- “...Under the electoral system, which this Clause proposes to set up [the Alternative Vote], no one is to have what he wants but everybody is to have their second preference. Why is a proposal of this kind brought before this House? It reminds me of a saying of Oscar Wilde regarding Whistler that: Whistler had no enemies, but he was intensely disliked by his friends. It would not be quite true to say that this Clause has no enemies; I am one, but it is true to say that it is intensely disliked by its friends”.
Sir Austen Chamberlain in H.C. Debs, March 4, 1931 Col. 524
On the 5th of May 2011 a majority of 68.9 percent of the British voters rejected a proposal to change the electoral system from the-First-Past-the-Post to the Alternative Vote (AV). The turnout was 42 percent. In the wake of the referendum several questions were asked? How could the yes-Campaign (which had enjoyed a double digit lead in January 2010) lose by a similar margin? Did the better financed side win by virtue of deeper pockets? Was the result driven by cues from party-leaders rather than based on enlightened decisions by the voters? And are referendums on electoral reform always unsuccessful?
Table One: Outcome of the 2011 Referendum on the Alternative Vote in the United Kingdom
Turnout: 42 percent
Yes: 32.1 percent (6152607)
No: 67.9 percent (13.013123)
Source: The Electoral Commission
Needless to say, some of these questions cannot be determined with any degree of certainty especially as we still await comprehensive survey data. However, based on referendums in other countries on similar and related issues, as well as compared with previous referendums in the United Kingdom, we can put together coherent picture that enables us to rise above the apparently idiosyncratic and seemingly unique factors that ostensibly determined the outcome of this plebiscite. In doing so we can conclude that the referendum was not, in fact, that unique, but rather followed some familiar patterns identified in other referendums around the world on electoral reform.
Reforms of the electoral system are seen as fundamental constitutional changes; i.e. changes that should not be undertaken lightly as the dangers of gerrymandering are ever present[1]. In the United Kingdom - it has almost become a convention of the constitution that electoral reforms and changes to the electoral system must be preceded by a referendum [2]. It was, therefore, not surprising that the Conservative-Liberal Democrat coalition government included this commitment in the Coalition Agreement:
- The parties will bring forward a Referendum Bill on electoral reform, which includes provision for the introduction of the Alternative Vote in the event of a positive result in the referendum, as well as for the creation of fewer and more equal sized constituencies. Both parties will whip their Parliamentary Parties in both Houses to support a simple majority referendum on the Alternative Vote, without prejudice to the positions parties will take during such a referendum
That this pledge was included was a result of bargaining. The Conservative Party – having campaigned against all forms of electoral reform – made a referendum on AV their “last offer” and the Liberal Democrats (a party that had campaigned for the more proportional Single Transferable Vote or STV) accepted this offer[4]. This concession from the Conservatives came after Gordon Brown had ostensibly proposed to put AV into law without a referendum, though the latter has not been independently confirmed[5]. That the Liberal Democrats choose not to go with Brown – despite this offer – may suggest that electoral reform without a referendum would be illegitimate, though it is probably equally, if not more, plausible to cite fundamental personal differences and the sense among the Liberal Democrat leadership that Gordon Brown and Labour – having just lost the election – lacked legitimacy[6]. In other words, the decision to hold a referendum was a result of bargaining; an agreement to disagree. Writing about the referendums in the 1970s, Dennis Kavanagh concluded that “the referendum had more to do with political expediency than constitutional principle or democracy”[7]. The same conclusion could safely be drawn in 2010-2011.
The Legislation and the Run-Up to the Campaign
Following the publication of the coalition agreement, the government moved swiftly and introduced the Parliamentary Voting System and Constituencies Bill in the House of Commons on 22 July 2010, which only met with cosmetic challenges[8]. The Bill passed its Third Reading in the Commons on 2nd November by 321-264. The aim of the legislation was two-fold, 1) to reduce the number of Constituencies from 650 to 600 (A Conservative manifesto commitment) and 2) to introduce (subject to a referendum) the Alternative Vote. Somewhat to the embarrassment of the Labour leader Ed Miliband (who was in favour of AV), several Labour peers (who were opposed to AV) sought to wreck the passage of the Bill in the House of Lords by imposing a forty-percent turnout threshold. Such a threshold had effectively sealed the fate of devolution in Scotland in 1979[9].This threshold would almost certainly have condemned the proposal to failure, as it has elsewhere. The amendment proposed by Labour’s (Lord) Charlie Faulkner (a former cabinet minister) failed – though only after all night sessions and bitter debates. The Bill was finally enacted on 16 February 2011 when it was passed by the House of Lords by 224 votes to 210.
Perhaps paradoxically, it was the Labour Party’s Manifesto A Fair Future for All which had included a commitment to holding “a referendum on introducing the Alternative Vote for elections to the House of Commons”, in order to “ensure that every MP is supported by the majority of their constituents voting at each election”[10]. Yet, despite their commitment to this system, many Labour politicians grew sceptical of the proposed change. While not officially hostile to AV, the Labour Party leadership seemed to have second thoughts and prominent members sowed doubts about the commitment to changing the system and to their contribution to the campaign. Andy Burnham – Labour’s Election Campaign Co-coordinator - expressed the view that Labour will not campaign for alternative vote, but will instead “focus…on [the] Scottish, Welsh and local elections taking place on same day”[11]. To be sure Ed Miliband, the party leader, threw his weight behind the change, but he was hampered by the fact that many prominent members of his party, such as Margaret Beckett (the former foreign secretary), John (now Lord) Prescott (the former deputy prime-minister), and former Home, Defence and Northern Ireland Secretary John (now Lord) Reid campaigned alongside David Cameron for a no-vote. In the end more than 150 Labour MPs were opposed to AV, and only 50 declared their support.
While many in the Labour Party campaigned against their manifesto commitment, the reverse was true for the Liberal Democrats. The small government party had forgotten their misgivings about the Alternative Vote and was actively campaigning for its introduction[12]. This was a bit of a volte-face, especially as the Liberal Democrats were at best lukewarm at the idea of introducing AV before election. Indeed Chris Huhne, the Climate Secretary, had observed that “the alternative vote is not the solution”, and opined that “only the single transferable vote will remedy the unfairness of the present system”, though he did admit that AV was “a small step in the right direction”[13]. Similarly Nick Clegg had referred to AV as a “miserable little compromise”[14].
Referendums on electoral reform are not as unsuccessful as most people may be inclined to think[15]. The adage, ‘if in doubt vote no’ does not universally apply to electoral reform referendums. Nineteen referendums held since 1980, nine have been successful (Andorra 1992, Russia 1993, Uruguay 1996, Ecuador 1997, New Zealand 1992, New Zealand 1993, Canberra 1992 and Italy 1992 – and under very special circumstances Iraq in 2005). Ten have been unsuccessful, though of these five were due to failure to meet the turnout requirement, namely in, British Columbia (2005), Italy (1999, 2000, 2009), and in Romania (2007). Though of course, without a referendum things are much easier. Of the 50 electoral system changes enacted without a referendum, law passed by the Czech parliament in 2000 is the only one that was not implemented. The law was vetoed by President Havel on the grounds that it violated the constitutional protection of the principle of PR[16].
While public enthusiasm for electoral reform is not great – the referendums in New Zealand in 1993, the Italian referendum in the same year, and Uruguay poll in 1996 are the only three examples of votes that have recorded turnout of more than 70 percent – it is inaccurate to suggest that electoral reform referendums are invariably a lost cause. Like all other referendums, polls on electoral reform are subject to ebbs and flows of popular support and the miscellaneous factors that determine the fate of political campaigns. It is beyond the scope of this chapter to give a blow-by-blow account of the campaign, and that such accounts have already been published by insiders in each of the two camps[17]. What is interesting in this context, and from the point of view of comparative politics, is the question as to whether the referendum fits the various patterns identified by political scientists studying referendums abroad. While political science studies of referendums were relatively – not to say entirely – empirical and case-studies based until the turn of the end of the 20th Century[18], political scientists have recently developed more advanced and comparative models and theories for studying referendums[19]. While it is outside the scope of this chapter to present a thorough overview of the literature, a few key findings are useful for understanding the 2011 referendum.
Table Two: Outcomes of Electoral Reform Referendums since 1980
|Country |Yes-Vote |Year |Success/Failure |
|Andorra |74 |1992 |Success |
|British Columbia |57 |2005 |Failed due to Turnout |
|British Columbia |38 |2009 |Failed |
|Canberra |65 |1992 |Success |
|Ecuador |59 |1997 |Success |
|Iraq |78 |2005 |Success |
|Italy |82 |1993 |Success |
|Italy |91 |1999 |Failed due to Turnout |
|Italy |82 |2000 |Failed due to Turnout |
|Italy |77 |2009 |Failed due to Turnout |
|New Zealand |84 |1992 |Success |
|New Zealand |54 |1993 |Success |
|Ontario |37 |2007 |Failure |
|Prince Edward Is. |36 |2005 |Failure |
|Romania |83 |2007 |Failed due to Turnout |
|Russia |58 |1993 |Success |
|Slovenia* |14 |1996 |Failure |
|St. Vincent & Grenadines |43 |2009 |Failure |
|Uruguay |51 |1996 |Success |
Sources Australian Electoral Commission, Elections Canada, Zentrum für Direkte Demokratie, Aarau (ZDA)),
Butlletà Oficial del Principat d'Andorra 1993, *Multi-option referendum none of the options secured 50 percent of the votes cast.
As suggested elsewhere by this author, referendums often follow an underlying logic and while it might be difficult to find law-like generalities like in the physical sciences, there are certain patterns which often repeat themselves in referendum campaigns[20]. Two of the most commonly recognised regularities are:
-* Referendums tend to be won by parties that capture the centre-ground and effectively appeal to the median voter[21]. Thus, the Irish referendum on divorce in 1986 was won by opponents as they successfully appealed to the centre ground and turned the issue into a matter of social security rather than a question about Catholic philosophy[22];
-* Citizens often decide on the basis of cues, such as ‘who’s behind it’ rather than based on an analysis of the issues[23], thus referendums are often won because voters use short cuts to acquire information rather than studying the finer points of the issues.
The Median Voter and the AV Referendum
Proponents of referendums – who have often argued that referendums lead to a more mature politics – may have been shocked by the campaign and the tone, attacks and apparent disinformation. The No-campaign more or less invented – by their own subsequent admission[24] - a claim that the introduction would cost more than 250 million pounds. And the Yes-side – despite Deputy Prime Minister Nick Clegg’s call for a ‘mature debate’ – compared the Conservative Chairman Baroness Sayeeda Warsi to Dr Josef Goebbels![25] These colourful interventions might have be entertaining, but they are more interesting because they can be interpreted as vindications – or perhaps, better – illustrations - of some of the trends and tendencies identified in other referendums around the world.
Realising that the Yes-to-fairer-votes were in line with the median voter (a majority of voters were in favour of electoral reform[26]), the No-to-AV camp effectively utilized a familiar tactic of referendum campaigns, namely that of capturing the middle-ground by changing the agenda. This tactic had been tried before in similar referendums. In 1993, the opponents of electoral reform in New Zealand sought with some success to change the agenda to economic issues – and the alleged reactions of the market by introducing the Mixed Member System[27]. This tactic differed considerably from the tactic followed in the first referendum in 1992, in which the proponents of reform had been allowed to outline the problems with the first-past-the-post system (which on two occasions had allowed the party that did not win the highest number of votes form a majority government)[28]. Although the No-side in New Zealand lost, they were able to narrow the margin considerably to 54-46 – a considerable feat given that the first referendum had been lost y a 84.7-15.3 percent margin[29]. This tactic of winning the middle ground, which also was employed in British Columbia to defeat the introduction of STV in 2009, ensured a reversal of the 2005 result that had resulted in a victory for the pro-reformers[30].
No to AV combined two strategies. The first of was focussed on the unpopularity of Liberal Democrat leader Nick Clegg. The main leaflet from the No campaign featured an unflattering photo of Mr Clegg holding a placard with a pledge to vote against tuition-fee rises and the caption ‘Nick Clegg is unpopular because he broke his promises’, a reference to the Liberal Democrats’ much debated decision to vote for student fees after the election. This much publicised leaflet -as well as billboards with the same message - effectively turned the poll into a referendum on the popularity – or otherwise – of the Liberal Democrat Leader[31].
In addition to this strategy, No-to-AV focussed on the alleged economic implications of a yes-vote. The opponents of electoral reform relied on, what a senior No-campaigner called a “simple, if contentious...figure of 250 million”[32] (a figure that suggested that the cost of AV would be over 250 million pounds due to the cost of voting machines). In full pursuit of their strategy, the opponents explained that the cost of voting machines and the cost of explaining the system to the voters could pay for 2,503 doctors, 35,885 hip replacements or 69,832 school places[33]. With these claims – dubious though they were - the No-to-AV effectively won the battle over public opinion, and was able to set the agenda to which the Yes-to-fairer-votes had to react and respond.
Yes-to-fairer-votes -dominated by the Liberal Democrats and reform groups such as Unlock Democracy and the Electoral Reform Society- took the bait, and reacted with fury. Chris Huhne, threatened with legal action – an empty threat as the Political Parties, Elections and Referendum Act 2000 does not give the Electoral Commission powers to intervene[34]. This intervention effectively secured that the claim became the focus of the debate. By reacting to the claim and by seeking to counter it, the proponents of reform moved the focus away from the discussion about the electoral system (on which they were in agreement with the voters) to the issues of economic reform.
Whether it was the 250 Million claim or the targeting of Nick Clegg ‘wot won it’ is debatable, but it is inarguable that the yes-campaign never recovered from the claim, and that they almost immediately lost the lead in the polls, which they had held – and even deepened - since polling on the issue began in earnest in December 2010.
Table Three: Poll of Opinion Polls 2010-2011
|Date |Yes |No |Don't Know |
|May-10 |59 |32 |10 |
|Oct-10 |41 |36 |18 |
|Nov-10 |48 |35 |17 |
|Dec-10 |36 |30 |34 |
|Jan-11 |37 |20 |37 |
|Feb-11 |40 |30 |30 |
|Mar-11 |33 |30 |30 |
|Apr-11 |39 |53 |9 |
Sources: Angus Reid, YouGov, ComRes, ICM
What is surprising, especially given that the yes-campaign was comprised of seasoned campaigners (like Gordon Brown’s former spin doctor Paul Sinclair) was that the Yes-side did not seek to come up with a counter-claim. The truth is, perhaps, that no issue was more salient than the No-to-AV’s second poster campaign ‘No to President Clegg’; a poster that played directly to the Labour voters’ hatred of the Liberal Democrat leader. The No-strategy – based on focus groups conducted by the Australian pollster Lynton Crosby (a man who made his name by securing four consecutive electoral victories for the Centre-Right under the AV system in Australia[35]) - was effectively based on the most salient issues of the day.
Conversely, the Yes-to-fairer-votes slogan, ‘Make your MP Work Harder’ had not been tested by focus groups[36] . Moreover, many MPs found it hard to campaign under a banner that effectively said that they were lazy and corrupt, and the Yes-to-Fairer Votes “bottled it and shelved it”, when “MPs kicked up about their anti-MP line”[37].
One of the reasons that the Yes-to-fairer-votes strategy never came off the ground could be that the proponents were short of cash. This was not the case. In December the Yes-to-fairer-votes campaign reportedly outspent their opponents by a factor of 3-1, and the no-camp struggled financially. As one no-campaigner lamented, “We couldn’t sign off budgets, which meant that we couldn’t buy add-space, and that meant that we couldn’t formally launch”[38]. In the end the No-to-AV raised funds, and outspent their opponents by a 2:1 factor. The No-Camp spent £3.4 million compared to £2.6 million by the No campaign[39]. Yet, the disparity in spending, say, when compared to the 1975 referendum in which the pro-marketeers outspent their opponents by a factor of 10:1 and were “pitifully short of money”[40], was not great. Both camps had received grants from the Electoral Commission, and it would difficult to sustain the view that the proponents lost due to lack of funding.
Off course, elsewhere, there is some evidence that campaigns can be won by the biggest spenders. But the correlation between spending and outcome is far from linear or conclusive[41]. Indeed, there is no evidence from comparative data that suggest that the richer side inevitably wins. What matters is not the amount of money spent, but rather how the money is being spent[42]. The 2011 referendum is no exception to the rule. The Yes-camp lost because they were unable to win the median voter and failed to target the swing-voters, and because they made a number of cardinal mistakes, such as uncritically relying on slogans that failed to capture the headlines or appeal to those who would have been positively disposed towards reform of the electoral system.
The defeat of the yes-campaign was in large measure the result of a failure to use communication strategies that money can buy. As a commentator put it “the yes-campaign never managed to make better arguments for AV than that it would “make your MP work harder” (although it never explained how)”[43]. But there were other reasons for the defeat.
Short Cuts, Cues and Policy Preferences
Recent years studies of referendums have taken on board a body of theoretical research known as the ‘Reasoning voter paradigm’[44]. The argument in this body of literature is that voters rather than acquiring costly information about referendums (and candidates) use “information short-cuts” and base their decisions on stated preferences of friends, elites and even celebrities whom they trust or otherwise[45]. The yes-camp – whether consciously or not – based their campaign on this tactic. It was assumed – though this was apparently not tested in focus groups[46] – that voters would react positively to celebrity endorsements. Thus Yes-to-fairer-votes enlisted the support of Oscar winning actor Colin Firth (who informed the voters that the Oscars were decided by the Alternative Vote[47]) and the public campaign was dominated by the comedian Eddie Izzard and the actor Richard Wilson of the aptly named television series One Foot in the Grave. But the Yes-to-fairer-votes did not anticipate – and seemed unaware – that such endorsements have had negative effects elsewhere. In Sweden in 2003, the support of a prominent businessmen and celebrities contributed to giving the yes-side an elitist outlook that effectively alienated the ‘ordinary’ voters, and contributed to the defeat of the proposal to introduce the Euro[48]. The same was seemingly true for Britain.
But there are other cues than those of celebrities. The ‘cues’ of the political parties are often assumed to be an indicator of how citizens vote. In referendums on European integration, for example, voters have often taken cues from party leaders[49]. This is also true for referendums elsewhere. Based on a study of the referendum in term-limits in Washington in 1991, Jeffrey Karp found that “the probability of supporting the initiative [depended] on feelings towards Speaker Tom Foley. Those with negative feelings towards Foley [were] almost twice as likely to vote for the initiative [which Foley opposed]”[50]. Whether the same is generally the case in referendums on electoral change is a more difficult question to ask as the number of cases is significantly smaller. But based on the limited comparative evidence there are indications that this is not generally the case. Voters in referendums on electoral change at the national level have not tended to follow cues, but have seemingly departed from the party-line. In Romania the President Traian Băsescu organised the referendum on the electoral system on the same day as the European Parliamentary Elections in 2007, and although he won a majority for his preferred option (a French two-round system), his party ALDE lost its two seats in the European Parliament on the same day. The referendum, which was ostensibly indented as a vehicle for gaining momentum for his party did not help! The same pattern much earlier was observed in the Republic of Ireland in 1959, when Eamon de Valera decided to call a referendum on the introduction of the First-Past-the-Post Electoral system on the same day as he contested the presidency[51], and in 1969 when Fienna Fail lost another referendum on the same issue, but won the election .Further, in New Zealand 1992 and in 1993 there was very little correlation between party preferences and voting choice in the referendum in New Zealand – indeed the old parties won 83 percent of the votes in the General Election on the same day as they lost the referendum in the latter year[52].
Does Britain fit the pattern? The Conservative Party was the only political party to oppose the referendum. Based on the ‘Elite Cue Theory’ espoused by Karp we would expect that the voters in conservatively dominated areas – i.e. areas where the Tories tend to win parliamentary representation – would have been the areas most likely to have rejected the introduction of AV. There is some indication that this was the case. The Southeast duly rejected the introduction of AV by one of the highest margins. The same rejection was equally unequivocal in the Labour heartlands in the North. The only thing that suggests that the voters did not take cues come from the South West, traditionally a heartland area for the Liberal Democrats. Here, in areas where one would, perhaps, have expected a stronger showing, an equally high number voted no (See Table Six).
Table Four: The Result of the 2011 Referendum on the Alternative Vote
|Region |Turnout |Yes |No |
|Scotland |1.975549 |36.4 |63.6 |
|Northeast |762300 |28.1 |71.9 |
|Northwest |2048744 |30.2 |69.8 |
|Yorks/Humber |1530921 |31.3 |68.7 |
|East Midlands |1432291 |28.7 |71.3 |
|West Midlands |1629969 |28.5 |71.5 |
|Wales |946942 |34.5 |65.5 |
|Eastern |1839419 |29 |71 |
|London |1860253 |39.5 |60.5 |
|South East |2786488 |29.5 |70.3 |
|South West |1797276 |31.5 |68.5 |
|Northern Ireland |668870 |43.7 |56.3 |
Source: Electoral Commission 2011
Only in Scotland – where the SNP campaigned for AV and won a majority of seats in the Scottish Parliament elections on the same day, was there an indication that voters did not take cues.
Needless to say, we cannot extrapolate from aggregate figures to individual preferences. To do so would be an ecological fallacy[53]. But the indications suggest that voters were taking cues from parties as suggested by Karp. The prominent constitutionalist – and advocate of referendums - A.V. Dicey once suggested that voters would be able to “distinguish between measures and men”[54], and would not be swayed by party allegiances. This might have been the case in Ireland in 1959, In New Zealand in 1993 and in Romania in 2007. It was seemingly not the case in the United Kingdom in 2011.
Bare-Knuckle Fight or Public Spirit?
The referendum in 2011 was a tough political battle. No quarter was given, and none was taken. The contrast to the only previous nationwide referendum was seemingly marked. Vernon Bogdanor, has described how the 1975 campaign “encouraged a sense of social unity by enabling some political activists to reach out across the party lines and establish contact with those holding similar views”[55]. By contrast the 2001 referendum was described by Vincent Cable – a Liberal Democrat Cabinet minister – as “ruthless, calculating and really rather tribal”[56]. It would appear that the 2011 referendum did not live up to the praise metered out to such exercises in direct democracy by the likes of Professor Bogdanor, and that it was closer to the description offered by American political scientist David Magleby who concluded that referendums serve to “intensify conflict and lead to a politics of confrontation”[57].
The 2011 referendum was, to be sure a bare knuckle fight. Most referendums are[58]. Yet the conclusion reached by Vince Cable is not entirely accurate, and while the campaign was undeniably tough and not a shining example of public spirit and deliberative democracy, it remains the case that erstwhile enemies from the Labour and the Conservative parties cooperated. The No-campaign was described as “a Conservative head with a Labour heart”[32], and there were according to press reports several remarkable “features of cross party alliances”[60]. In 1975, Butler and Kitzinger found that almost every group seemed to have come to the “realisation how easy it was with a little effort, to find common ground on a range of issues”[61]. The same conclusion could be drawn from the 2011 referendum.
Conclusion
Referendums on electoral reform are extremely rare[62], though contrary to the impressions, referendums are not generally unsuccessful. Of the nineteen referendums on electoral reform since 1980 nine have been successful and ten have been unsuccessful, and of these five, or half of them, failed due to turnout requirements and not due to outright rejection by the voters.
The campaign showed many traits seen in other similar referendums. The evidence from the campaign suggests that the opponents of AV were able to set the agenda and convey the impression that they represented the median-voter. Like in Ireland in 1986, the no-campaign did this by changing the agenda and the focus of the debate away from the issue and into a debate about unrelated issues such as the economy and the unpopular cuts in public spending.
As proposed by proponents of the Reasoning Voter Paradigm, the voters took cues from politicians’ positions on the issue, and the no-side effectively used this tactic to turn the referendum into a vote of no confidence in the exceptionally unpopular Liberal Democrat leader Nick Clegg. The result was a massive defeat.
Does this mean that electoral reform is a dead issue in British politics? It is too soon to tell, but it is certain that none had much enthusiasm for the Alternative Vote, but given the lack of enthusiasm among proponents of AV and the fierce determination shown by No-to-AV it is perhaps appropriate to quote Nicoló Machiavelli’s quip about the difficulties of undertaking large-scale political change:
- There is nothing harder to undertake, nothing more likely of failure, nothing more risky to pull off, than to set up oneself as a leader who plans to found a new system of government. For the founder makes enemies of those who are doing well under the old system, and has only lukewarm support from those who hope to do better under a new one... So it is that whenever those who are enemies of a new order have a chance to attack it, they do so ferociously, while others defend it half-heartedly
- .
Matt Qvortrup est maître de conférence de l'Université de Cranfield (Resilience Centre), spécialiste des questions électorales et en droit constitutionnel comparé. Matt Qvortrup est docteur en droit (Oxford University) pour une étude portant sur les référendums qui fut ensuite publiée sous le titre : A Comparative Study of Referendums (Second Edition 2005). Son prochain ouvrage s'intitule Balloting to Stop Bullets. Referendums on Nationalism and Ethnic Issues (à paraître, en 2012, University of Pennsylvania Press).