimer, B., editors, Congress Reconsidered, pages 77 106. CQ Press, 8th edition. Feddersen, T. J. and Sandroni, A. (2006). A Theory of Participation in Elections. American Economic Review, 96(4):1271 1282. 20
Franklin, M. (2004). Voter turnout and the dynamics of electoral competition in established democracies since 1945. Cambridge Univ Pr. Gerber, A. S., Green, D. P., and Shachar, R. (2003). Voting May Be Habit-Forming: Evidence from a Randomized Field Experiment. American Journal of Political Science, 47(3):540 550. Goldwater, B. (1988). Goldwater. Doubleday, New York. Green, D. and Shachar, R. (2000). Habit formation and political behaviour: Evidence of consuetude in voter turnout. British Journal of Political Science, 72(1978):561 573. Imbeau, L. (2001). Left-right party ideology and government policies: A meta-analysis. European Journal of Political Research, pages 1 29. Lewis-Beck, M. and Nadeau, R. (2004). General election forecasts in the United Kingdom: a political economy model. Electoral Studies, 23:279 290. Middendorf, J. (2006). A Glorious Disaster: Barry Goldwater s Presidential Campaign and the Origins of the Conservative Movement. Basic Books, New York. Miller, W. and Shanks, J. (1996). The new American voter. Harvard University Press Cambridge, Cambridge. Palfrey, T. R. (1984). Spatial Equilibrium with Entry. The Review of Economic Studies, 51(1):139. Plane, D. and Gershtenson, J. (2004). Candidates ideological locations, abstention, and turnout in US midterm senate elections. Political Behavior, 26(1):69 93. Plutzer, E. (2002). Becoming a habitual voter: Inertia, resources, and growth in young adulthood. American Political Science Review, 96(01):41 56. Riker, W. and Ordeshook, P. (1968). A Theory of the Calculus of Voting. The American Political Science Review, 62(1):25 42. Roemer, J. (1994). A theory of policy di erentiation in single issue electoral politics. Social Choice and Welfare, pages 355 380. Segal, D. (1968). Partisan realignment in the United States: the lesson of the 1964 election. The Public Opinion Quarterly, 32(3):441 444. 21
Wittman, D. (1983). Candidate Motivation: A Synthesis of Alternative Theories. The American Political Science Review, 77(1):142. Zipp, J. (1985). Perceived Representativess and Voting: An Assessment of the Impact of" Choices" vs." Echoes". The American political science review, 79(1):50 61. 22