The problem of poverty among old-age pensioners

7One of the aims of the Pensions Act of 21 August 2003 (the so-called “Loi Fillon”) was to increase small pensions. One of these, the minimum vieillesse—the first social policy for minimum income enacted in France, in 1956  [16]—was paid to 586,000 people in 2007.  [17] It takes the form of a differential allowance, and guarantees everyone over the age of 65, residing on French territory, a minimum standard of living close to the poverty line. It is drawn from the FSV, which was created in 1993 and funded by the contribution sociale généralisée (CSG) (general social contribution) and various taxes to pay for national solidarity benefits. The minimum vieillesse is funded by the state, but belongs to the framework of a social security system based on a contribution approach.  [18]

8The general question of the poverty and purchasing power of the poorest French citizens attracted considerable attention during the 2007 campaign and at the start of Sarkozy’s term.  [19] This prominence was visible in opinion polls, public debates, and activity by organizations like Alerte and the Fondation Abbé Pierre.  [20] In August 2007, Sarkozy used these preoccupations to justify adopting the fiscal measures in the TEPA Act (a law promoting work, employment, and purchasing power), including those meant to make it possible to “work more to earn more.” In this context, the material situation of the poorest pensioners received a certain amount of media coverage; several reports pointed out that the allocation de solidarité aux personnes âgés (ASPA) (Solidarity Benefit for the Elderly) had not kept up with the median income and the poverty line.  [21] These reports were systematically compiled and integrated into the strategic planning that took place at the first thematic workshop, on the “social question,” organized by Emmanuelle Mignon in preparation for creating the manifesto.  [22] As we shall see, unions and associations for the elderly also helped put the question on the agenda by lobbying the candidates.  [23]

9A consensus on increasing pensions developed quickly among the main presidential candidates.  [24] Ségolène Royal, the Parti Socialiste (PS) candidate, was the first to propose increasing all pensions which were below the salaire minimum de croissance (Smic) (minimum wage), including the minimum vieillesse; from 2007, she proposed an increase of 5%, having considered raising small pensions to the level of the Smic by the end of her term. François Bayrou, the UDF candidate, went one better and proposed an increase in the minimum vieillesse to a level equal to 90% of the Smic over five years. In his presidential program Mon projet: Ensemble, tout devient possible, of which eight million copies were printed at the end of March 2007, Sarkozy promised to raise the minimum vieillesse by 25%. The promise seems to have become obligatory for all the candidates—a scenario that has not been theorized by the literature on electoral pledges. Besides policy-, vote-, and office-related considerations,  [25] this example suggests that the fact of an issue being placed on the agenda by rival candidates can in itself constitute a strong incentive to make an electoral pledge.  [26] Does this consensus between candidates make the promise more likely to be kept?

Democratic mandate versus budgetary constraints

13The rise in the minimum vieillesse was likely to come into conflict with budgetary constraints, since it had not been incorporated in the overall costing of the presidential manifesto  [37] and the source of its funding had not been stated with any precision. Sarkozy announced he would use the proceeds from the reform of the special pension regimes, even though this would only produce financial savings over the long term, as Royal pointed out in a televised debate between the two rounds of voting.  [38] Once he was elected, Sarkozy remained vague about the means of financing the ASPA reform, but mentioned the FSV, which already financed the benefit.  [39] Despite this vagueness, there was no resistance either from the finance ministry or the civil service. Sarkozy’s use of powers of decree helped in this regard: the decision could be made quickly, without any explanation—particularly about how the measure would be funded—and without giving the opposition a platform to contest it.

14In any case, Sarkozy could justify himself, less than a year after he had been elected, by pointing to an explicit and electoral promise with precise figures: a 25% increase over five years. This precision left little room for debate, especially as he had made a point of his desire to be precise in his manifesto, to then be able to keep his word.  [40] To explain the absence of any need for compromise with ministries, the presidential advisers we interviewed stressed that a promise contained in a detailed manifesto presented to and approved by voters conferred substantial legitimacy; so did the fact that the decision was made by the president himself, rather than a minister. This was a point made by Éric Aubry, a member of the team responsible for the UMP manifesto in 2007, who served as social adviser for Fillon from 2007 to 2011:

15

It needs to be a political commitment, whether presidential or legislative. We could say: “There was a democratic debate. Our manifesto was chosen. The manifesto proposed this reform, we have the legitimacy to do it, we’re doing it.“  [41]

16These arguments were particularly strong because the initiative was undertaken soon after the election.  [42] The president still had considerable political capital, his authority over his entourage was undiminished, he had good relations with the prime minister, and the government was relatively unified. These “democratic” resources had an important effect on power relations with actors in the finance ministry and the civil service responsible for social security,  [43] and explain the president’s success in keeping the project from getting bogged down in inter-ministerial meetings and dedicated commissions, particularly those meant to establish how the measure would be funded.

17

For example, for the minimum vieillesse, we had the finance minister saying: “It’s not possible! Not now!” Then we have an inter-ministerial meeting to act on the president’s commitment. We’re discussing the schedule for the increases, but its already decided. We draw up a “blue paper,” in the jargon of the prime minister’s office, that is, a detailed list of decisions, and then there’ll be a first increase on such and such a date of such and such percent.  [44]

18The legitimacy conferred by the election gave the president significant resources for carrying out his manifesto, but room for maneuver remained limited. A common theme of our interviews was the realization of promises in a watered-down form:

19

There was soon a problem of economic reality, which got worse with the financial crisis. And there were lots of measures we didn’t carry out. For example, we didn’t do the “suburban development plan,” because there wasn’t enough money. We almost didn’t do the increase in small pensions because there wasn’t enough money. [...] It’s easy to carry out reforms when the money’s flowing. But when you don’t have the cash...  [45]

20When the financial crisis took place in autumn 2008, things changed steadily, and the logic of decision-making changed along with it.  [46] The main aim of Sarkozy’s presidency was no longer to keep his campaign promises, and he even wheeled back on some of them to focus instead on the urgent problems created by the crisis. Power relations between the president and the civil service also changed. During his campaign, Sarkozy took pride in not being an énarque, and claimed he would have the upper hand over the civil service and would not let his decisions be influenced by “technocrats.” More broadly, he suggested that political logic takes precedence over administrative logic, in the name of democracy and the mandate conferred by the citizens. But the financial crisis called this approach into question:

21

Nicolas Sarkozy was the victim of a shock that came from outside, which no one had foreseen, and which meant he lost his footing in terms of methods... In the midst of the 2008 crisis, you needed énarques. When you have technical negotiations with ministers or bankers that last right through the night, if you don’t go in armed with five énarques to the right of you and five énarques to the left of you, who’ve had files on every technical aspect of the issue going back ten years, you lose your footing. And so the crisis not only got in the way of his political project, but also his method. At that moment, after the 2008 crisis, there was a complete restoration of the énarchie—the highest level of the president’s cabinet included.  [47]

22This change in context considerably reinforced the influence of the senior finance ministry officials. They did not necessarily gain the ability to oppose the reforms, and had to accept those that had already been voted through. But Bercy used its power to oversee the execution of the budget through the use of memos. The ministry can establish a so-called “precautionary” reserve, which holds back part of the budget allocated for a given program. At the end of the fiscal year, the ministry then announces whether this funding will be “released” or simply cancelled.

23The relations between the presidency and senior civil servants also developed according to the logic of power relations in the strict sense. Elected officials and ministers come and go while senior civil servants remain,  [48] but this permanence cuts both ways: a civil servant’s career can be affected in the long term by a conflict with political representatives in power. In the period immediately after an election, technocrats tend to adopt a wait-and-see strategy. This was particularly true when faced with Sarkozy, who had proven himself determined to drive through certain decisions. The financial crisis of 2008, which coincided with a drop in the president’s approval ratings, changed the situation by severely reducing his chances of obtaining a second term.

24As the literature on the political economy of state budgets suggests, budgetary constraints are not a simple given, and the president can use room for maneuver to convert resources from his “democratic mandate.” This freedom diminishes over time, particularly during periods of poor economic conditions. It remains for us to determine the factors that explain why certain commitments were prioritized for funding.

Distributive policies and promises designed to target specific publics

26Various studies suggest that electoral pledges can be conceived as a form of exchange between, on the one hand, political parties and their candidates, and on the other, certain groups of voters. Studies of distributive policies show that elected representatives hoping to be re-elected  [49] tend to prioritize spending on certain groups, in order to benefit from the phenomenon (whether real or merely anticipated) of “voting with your wallet”: voters reward elected representatives who promise to do or have done something for them. Taking account of electoral factors may be a trait of elected representatives all over the world—particularly with regard to the social groups at the heart of their target base,  [50] or, conversely, those who seem most volatile.  [51] This approach leads elected representatives to concentrate spending on target groups, something James Q. Wilson describes as “client politics.”  [52] Following a similar line of thought, members of the US Congress try to get as many budgetary resources as possible allocated to their constituency.  [53] Such strategies can also be observed in the analysis of public action decisions in particular sectors. The literature on the welfare state calls into question the party-based model of spending, which assumes that spending is higher when a left-wing party is in power, and reveals more complex, fluid, and volatile dynamics in which political parties spend money by targeting the most strategic electoral groups.  [54] As a recent review of work on the effects of party politics on the welfare state observed:

27

The crucial insight that we can gain from the literature on party-electorate linkages is that—in contrast to the programmatic bias in partisan theory—the electoral strategies of parties may as well be particularistic. Parties sometimes adopt narrowly defined policies in order to attract specific groups of voters.  [55]

28In their most radical form, these transactions based on particularistic considerations can take the form of clientelist relations or party patronage.  [56] Without necessarily going this far, the elected representatives of many democratic countries practice non-contingent and indirect political exchanges.

29

They devise policy packages knowing that they are likely to benefit particular groups of voters (typically, a party’s swing voters) rather than others, and that this in turn will make it more likely in general that members of these groups will vote for the party.  [57]

30The existence and importance of “voting with your wallet” have been keenly debated for several decades, some researchers finding that it has a large effect  [58] and some arguing, to the contrary, that it is limited or insignificant.  [59] Having said that, the countless examples of tactical budgetary allocations documented in the literature, in countries as different as the United States, Sweden, Brazil, and Romania,  [60] suggest that David Mayhew’s observation from 1974 still holds:

31

How much particularized benefits count for at the polls is extraordinarily difficult to say. But it would be hard to find a congressman who thinks he can afford to wait around until precise information is available.  [61]

32The practice of targeting certain distributive policies is likely to have effects both on the formulation of electoral manifestos and on their fulfillment. Candidates have every reason to appeal not only to voters’ retrospectives concerns, but also, by promising specific advantages, to their prospective expectations.  [62] A Swedish study even suggests that voters put greater weight on the policies promised to them than on those carried out by the outgoing government.  [63] Distributive measures are an essential component of electoral manifestos. Whether trying to attract strategic segments of the electorate, or to provide a compensation for difficult reforms, social policy proposals aim to make the manifesto more attractive.

33

In an electoral manifesto, let’s be honest, there’s always something hard to swallow—the special regimes reform—but there’s also a sweetener. The sweetener, in this case, was the commitment to increase the minimum vieillesse by 25% [...]. And there were other commitments too, but relatively minor ones, like increasing pensions for widows and widowers.  [64]

34We are then faced with a new question: what factors explain which groups are more likely to be targeted by electoral pledges and prioritized when governments make compromises? We can form hypotheses based once again on the literature on distributive policies, which discuss two main types of characteristics: the resources the group can mobilize and the way it is perceived within society.  [65]

35The first dimension involves the resources a group possesses to have its demands heard by the government. This obviously involves both material and organizational resources  [66]: faced with representatives with limited cognitive and material resources, the most powerful interest groups are more likely to see their demands fulfilled. But it is also, broadly, a question of the group’s ability and inclination to mobilize, notably by voting. Paul S. Martin recalls Valdimar O. Key’s observation that “the blunt truth is that politicians and officials are under no compulsion to pay much heed to classes and groups of citizens that do not vote,” and suggests that distributive policies prioritize “active, hence attentive, publics,” since these are most likely to express their gratitude by voting.  [67] He supports this hypothesis by showing that US federal funds are concentrated on districts where the voting rate is the highest, whereas the greatest need is often found elsewhere.  [68] The lower degree of mobilization of the most disadvantaged social classes therefore limits incentives to fund anti-poverty programs. Although the literature agrees that levels of participation by groups affects the advantages they are likely to obtain, there is debate between those who believe the parties in government favor their own electorate in their use of distributive policies,  [69] and those who argue that the segments of the electorate containing swing voters are targeted as a priority.  [70]

36In a widely discussed article, Anne Schneider and Helen Ingram argue that the ability of different social groups to make themselves heard by the government depends not only on their resources, but also on their social construction—“the cultural characterizations or popular images of the persons or groups whose behavior and well-being are affected by public policy. These characterizations are normative and evaluative, presenting groups in positive or negative terms through symbolic language, metaphors, and stories.”  [71] These come into play when elected representatives develop re-election strategies which anticipate the reaction of the target group (and other groups) to a particular measure that might benefit or harm them. Elected representative tend to target distributive policies at groups that are both powerful and perceived positively by society, whereas “dependent” groups, who elicit sympathy but do not mobilize in their own right, are instead the object of symbolic measures that enable an appearance of concern, but without having to allocate resources.  [72]

The strategic nature of the public of pensioners

37These hypotheses illuminate the start of Sarkozy’s presidency. Many distributive measures benefited parts of the electorate that could be considered strategically important: this was the case for the “tax shield” and inheritance tax exemption for the wealthiest households, the reimbursement to expats of taxation earmarked for school budgets, reductions in tax and social security contributions for entrepreneurs, and the reduction in VAT for restaurants. Although it was aimed at a less privileged part of the population, the increase in the minimum vieillesse followed the same logic, targeting a population that possessed several of the characteristics of the groups likely to benefit in theory from distributive policies: a high degree of mobilization, a body of swing voters, and, in the case of the poorest pensioners, positive social perception.

38Pensioners also made up a mobilized social group in the broad sense. Those on the minimum vieillesse were living in poverty; a collection of associations and pensioners’ unions supported putting this problem on the agenda. In 2000, to make up for a low level of pensioner mobilization, lack of public funding for their representative bodies, and preoccupation among unions with the problems of the working population, a number of associations and pensioners’ branches of unions created the Confédération française des retraités (CFR) (French Confederation of Pensioners). At each presidential election, representatives of the CFR meet members of each candidate’s team to submit their proposals. In 2007, two representatives—Jean Catherine, deputy chairman of the CFR, and Sylvain Denis, president of the Fédération nationale des associations de retraités (FNAR) (National Federation of Pensioners’ Associations)—had a long meeting with Fillon, who was open to their demand for an increase in small pensions.  [73]

39Once the elections were over, in accordance with the agreements reached in these meetings, these actors stayed in contact with several ministerial cabinets, especially those of Fillon, Xavier Bertrand (labor), and Roselyne Bachelot (health), while trying to attract media attention to the problems of pensioner poverty. This mobilization resonated with the alerts produced by the COR, whose report, submitted to the prime minister at the end of 2007, included a chapter on standards of living among pensioners, especially those receiving the minimum vieillesse,  [74] and by the CNAV, whose president called for an increase in the ASPA, urging that “Nicolas Sarkozy should not just be satisfied with the effect caused by an announcement before the municipal elections.”  [75]

40The pressure on Sarkozy’s campaign team was maintained throughout his term. In a joint announcement three days before the first round of the municipal elections, six pensioners’ groups called for a national day of action to demand “an immediate increase” in pensions, with a large demonstration in Paris and 70 other events. On this occasion, a delegation of representatives from unions was received at Matignon by an adviser to the prime minister.  [76] Several such meetings with unions and pensioners’ associations had already taken place, as the prime minister’s social adviser explains: “At regular intervals, they would protest: ‘Commitments were made, where have we got to?’”  [77] These organizations not only lobbied the executive branch, but also influenced parliamentarians by mobilizing considerable resources. Beginning with the 2007 legislative elections, for instance, members of the CFR approached all candidates. After the vote, this costly strategy was altered to target those deputies who were most influential, who were familiar with pensioner issues, who had supported their interests during the campaign, and who were likely to be mobilized on these issues.  [78]

41Although the unions and pensioners’ associations are far from the most powerful interest groups, the period in question was marked by intense activism on their part—particularly influential since pensioners have a far higher level of voter turnout than average.  [79] The literature increasingly uses the concept of “grey power” to account for the over-representation of older voters, electoral strategies which target them, and the distributive policies this leads to.  [80] Furthermore, they make up the center of the target electorate for the right in general,  [81] and for Sarkozy in particular: at the second round of the 2007 presidential election, more than 70% of voters over sixty voted for him, making up more than a third of his total vote.  [82]

42Pensioners correspond well to what Martin describes as an “active, hence attentive, public.” Their importance to Sarkozy’s support soon weakened the president. The deterioration of his popularity, after the honeymoon period of the summer of 2007, was particularly rapid among older voters. For example, Le Monde remarked on 22 January 2008:

43

The crisis is even clearer among older people, even though they were a foundation of the president’s base. He has lost ten points over two months in this age category, according to the latest Ifop survey for Le Journal du dimanche, January 20. And the survey of TNS-Sofres for Le Parisien makes the figure as high as eighteen points. The president wants to try winning this population back over, which amounts to 13 million pensioners.  [83]

44Older voters seem to have been disappointed by the portrayal of the president’s private life, his plans to change the conditions of exemption from television license fees, the adoption of excess payments for medical insurance, and the lack of measures aimed at pensioners—who were not particularly affected by the loan interest repayment policy—and by measures meant to make it possible to “work more to earn more.”

45The minimum vieillesse favors the poorest pensioners, whose material situation was widely considered intolerable—“shameful,” according to Bayrou.  [84] While Anne Schneider and Helen Ingram have pointed out that the target group of the “poor,” taken as a whole, can be subject to negative social constructions—presenting them, for example, as “lazy persons who are benefitting from other people’s hard work”  [85]—the recipients of the minimum vieillesse escape this negative representation because they are retired.

46Furthermore, there was clever maneuvering regarding the publics who were, on the one hand, genuinely affected by the reform and, on the other hand, those who only seemed to be affected by it. The president and government presented the increase as a gesture benefiting those receiving “small pensions”; this was the only response given to parliamentary questions about low levels of pension income. More broadly, it was presented as a solution to the decline in purchasing power of pensioners. In fact, however, it concerned only a small part of this population and, as we will see, even excluded couples receiving the minimum vieillesse. Despite opposition by pensioners’ unions  [86] and associations to the exclusion of “3.5 million small pensions,”  [87] the executive deliberately maintained the confusion: the cost of the reform was reduced by basing it on a limited number of people, but its electoral benefits were maximized by an announcement which spoke in terms of small pensions as a whole.  [88] Appreciating these maneuvers with target groups is essential to understanding a decision that, in itself, presented an unappealing cost-to-benefit ratio.

47Finally, the revaluing of the ASPA received wide approval among the population. This consensus was shown by various polls in 2008, which undoubtedly influenced Sarkozy’s decision: according to a survey by TNS Sofres published by Le Parisien on 5 February 2008, 74% of respondents considered pensioners’ standard of living unsatisfactory; in December, another opinion poll carried out by OpinionWay ranked the increase in the minimum vieillesse as the most widely approved measure, with 94% of respondents in favor.  [89]

The shifting importance of target populations depending on the context of political competition

48Research into the electoral considerations involved in targeting distributive policies suggests that the most “contested” electoral segments are prioritized. As we have seen, this depends on the electoral behavior of these groups—their mobilization and their propensity to change their party support from one election to the next—but also on positions adopted by other parties (or candidates) and the degree of political competition around the issues important to these groups. Political parties are incentivized to concern themselves with electoral clienteles to whom rival appeals are likely to be made.  [90] Notably, this explains the positive correlation between the proportion of targeted expenditure and the group’s electoral volatility,  [91] and why US congressmen are particularly active in “pork barrel politics” when their re-election hangs in the balance.  [92]

49In the case at hand, the context of political competition is also likely to have played a role. The anticipation of a second round—which from early on had seemed likely to involve a head-to-head contest with the Socialist Party candidate—led Sarkozy to adopt a strategy of openness towards certain themes, ideas, or personalities from the left.  [93] As several members of his team explained to us, the aim was both to balance out his liberal, authoritarian image, and to avoid being boxed into certain party positions he did not identify with. He asked his team to work “without taboo,” to “open the door,” and, alongside Mignon, Research Director for the UMP and responsible for drawing up his manifesto, he developed a method of working based on discussion with experts, many of them from the left. According to Mignon, a major advantage for Sarkozy’s campaign was this ability to “cherry pick” from among the “fairly pluralist” proposals of his team, which included liberals and social Gaullists, those more concerned with issues of sovereignty and those more concerned with values: “Each time a proposal allowed him to add some balance to his manifesto on the left, he adopted it.”  [94]

50Several other analysts of electoral competition even describe strategies of “preempting” or “appropriating” issues from the left.  [95] As it happened, it was a matter of contesting the left’s dominance of important themes such as purchasing power and the fight against poverty.  [96] Many witnesses confirm this strategy of appropriation. In his account of the campaign, Georges-Marc Benamou, for example, describes Sarkozy’s remarks during a meeting:

51

“You’ll see... That’s not all... I’m going to surprise you... We’re going to pinch everything from the left. They’ll have nothing left. [...] It’s simple: we only need to take on words, their words, to preempt them in a way.” He lets this sink in and starts again, with an excited look in his eye: “You rob them, and suddenly, you’ve got them. Bam, you leap on them, and you’ve got them.“  [97]

52Those who wrote Nicolas Sarkozy’s social manifesto also make a connection between the increase in the ASPA and the promise to increase the RSA:

53

That was the idea: there should be a boost for the most disadvantaged, the left behind. With a concern for social balance in the presidential manifesto. [...] The idea that there should be a policy of fighting against social inequalities: the left doesn’t have a monopoly over that! The fight against social inequalities is an inclination and a policy that is also shared with the right [...]. After all, there are a certain number of points in the social field that are really a matter of consensus.  [98]
You’ll also notice that, in the Fifth Republic, a left-wing president sometimes has to deal with the economy, business—that was true for Mitterand, Hollande—and a right-wing president has to deal with the social question and consider that, when we’re going through a crisis, we have to keep a few sticking plasters on, and maybe add new ones.  [99]

54Once Sarkozy was elected, the desire to integrate attitudes from the left was manifested in the government’s “openness” to personalities from the left, and more visibly in the choice of advisers at the Élysée, who were recruited from far beyond the president’s immediate circle—among the supporters of Chirac, the “social Gaullists,” and even Royal’s team.  [100] The formulation and realization of the commitment to increase small pensions belonged to this same logic, which allowed Sarkozy to compete with the left on its own territory.

The timeframe of decision-making

56A rich tradition of research has developed around the question of political and electoral cycles, starting from the hypothesis that government actors tend to order their fiscal reforms and expenditure in relation to electoral deadlines.  [101] To the extent that citizens seem to make up their minds based on the most recent political information,  [102] it would be rational for the executive to take unpopular measures at the start of their term, and adopt targeted distributive measures as electoral deadlines approach.

57This is precisely what we observe in the case in question, since the announcement occurred a month before the municipal elections of 2008. This chronology, and the speed of the decision, are all the more remarkable as the promise had barely been mentioned in political and media speeches after Sarkozy’s election. During summer 2007, several deputies from both the opposition  [103] and the majority  [104] had pressed the government on the details, but they received no response before the decisions of 2008. However, on 11 September 2007, the president repeated his commitment to increase small pensions.  [105]

58The chart in Figure 1, produced from close analysis of the press and the archives of the National Assembly, shows that it was only from the start of 2008, when advisers to the president and prime minister became aware of the rapid decline in their popularity rating and began to fear they would be punished by voters in the municipal elections in March, that references to the increase in the ASPA became more common. On 3 January, Claude Guéant, General Secretary of the Élysée, referred to “small pensions” as a “very important concern of the president,” which would be addressed by the government “before the summer.”  [106] This was confirmed by Bertrand, who announced “the delivery of Nicolas Sarkozy’s campaign promises,” including the increase of the minimum vieillesse by 25% “over his term.”  [107] On 22 January, Le Monde revealed that the government intended to accelerate the increase; although the plan had been to wait until after the negotiation of the reform of the general pension system, the newspaper stated that the president intended to act “before the municipal elections.”

Figure 1. Nicolas Sarkozy’s popularity rating and the prominence of the minimum vieillesse in the media and parliamentary sphere

Figure 1. Nicolas Sarkozy’s popularity rating and the prominence of the minimum vieillesse in the media and parliamentary sphere

59As the municipal elections approached, pensioners’ associations intensified their lobbying of deputies from the ruling party, many of whom were up for election. Some of the most influential UMP officials called for an acceleration of the policy and applied pressure by asking the government questions  [108] and making public statements.  [109] More informally, they met with ministers and their cabinets.

60

Of course there is pressure on parliamentarians, which takes different forms... Spoken questions, written questions, direct interventions, questions in group meetings, in the Assembly. It was the same for the increase in the AAH, the allocation adulte handicapé (Adult Disability Benefit). So, at the same time, the associations and unions mobilize and speak directly or indirectly to their elected representatives, and then also to different political groups, in Paris... So of course, there’s pressure on parliamentarians, most of them saying (he knocks on the table): “People are asking me: where has it got to, this commitment? What are you doing?“  [110]

61But the president’s advisers were divided, and did not believe the measure would be enough to reverse the trend in time for the municipal elections. They argued that it would be better to keep it in reserve for negotiations over extending the contribution period.  [111] They eventually opted for a rapid decision: on 3 February, Guéant announced that the increase in the minimum vieillesse would only take “a matter of weeks,”  [112] but was criticized by several ministers who considered the announcement premature.  [113] Two days later, during a session for questions in the Assembly, Fillon and Bertrand confirmed that they would deliver on the promise to raise the minimum vieillesse by 25% over five years.  [114] On 6 February, in a meeting with unions to discuss social security reforms, the president himself finally confirmed this increase, starting with 5% in 2008.  [115]

The improvised nature of the reform

62The interviews that we carried out with advisers to Sarkozy and Fillon confirm the impression that the execution of some commitments was not planned out, but instead decided more tactically, in response to the shifting political context. Although the presidential team quickly put certain priorities on the agenda, other issues were not approached methodically or associated with a specific timetable. Actions were decided “on a day-to-day basis, in response to events”: televised speeches by Sarkozy, external shocks (from the economy or foreign politics), and elections. Advisers “mined” the manifesto to come up with the characteristic “announcements” the president liked to make in speeches. This search did not follow a particular logic, and was often done in a state of emergency, as Mignon explains:

63

My greatest fear was when there were television programs: either a speech to the press or a television appearance. [...] They would tell us at the last minute, [...] and we had to prepare what we called “announcements.” Everyone would go to look in their own basket to see what they could come up with for an “announcement.“  [116]

64This absence of a clear ordering of tasks over the term of office, and the need to work in a state of urgency in the lead-up to media or electoral meetings, casts light on the hesitant, rushed decision to raise the minimum vieillesse. By creating incentives to act in the short term, intermediate elections create opportunities to remind elected representatives of their responsibilities:

65

Shrewdness wasn’t necessarily incompatible with conviction, you can genuinely want to carry out a measure, and also try to do it at the right time with regard to elections.  [117]
But it’s also very democratic. You’ve made commitments. There’s an intermediate election. Obviously everyone looks at the list of measures that you produced previously: “Where have we got to? What’s wrong with them, how has that happened?” You could say it’s a booster shot.  [118]

Increasing visibility with a check

66Another indication of the strategic nature of the increase in the ASPA lies in the way it was carried out. During the announcement, Sarkozy specified that those eligible would receive 200 euros, in the form of a check, from April 2008, as an “advance” on the increase to come. This measure, which was made official by decree on 9 March 2008, and which meant the increase could be paid before the end of the month,  [119] allowed him to maximize the desired effect of this decision in advance. The measure was presented by the media as an “electoral gift,” aimed at restoring the popularity of the president and his party before the municipal elections, and the opposition was quick to criticize its use for electoral ends.  [120] The measure was not unprecedented. In 2004, Jean-Pierre Raffarin’s government had granted a bonus of 70 euros to those with small pensions;  [121] more recently, in 2014, a bonus of 40 euros was paid by Manuel Valls’ government to all pensioners who received less than 1,200 euros each month.  [122] More generally, studies of distributive strategies show that they are conceived with visibility in mind, more than effectiveness.  [123] Rather than investing in public services that generate widespread benefits, such as education, governments prefer to establish more targeted and tangible projects, like infrastructure projects or checks that help sustain consumption.  [124] For example, they tend to prioritize funding for managing the consequences of natural disasters over preventative measures, since the latter will often remain invisible to voters.  [125]

67The indecision about how to deliver the check for 200 euros was also indicative of its improvised nature. During the initial consultation with the pension funds that would have to carry out the measure, Danièle Karniewicz, president of the CNAV, condemned the lack of precision in the decree about the nature of the check and how it would be financed.  [126] The council of the CNAV came to a favorable opinion after it was confirmed that the payment would be a bonus (and not an advance), which would not need to be recuperated, but which would still be taken into account in the timetable for the increase, and that the cost of this measure, estimated to be 120 million euros, would not be borne by the funds of the CNAV.  [127] The bonus would actually be paid to all recipients of the minimum vieillesse, including 105,000 recipients living in couples who would therefore be excluded from the increase carried out in the 2009 social security budget. The payment would remain with the recipients, since there was no practical way of taking it back.  [128]

An illusory reform

68The processes of political mobilization reflect the characteristics of the groups with an interest in the advantages delivered by the measure, but also of those who would shoulder the cost. For instance, Wilson argues that a measure with concentrated costs and dispersed advantages would be more likely to face mobilization by hostile interests than a measure with concentrated advantages but dispersed costs. From this perspective, social policies which benefit precise groups but are funded collectively would be favored, since the beneficiaries mobilize but those who fund the measures do not.  [129] Contrastingly, Walter Korpi and Joakim Palme claim that social policies targeted at a specific part of the population tend to be less socially acceptable than universal benefits.  [130] Studies have shown that in France, as in other European countries (Germany, the United Kingdom, Italy), the social acceptability of payments to pensioners increases with their age.  [131] In a report drafted in preparation for the 2008 “meeting” on pensions, the COR advised against an immediate increase in the minimum vieillesse of 25%, arguing that the gap between this pension and those of employees who had spent a career working at the minimum wage should “remain at a significant level if we want to keep valuing work.”  [132]

69Although these studies arrive at conflicting conclusions, they lead us to consider not just the aims of a reform, but also the reactions of those who finance it. In the case in question, the measure met with no political or administrative resistance. Beyond the “positive” social construction of the poorest pensioners as a group and a consensus about the urgent need to improve their material situation, our investigation showed that the reform could be financed easily because it cost less than expected and because the executive used a very opaque legal and budgetary framework.

70Although the promise was perceived as a significant gesture for small pensions as a whole, the impact of the actual increase was, as we have seen, much more modest. The decision not to apply the measure to couples, a compromise to the finance ministry with the aim of limiting the “cost to the public finances” and “buying some time,”  [133] was strongly criticized by deputies. It was defended by Bertrand,  [134] who nonetheless agreed to form a working group to consider the question. In the long term, however, apart from the check for 200 euros, which they kept, the minimum vieillesse for married couples only received the required, inflation-indexed increases.  [135] More broadly, the question of increasing the benefits generally linked to the minimum vieillesse (the supplementary disability allowance, the allowance for disabled adults, work accident annuities) was completely obscured.  [136]

71Besides, the increase in the minimum vieillesse was apparently painless. The reform was recorded on the budget of the FSV, and not on the public budget sheet. The president of the CNAV and unions immediately expressed doubts about this approach.  [137] Danièle Karniewicz noted that the FSV was already indebted to the CNAV: “Will the FSV increase this debt, or will it pay something into the pension schemes, and if so when?”  [138] A recent Senate briefing  [139] revealed that the FSV never received additional revenue to offset these expenses, which simply worsened its deficit.  [140] The briefing specifically condemned the bonuses granted by decree without any provisions for funding.  [141] It described the FSV as a “catch-all” fund which has borne the cost of various reforms without any direct connection to its original purpose. In another report, senators noted that the FSV presented serious legal risks and a considerable deficit, and that its resources were very insecure, since they had been changed eighteen times over twenty-three budgets.  [142] The specific legal framework for social security funding involved significant room for maneuver, which made it possible to avoid the question of the cost of a measure which, in any case, was popular:

72

The president’s campaign promise to increase the ASPA by 25% was kept, but funded from the social security budget rather than the state budget. The division of tasks meant that I had no direct responsibility for this decision-making related to social security. Personally, I regret the existence of these complex relations between state and social security, established decades ago, which mean that decisions about social security expenditure and state expenditure are taken separately.  [143]

73The absence of funding for the increase in the minimum vieillesse was not an isolated case, but part of a major trend in public policy in the last few decades. For ministries, the budget constraints are real: there are many mechanisms of control which allow the finance ministry to ensure that budget limits are respected. But equivalent rigor is not applied to social security spending.  [144]

74Although the decision to revise the ASPA upwards did not encounter any opposition, this was not only because it benefited the poorest pensioners, who were broadly viewed sympathetically, but because its cost was not borne by any particular group, or even by the annual state budget. Such a measure had a negative impact only collectively and in the long term, by increasing the deficit of a fund, with unforeseeable consequences. This is the exact opposite of policies that draw on a specific tax, as was the case of the minimum vieillesse when it was created in 1956, which was financed by taxing car use.  [145]

75The example of the minimum vieillesse shows that the criteria normally used to explain the fulfillment of an electoral pledge—the absence of major expense, benefits for as many people as possible, media coverage of the issue, and its centrality to the manifesto and ideology of the candidate and their party—are not sufficient for every situation. The aim of this article has been to use analytic frameworks and approaches from other literatures to understand the conditions in which this promise was kept, in order to reveal supplementary factors that affect the exercise of the president’s responsibility towards citizens.

76Our observations confirm that, within the framework of the literature on electoral pledges, it is useful to reconsider the significance of budgetary constraints—which, far from acting in an invariable way, undergo (re)definition depending on the issue and how much time is left in office. It becomes possible to create room for maneuver to deliver the presidential project, especially by converting resources of legitimation drawn from the “democratic mandate” and by inscribing the decision in a flexible legal framework. But it becomes increasingly difficult to finance the expenses which arise as the term in office progresses, and when economic conditions are poor.

77The question of the incentives to keep one’s word is therefore particularly relevant with regard to high-cost electoral commitments. Our case study leads us to include the characteristics of the publics targeted by these commitments when analyzing these incentives. This parameter is central to distributive public policy analysis, but has not been included in the analysis of the fulfillment of manifestos, even though it is particularly relevant for taking into account the delivery of promises involving an expense. Based on our analysis of the increase in the ASPA, we propose that the delivery of a promise is more likely when the group that benefits from it is viewed positively in society and possesses significant resources for mobilization: these resources include interest groups, networks of influence, and above all electoral mobilization. As it happens, unions and pensioners’ associations managed to increase the visibility of this promise using numerous operations “from above,” through meetings at the Élysée and in ministerial cabinets, and “from below,” by mobilizing elected representatives from the ruling party and opinion leaders. The large numbers of pensioners in Sarkozy’s base meant the group received heavy attention. The increase in the minimum vieillesse allowed the president to improve his reputation with pensioners without facing significant opposition from other social groups.

78Decision-makers involved in the policy made sure to maximize the perception of the benefits of the reform through the use of a check for 200 euros and a communication strategy that maintained ambiguity about the people who would benefit, in order to let people believe that it was a gesture towards all recipients of “small pensions.” There was some very clever maneuvering around target groups, using a framework that allowed Sarkozy both to adopt a widely supported reform, given that the poorest pensioners evoked a high degree of sympathy, and to speak as if the policy benefited everyone who received “small pensions”—or even all pensioners. At the same time, it was difficult to ascertain its cost. Room for interpretation in what had been promised resulted in the promise being fulfilled to the minimum possible extent, and this likely explains the gap between the significant impact manifestos have on public action, and the dominant perception of them as a series of promises that only involve “those who want to hear them.” These observations invite us to integrate the way in which promises are constructed, invested, and redefined over time into our analysis of the electoral pledges. This flexibility is still difficult to address using large databases, which have to express what has been promised in specific terms in order to evaluate whether it has been delivered—which, we have suggested, is likely a major reason that the literature on electoral pledges fails to include certain factors traditionally considered in public policy analysis.

79We have also observed that strategic considerations about target populations are particularly important when there is strong political competition over the issue at hand. The commitments fulfilled are not limited to the key themes of the president and their party; they may also be related to the favored issues of their political opponents, following a logic of competition that is essential to understanding Sarkozy’s actions as a candidate and while in office. Keeping a widely supported promise limits the risk of resistance, while enabling competition with the other parties on their own territory.

80Two other contextual elements affect how important the clienteles targeted by electoral promises are. On one hand, crises of popularity can push the executive to fund distributive measures to win back unhappy voters. On the other hand, major upcoming intermediate elections increase strategic considerations, as they are viewed as a way to measure support for the executive and its policies. Very often, the literature on electoral cycles only takes single elections into consideration, analyzing the electoral incentives for an actor or category of actors over the term of office involved: in this view, such incentives are weak at the beginning, growing as the electoral deadline approaches—particularly if they want to be re-elected. As the present case shows, this rather simplistic model is upset by the occurrence of other elections, which act contra-cyclically and incentivize elected representatives to keep sight of their electoral commitments throughout their term. These intermediate votes become electoral tests when the executive branch suffers a popularity crisis, and when they are used as opportunities by certain social groups. They may then contribute to the realization of electoral promises that are, in themselves, unlikely to be kept, notably when they target strategic clienteles and when opposition parties adopt the same policies, thereby making them obligatory in the campaign. Sufficiently strong electoral incentives can provide motivation for funding a promised measure, even when costly. The increase in the minimum vieillesse is not a unique case: it was part of a series of reforms by which the executive branch tried to satisfy specific, mobilized, influential clienteles, with an eye on a difficult electoral calendar. Overall, candidates are incentivized to make promises aimed at these types of publics, and possess a certain capacity to keep them.

81Curiously, however, the increase in the minimum vieillesse was barely mentioned in discussions about the results of Sarkozy’s time in office, or in his 2012 campaign. More generally, the latter was based on new proposals rather than on his record in office, following the principle that it is the manifesto that wins an election. In hindsight, certain people from Sarkozy’s circle felt this lack of messaging about his record was a strategic error, which gave the impression that nothing had been accomplished.  [146] But pensioners, especially the oldest, voted for him in large numbers in May 2012, continuing to be one of the socio-professional categories most favorable to the incumbent president.  [147]

82From a methodological point of view, this study encourages us to supplement quantitative analyses of this subject with case studies. Once again, this case study does not aim to invalidate theories produced by quantitative approaches which use large samples to analyze promise-keeping. Rather, the aim is to supplement the explanations they offer. We do not believe the decision to raise the minimum vieillesse was an unlikely one in absolute terms; rather, the literature on the fulfillment of electoral pledges did not offer any satisfactory explanation for this case. It does not take account of the particular status of promises that are a matter of consensus among the candidates in an election, pays little attention to the nature of the policy’s target populations, and tends to overestimate institutional capacities and budget constraints, or at least to overlook the changes that the latter may undergo. By closely examining a deviant case which cannot be fully explained by the existing models, we have proposed new hypotheses which merit systematic testing. Our results led us to return to the coding of French electoral promises following a similar methodology to that of the CPPG, with the aim of supplementing, where possible, the classic variables with additional ones relating to the groups targeted by the promises (both positively and negatively), the timing of their fulfillment, if it occurs, and the cycles of popularity ratings of the executive.

83Qualitative methodologies allow us to avoid thinking of promises as invariant—either made or not, kept or not—and to reconstruct more clearly how candidates construct and commit to them. They make it possible to avoid reducing the cost of promises to a fixed figure, and to observe the various ways this figure can be manipulated, depending on how the measure’s target population is defined and how it is funded. Overall, it is important to provide the whole genealogy of a promise, from the moment it is put on a candidate’s agenda to the moment of its realization, with all the adaptations, representations, abandonments, and amendments this entails. These factors occupy a central place in the sociology of public action; integrating them into a systematic study of electoral pledges turns out to be difficult, as the databases used for this purpose are made up of items that are assumed to be invariant. This methodological limitation justifies the combined use of different approaches and methods which allow us to explain a larger number of cases.  [148]