The money illusion and democratic accountability: the democratic stakes of indexing government benefits

Abstract

The decision whether to index government benefits can have significant economic and political implications. It can affect whether or to what extent benefits maintain their real value over time, affect the policy levers available to fight inflation, and shape discretionary budget priorities. Most of the attention in the literature has focused on understanding the economic pros and cons of indexing and the politics and political use of indexation in the context of welfare state reform and retrenchment. Less attention has been paid to what indexation means for democratic accountability. This paper seeks to rectify this by investigating the democratic stakes of indexing government benefits. It argues that there are, other things being equal, strong democratic reasons to index government benefits in a way (or according to metric) that preserves their publicly articulated purpose. However, concerns about lack of discretion and ownership suggest indexation rules should be designed to provide governments with some discretionary power over the size and perhaps timing of automatic adjustments—though accompanied by requirements that the exercise of this discretionary power be justified publicly.

Introduction

The decision whether to index government benefits—whether to automatically adjust benefit levels in line with changes to “a consumer price index (CPI), a wage value index, or a combination of both” (Fernandez, 2012, 242)—may seem technical, but it can have significant economic and political implications. It can affect whether or to what extent benefits maintain their real value over time, affect the policy levers available to fight inflation, and shape discretionary budget priorities. This is perhaps made most apparent during times of high inflation, when inflation threatens to hollow out the purchasing power of those who depend on government benefits to afford their basic necessities and governments struggling to bring inflation under control—as the last few years have demonstrated (Pallotti et al 2023). Yet the stakes, at least from a democratic perspective, can be equally high during times of low inflation. From a political standpoint, whether a benefit is indexed affects the “choice architecture” (Thaler et al. 2013) surrounding the policy. If a benefit is not indexed, the default outcome—i.e. what happens if no decision is taken that changes things—is for the real value of the benefit to gradually diminish (assuming there is at least some inflation and/or real wage growth). Whether the default outcome diminishes or preserves the real value of a benefit thus has significant political implications, particularly since it affects the power and relative bargaining position of potential veto players in the political system.

Lots of policy decisions have significant implications, of course. What makes the indexing decision particularly interesting, from a democratic standpoint, are two things. First, the tendency of people to “think more in nominal monetary values than in real ones” (Chytilova 2020, 4)—i.e. what economists refer to as “the money illusion”. With regards to indexing decisions, the money illusion can enable policymakers to engage in a sort of political deception, claiming credit or avoiding blame in ways that seem inconsistent with democratic norms regarding transparency and accountability. Second, indexing decisions raise questions regarding the scope of the discretion and flexibility that should be afforded to democratically elected policymakers. Simply put, more indexing means less discretion and flexibility for elected policymakers. Does this mean indexing is problematic from a democratic perspective? The answer is far from clear.

This paper examines the democratic stakes of the decision whether to index government benefits. It identifies and considers the implications of five democratic problems, or concerns implicated in the decision—deception, neglect, lack of discretion, ownership, and disappearance from the political agenda—and argues that there are, other things being equal, strong democratic reasons to index government benefits in a way (or according to a metric) that preserves their publicly articulated purpose (e.g. to provide the recipients with the ability to afford some basic necessities, or to ensure recipients do not fall into relative poverty). At the same time, it further argues, concerns about lack of discretion and ownership suggest indexation rules should be designed to provide governments with some discretionary power over the size and perhaps timing of automatic adjustments—though accompanied by requirements that the exercise of this discretionary power be justified publicly.

The existing literature on indexation has considered the policy related and economic pros and cons of indexing pensions (Biggs et al. 2005; Whitehouse 2009), the politics surrounding indexation (Fernandez, 2012; Weaver 1988), and the role of indexing in welfare state retrenchment (Green-Pedersen et al. 2012; Pierson 1996). Though the implications and pros and cons of indexing from a democratic perspective have animated and to some extent been explored in this literature (see particularly: Green-Pedersen et al 2012; Pierson 1996; Weaver, 1988), the democratic stakes of indexation have remained undertheorized. By identifying and fleshing out the democratic stakes and implications of the decision whether and how to index government benefits, this paper aims to contribute to a better understanding of a neglected but nevertheless relevant area of consideration when deciding whether and how to index government benefits.

Though there are democratic stakes in the decisions whether and how to index all kinds of government expenditures and revenue related tax brackets, income ceilings and other things, this paper is only concerned with sketching the democratic stakes and implications of indexing/not-indexing government provided social or welfare benefits—by which is meant benefits that are paid out, in cash, to individuals or households to support them financially. The type of benefits that the analysis is meant to apply to are things like pensions, unemployment insurance, minimum income guarantees, child benefits, and the like.¹ The democratic stakes of indexing other expenditure and revenue related programs and items would to some extent overlap with what is sketched below regarding government benefits, but they would not be identical, and so deserve to be analysed and treated separately.

The politics and economics of indexing

It is worth beginning by clarifying what indexing means and how indexation rules can vary. A government benefit or output is indexed if its nominal value is automatically adjusted to track changes to prices, wages, or some combination of both (Fernandez, 2012; Biggs et al. 2005; Weaver 1988; Whitehouse 2009). Automatic here does not necessarily mean without the government having any discretionary power to affect what happens. The government may be empowered to delay or stop automatic adjustments, or to propose an alternative adjustment (see Weaver 1988). But it does mean that the default outcome is shifted. Without indexation, the default outcome—i.e. what happens if no action is taken—is for no adjustment to be made (up or down), whereas with indexation, the default outcome is for an adjustment to be made, in accordance with the indexation rules (i.e. for instance, for the nominal value of a benefit to be adjusted upward to account for changes to a consumer price index).

The choice of what index to track—whether to track prices or wages, or even the choice between different measures of changes to prices—can have significant long-term effects on the real and comparative value of a benefit, as well as, of course, the size of the government’s expenditures (Biggs et al. 2005; Whitehouse 2009). Since wages tend to rise more than inflation (at least over long periods), the choice to let pensions track prices rather than wages, for instance, while it will help preserve the real value (i.e. the purchasing power) of the pension benefit, it will also mean that the pension income is reduced (gradually, over time) compared to the incomes enjoyed by the working-age population (Whitehouse 2009). Indexing rules can thus affect whether benefit programs remain adequate and well-tailored to the problems they are meant to solve or alleviate, in light of changing circumstances. If the indexing rules of a minimum income program do not adequately track the changes in prices faced by people near poverty, for instance, the adequacy of the benefit may be eroded. Indexing rules can also affect other parameters often relevant to the design of benefit programs, such as its coverage and the impact of the benefit on work incentives (Figari et al. 2013; Sutherland et al 2008; Vandelannoote and Verbist 2020). The design of indexing rules and the choice of what index to track is thus going to reflect a particular set of policy priorities –related, among other things, to the purpose of the benefit in question (whether the aim, primarily, is/should be to protect the recipients’ purchasing power or their relative economic standing) and regarding fiscal spending priorities and concerns.²

No doubt partly for these reasons, indexation rules—even among benefits that are indexed—vary between different countries. Sometimes they track changes in prices, sometimes changes in wages, sometimes whichever is higher (reflecting a desire, perhaps, to protect both the purchasing power and the relative economic standing of the recipients). In some cases, the indexing is not “full”—i.e. a benefit might be set to track changes to average wages or prices minus a small percentage, or only up to a certain point—reflecting a desire to control the cost curve of the program (Fernandez, 2012; Biggs et al. 2005; Weaver 1988; Whitehouse 2009). Most OECD countries use some form of indexation to automatically adjust pension benefits (Fernandez, 2012; Whitehouse 2009). But far from everything that could—or perhaps should—be indexed is. Instead, what one finds is a much more mixed picture. Sweden, for instance, uses a form of indexation in the calculation of pension benefits, but does not use indexation to automatically adjust the child benefit or the income ceiling in its unemployment insurance system, to give just two examples. The picture is similar if one looks at other countries (see Weaver’s [1988] study of indexation in the US).

A number of studies have sought to make sense of and explain (at least some of) this variation—both between countries when it comes to the indexation of pensions (Fernandez, 2012) and within countries (Weaver 1988). For Fernandez (2012, p. 242), “macroeconomic conditions and the partisan structure of government” (at key moments) best explains differences in the indexation of pensions between countries—with countries with higher inflation rates and more entrenched Christian democratic parties being more likely to index pensions and to tie them to wages. Similarly, for Weaver (1988), whether and how programs are indexed reflects a combination of factors—from the credit-claiming or blame-avoidance incentives of the politicians involved to the presence or absence of powerful interest groups or stakeholders pushing for a particular design choice. Other studies have looked at how indexation, or the altering or removal of indexation rules, can and has been used as a tool to partially dismantle (or at least make less generous) certain welfare state provisions (Green-Pedersen et al 2012; Lindbom 2010; Pierson 1996).

There are both economic and political arguments for and against indexation that are advanced and discussed in the literature, including arguments that are specific to the indexation/non-indexation of specific social benefits (see Weaver, 1988; Whitehouse 2009). Broadly speaking, economic arguments in favour of indexation tend to emphasise the predictability and certainty it helps provide regarding the real value of benefits (which can erode substantially because of inflation), as well as the supposed benefits of transferring the “determination of benefit levels … from the political arena to the realm of the law” (Whitehouse 2009, 11). When it comes to pensions, for instance, this can help “minimi[ze] political competition and intergenerational social conflict over pensions” (Whitehouse 2009, 12) partly because of the way it “anonymises” responsibility for the changes that are made (Weaver 1988; Whitehouse 2009, 11). The economic arguments against indexation tend to emphasise the loss of flexibility over budgetary policy that indexation can entail, and the impact this can have both on control over fiscal policy—and not least the ability to limit spending during times when inflation is high—as well as the ability of the government to distribute the costs of fiscal cutbacks fairly. The argument is also sometimes made that indexation can contribute to more inflationary fiscal policy by effectively shielding part of the electorate from the negative consequences of inflation, making them less willing to bear the costs of bringing inflation down (Whitehouse 2009).

From a political standpoint, there can be both strategic and more ideologically rooted reasons to favour indexation/non-indexation. From a strategic perspective, depending on the context, indexing/not-indexing can help politicians claim credit or avoid blame. In a context where politicians want to be able to take credit for looking after pensioners, there can be benefits to pensions not being indexed, as this provides politicians with opportunities to raise pensions and claim credit for doing so (Weaver 1988). Not-indexing benefits can, in a different context, help politicians avoid blame for cutting benefits (in real terms), since these cuts will occur gradually, by default, and with low visibility (Green-Pedersen et al 2012; Lindbom 2010; Pierson 1996; Vis 2016). In yet another context, the decision to index benefits can be what politicians have a strategic, political, interest in claiming credit for.

Looking at the more ideologically rooted political reasons to favour indexation/not-indexation, politicians have good reason to make their ideologically preferred policy outcome the default one. Changing the status quo is (nearly) always going to be more difficult than preserving it, and so to the extent that the default outcome aligns with their ideological commitments, politicians will have a reason to push for it. This can mean favouring indexation or being opposed to it, or it can mean favouring particular indexation rules (i.e. that the value of the benefit should track changes in wages rather than prices, or vice versa, etc.). A politician who wants to cut benefits would thus have a reason to not want benefits indexed, while a politician who wants benefits to stay generous would have a reason to index benefits.

Indexing as a democratic problem

Indexing decisions also implicate democratic norms and values. As was discussed above, indexation rules have been used to cut welfare benefits and to enable politicians to avoid blame—which raises questions regarding the implications of indexation rules for transparency and democratic accountability. The literature on the politics of indexing also points towards the relevance of questions concerning the discretion and flexibility that ought to be afforded currently elected policymakers, and the democratic implications of the type of choice architecture that surrounds government benefits. There are at least five democratic problems, or concerns, that are worth distinguishing, each of which will be discussed below. But before proceeding to consider these, it is important to say a bit more about democratic accountability and the democratic principles at stake in indexing decisions.

Implicated democratic principles

Indexing decisions implicate questions regarding democratic accountability, I argue in this paper. But what does democratic accountability mean? And what is the relationship between accountability and other democratic ideals that are often mentioned alongside it, such as transparency and good governance? The literature on these questions is vast and reflect disagreements not only about the best way to make sense of the meaning of these concepts and how they relate to each other, but also broader questions regarding how best to further them in a democratic context (see Bovens, Schillemans and Goodin, 2014; Philip, 2009; Warren, 2014). In a broad sense, democratic accountability denotes the idea that elected representatives—and perhaps especially democratically elected governments—are (or should be) answerable to the people for their conduct in office. A (perhaps the) central component of democratic accountability is thus the idea that elected representatives owe the people (however delineated) reasoned explanations/justifications for the decisions, priorities and choices they make while in office. Implicit here is also the idea that in a democracy the people are supposed to be able to hold elected representatives accountable for their conduct—through their choice of who to vote for or through some other means.

Questions regarding democratic accountability thus also implicate adjacent principles regarding transparency and how best to further good governance. Indeed, a key question in the literature is how best to understand the relationship between accountability, transparency and good governance (Fishman 2016; Hood 2010; Meijer et al. 2018, Papadopoulos 2010). Does transparency with respect to the business of government necessarily or always promote democratic accountability? Does it “strengthen the (information) position of citizens” (Meijer, Hart, Worthy 2018, 4) and enable citizens to hold public office holders accountable or does it, rather, undermine accountability by leading to “one-way communication rather than real answerability in effective dialogue” (Hood 2010, 992)? And does greater accountability necessarily improve governance and responsible policymaking? If there is an answer to these questions that emerges from the literature, it is that accountability, transparency and good governance are goods that can and do come into conflict, even though they also can be and often are mutually supportive (see Hood 2010; Papadopoulos 2010). This then raises the further question—also central to the literature on these issues—of how mechanisms for promoting the right kind of accountability and the right kind of governance can be institutionalised. Though not framed in exactly those terms, the different perspectives on indexing that the next section will describe and discuss can be seen as reflecting different interpretations of what would be required to honour norms regarding accountability and transparency when it comes to indexing—would indexing “strengthen the (information) position of citizens” or obfuscate things?—as well as different ideas about how best to promote good governance.

Five democratic concerns

This section outlines and discusses five democratic concerns related to the indexation of government benefits. The first two concerns could be said to emerge from the literature on welfare state retrenchment. These are rooted in concerns about the consequences of not indexing benefits—the first having to do with how non-indexation can enable politicians to engage in a kind of deception, and the second having to do with how non-indexation may contribute to a kind of policy drift. I then proceed to consider three concerns related to indexing benefits that emerge from Weaver’s (1988) influential work on the politics of indexation in the US, though they are framed and described below in slightly different terms than Weaver uses. At the heart of Weaver’s concerns about indexation are concerns about its impact on democratic deliberation, responsible policymaking, and democratic accountability. Below I attempt to sketch and critically analyse the roots of these concerns, distinguishing concerns about how indexation reduces discretionary authority, from how it may reduce a sense of ownership over certain policy outcomes, from the concern that it may cause some policy questions to disappear from the political agenda and anonymise responsibility. Section "Rules of thumb for when and how to index: a democratic perspective" attempts to distil the key takeaways from the discussion below into two principles, or rules of thumb, for how to think about the indexing of benefits from a democratic perspective.

Deception

When benefits are not indexed, the tendency of people to think of the value of money in nominal rather than real terms—the money illusion—can enable politicians to engage in a kind of deception. They can claim to be increasing benefit levels when they are merely raising them to preserve their real value—or even when they are raising them to levels below what would be required to preserve their value in real terms, i.e. when they are de facto cutting the benefit. And they can claim to be preserving the generosity of a benefit in not raising its nominal value, while de facto reducing its real value. Indeed, they can make cuts to programs without having to take or justify any action whatsoever, since inflation and wage growth will gradually erode the real or comparative value of a non-indexed benefit automatically (if no action is taken to preserve it). The concern about deception could thus be seen as a concern about transparency and accountability. The money illusion can enable politicians to avoid having to account for their actions.

And, indeed, there is evidence that non-indexation is used in precisely this way, to obscure cuts to welfare programs. In his work on welfare retrenchment, Pierson (1996) describes “poor indexation” as a “low-visibility strategy” of gradually lowering the value of benefits, using it as an example of how governments can try to “diminish their own accountability for unpopular reforms” (Pierson 1996, p. 26, italics in the original). Green-Pedersen et al (2012), in a study of indexation in four countries—Denmark, Sweden, Germany, and the United Kingdom—found that non-indexation had been used to stealthily reduce the real value of social benefits in the countries studied. This is perhaps not surprising. Politicians can have good (political) reasons to use policy tools that enable them to avoid being blamed for cutbacks.

This is also, however, part of what makes it relevant to consider the issue of indexation through a democratic lens. Citizens are entitled to expect that meaningful changes to welfare state programs are justified and defended publicly. If non-indexation is used as a tool to obfuscate gradual, but meaningful, changes to welfare state programs, an important democratic norm seems to have been violated. Of course, there is nothing inherently or necessarily deceptive about the non-indexation of benefits. In particular contexts there may be good policy reasons to use non-indexation to gradually lower benefit levels, or other reasons for choosing not to index a particular benefit. And there is nothing that says that it cannot be done in a transparent manner, which does not lead to or facilitate deception. Using non-indexation to make cuts to benefit programs becomes a democratic problem only when (and if) policymakers take advantage of people’s tendency to conceive of the value of money in nominal rather than real terms to disguise the fact that benefit levels are being cut.

But the fact that non-indexation can facilitate this sort of deception provides, I want to suggest, a democratic reason in favour of indexing benefits, making the benefit track some metric—be it inflation or wage growth or some combination of the two—that preserves its publicly articulated purpose. Doing so would force politicians wanting to cut benefit levels to take active steps to do so, making it harder to make meaningful changes to programs without having to publicly articulate and justify the decision to do so. There may of course be countervailing reasons against indexing benefits that outweigh this deception-based concern. But, other things being equal, the concern with the way in which non-indexation could facilitate deception and undermine democratic accountability provides a reason in favour of indexation.

Neglect

Non-indexation means the real value of a benefit will gradually decrease unless active steps are taken, at regular intervals, to counteract the effects of inflation and/or wage growth. Simply preserving the value of a non-indexed benefit thus requires the repeated attention and effort of policymakers. Since the attention and effort of policymakers is a finite commodity and since it can be hard—and in some institutional contexts extremely hard—for policymakers to agree on and enact policy changes (even ones that are status quo preserving), not having benefits indexed can create a risk of benefits being cut in ways that undermine their publicly articulated purpose, not because of a policy choice but simply as an unintended consequence of the difficulty of overcoming policymaking or legislative inertia.

This does not mean that non-indexed benefits will necessarily see their real value erode over time (Weaver 1988). Politicians might stand to gain politically from being seen as protecting the real value of a popular benefit—like pensions, for example—and be able to use the fact that it is not indexed to get the issue of pensions onto the political agenda. Indeed, this could even result in the real value of the benefit being pushed higher than it would have been had it been indexed. But the fact that repeated legislative initiatives are necessary just to preserve the status quo nevertheless creates a real and tangible risk of the benefit seeing its real value erode due to neglect.

Moreover, it makes the value of non-indexed benefits more vulnerable to the vagaries of political calculations regarding credit-claiming and blame-avoidance. While non-indexation can present politicians with opportunities for credit-claiming actions that preserve the real value of a benefit, a desire to not be seen as a spendthrift or to be blamed for raising benefits paid out to stigmatised or unpopular groups can result in politicians not being willing to raise certain benefits in ways that would preserve their real value. Indeed, the money illusion—to the extent that it is shared among the public and among politicians and media actors—can perhaps make it difficult for politicians to convince a sceptical public that nominal increases in the amount a benefit pays out are in fact status quo preserving, thus making it even less likely that the real value of “unpopular” benefits are preserved over time if they are not indexed. An illustrative example of this policy dynamic can be seen in the size of the economic support Sweden pays to persons with refugee status who are unable to work. As of 2023, the amount paid per day had not been changed since 1994, a period of almost thirty years. Accounting for inflation, the real value of the benefit had been cut by close to 40% during this time, drastically reducing the ability of the benefit to support the people living off it economically.³ Though one can only speculate as to the precise reasons why this occurred, a plausible explanation seems to be that it reflects a combination of neglect and blame-avoidance among elected policymakers.

This also gets at what makes this issue a democratic problem, which is that it can produce outcomes that are unintended and that do not reflect or realise the publicly articulated aims of the (democratically enacted) underlying policy. Another way to frame the issue would be to say that non-indexation makes policy drift more likely (Béland et al. 2023). Policy drift refers to “a shift in the context of policies that significantly alters their effects” (Hacker 2005, 45)—the context in this case being the gradual depreciation of the value of the nominal amount paid in benefits due to inflation. When the effects of this shift are that the benefits no longer serve the function they were intended to serve—of enabling a targeted group of people to afford basic necessities, for instance—the publicly articulated purpose of the social policy will have been undermined. To the extent, therefore, that well-tailored indexation rules help prevent policy drift—and, more precisely, deleterious effects of neglect—and in that sense promote good governance, there are democratic (in addition to policy preservation) reasons to embed them in the structure of benefit programs.

Lack of discretion

Numerous concerns have also been raised regarding the implications, from a democratic perspective, of indexing benefits. At their core, these are about fears that indexing will undermine accountable and responsible policymaking by automating and anonymising impactful decisions and reducing the power of elected policymakers to govern. Let us begin by considering the last of these, that indexing reduces the discretion and decision-making power of policymakers in ways that are problematic, from a democratic perspective.

Though they remain free to change benefit levels and indexation rules, indexing benefits can make it harder for policymakers to control how much the government spends and to prevent nominal benefit level increases. In effect, indexation creates a status quo bias that favours nominal increases in spending. And to the extent that indexing reduces budgetary discretion, it can effectively force a government that wants to limit or reduce spending to only cut those parts of the budget that it enjoys more discretion over—which may result in a budget that does not reflect the (actual) will of lawmakers or of the electorate.

There are two ways of making sense of what the democratic “problem” is here: one is that there is something inherently problematic, from a democratic perspective, when the hands of current lawmakers have been tied by a previous crop of lawmakers. Another way to frame the issue would be to say that the lack of discretionary authority can lead to poor governance—e.g. budgets that are skewed in ways that produce unwanted negative consequences. Both ways of framing the issue alert us to a potential democratic problem, but it is important to add a few caveats.

Elected policymakers, to begin with, never enjoy unlimited discretion and flexibility. They have to operate under the rules and within the constraints set by previous lawmakers and governments and are bound by a complex web of laws, regulations, commitments and other types of constraints. Laws, to take just one example, tend to remain in effect until they are changed or amended. Newly elected policymakers do not start off with a clean slate. Thus, to the extent that indexing government benefits reduces the discretion and flexibility of elected policymakers, it would be fair to say it is only doing so at the margin. And though indexing changes the default outcome—what happens if no decision is taken—and thus changes the “choice architecture” in which the decision takes place, it is important to keep in mind that indexing does not deprive lawmakers of the power or the ability to change or reduce benefit levels (in real terms). Indexing merely makes it more difficult to reduce the real value of the benefit by freezing the amount it pays out in nominal terms. The bar for making other types of changes—including changes that would make it more generous in real terms or to make more substantial cuts to the program—remain unaffected. Thus, while there is reason to factor the lack of discretion concern into the design of indexation rules, it is not clear it provides a particularly strong reason against indexing government benefits altogether.

However, while the lack of discretion concern does not provide a strong reason against indexing altogether, it does suggest that indexation rules should be designed in ways that preserve at least some discretion on the part of policymakers, at least in certain contexts or with regards to certain benefits. This could counteract the potential for indexation rules to have a distorting effect on budget priorities and/or reduce (at least at the margin) the ability of elected policymakers to govern effectively, as could be the case if indexation rules make it hard to constrain fiscal spending during times of high inflation, for example.

Ownership

Another concern regarding indexation centres around ownership, and the implications for accountability and responsible policymaking of having adjustments to benefit levels be automated in the way indexing implies. The concern here is partly that indexation will erode the sense of ownership policymakers ought or need to have to govern responsibly. Weaver (1988), for instance, argues that indexation could erode responsible policymaking by making policymakers less able and willing to make necessary changes to indexed programs—perhaps because of a fear that they would be blamed by voters for cutting popular programs. With indexation, he argues, politicians are held accountable “for maintaining specific policy outcomes”, not “for making policy in a responsible manner” (Weaver 1988, 251). Put differently, indexation (in effect) threatens to create too much transparency regarding decisions to reduce (the real value of) benefits—causing policymakers to distance themselves and to not take ownership of and responsibly for the fiscal and economic consequences the indexed benefit programs produce. A related ownership-concern focuses on the way indexation anonymises responsibility. Though sometimes framed as an advantage of indexation (see Whitehouse 2009), anonymization also risks eroding accountability by making it less clear who is responsible for (and who should be blamed for) problems created by automatic increases to indexed benefits.

Here too, however, it is important to add some caveats to these concerns. To begin with, the potential problems described above are not unique to the policy areas affected by indexation. As was discussed earlier, policymakers are not handed a clear slate on which to make policy when they are elected. Laws, social programs, and other sorts of commitments already exist and constrain and shape what elected policymakers can do. Though problems concerning accountability and the toleration of poor outcomes are (or can be) real and pervasive, these are problems that touch on much broader issues of democratic concern—regarding the incentives that shape the actions (and inactions) of elected policymakers. One could even argue that indexation is conducive to ownership and accountability in at least one way: by making it harder to “dismantle by default” and with “low visibility” (Green-Pedersen et al. 2012; Pierson 1996), indexation can help force policymakers to own and justify—and therefore be accountable for—cuts they want to make to the real value of benefits. It is of course true that this may make policymakers reluctant to enact cuts to popular programs even if it would be wise from a broader economic or budgetary perspective to do so. But to the extent that that is a problem, it seems the solution should be to find ways of ensuring that policymakers are held accountable also for their sins of omission, for not addressing problems they have the power to address and for tolerating poor outcomes—particularly since it seems far from clear that not indexing benefit programs would make elected policymakers more responsible and focused on good governance, given the concerns regarding neglect and policy drift that were discussed earlier.

Rather than providing a reason against indexation, the concern about ownership would thus seem to counsel in favour of indexing benefits in way that preserves their publicly articulated purpose.

Disappearance from the political agenda

This concern centres around the potential that indexation could be used to help take and keep an issue off the political agenda, removing it from public discussion and shifting power over important policy decisions from elected officials to bureaucrats and administrative agencies. The disappearance concern thus draws on a broader worry about the democratic implications of too much technocracy—of too much “automatic government”—and the implications it carries for who wields power and sets policy. To make sense of this concern, it is important to remember that if pension benefits are indexed, there will not be the same need to consider, year-after-year, whether and by how much pension benefits should be raised (in nominal terms) as there would be if the real value of the benefit was being eroded due to inflation. Indexation could thus be said to help remove the need for a certain kind of recurring public discussion that would otherwise be necessary. Thus what, from a certain kind of economic perspective, can be seen as a benefit of indexation—that it can help “minimi[ze] political competition and intergenerational social conflict over pensions” by transferring the “determination of benefit levels … from the political arena to the realm of the law” (Whitehouse 2009, 11)—can, from a democratic perspective, be seen as a real downside. It is, after all, through public deliberation that democratic will formation is supposed to occur.

Whether indexing in fact has this kind of side-lining effect and whether it undermines the quality of public deliberations concerning things like pensions and other benefits are empirical questions that cannot be answered here. But it is worth noting that it is not obvious how not indexing benefits would serve to improve the quality of public deliberation about whether and how they should be adjusted. People’s tendency to think of monetary values in nominal rather than real terms undercuts the suggestion that people would be in a better position to understand the stakes of a debate about changes to the pension system if benefits were not indexed.

The argument that non-indexation would be preferable because it would force politicians to confront the question of whether and by how much to raise benefit levels regularly also has at least one serious flaw. The mechanism doing the “forcing” here seems to be the bad consequences that would befall the people who rely on the benefits if they are not raised to account for inflation. People would complain and worry, and this would “force” the issue onto the political and legislative agenda. One problem here would seem to be the time lag between when the problems created by not having the benefits indexed would start to appear and be felt and when they would be addressed—if, indeed, the policymakers decide to raise the benefit levels to protect their real value. Conceivably, things would need to get fairly bad, at least for some people, before the issue would be able to force its way onto the political agenda. Another problem is that not all groups and constituencies are equally able, sufficiently well connected, and/or perceived sympathetically enough by the media and politicians, to have their concerns reach the ears of elected policymakers. Though it seems conceivable that politicians would regularly adjust pension benefits to protect them from the effects of inflation even if they were not indexed, it seems far less likely the same would be true for cash benefits given to more stigmatised groups.

Thus, while the disappearance concern describes a real potential problem with indexation and how it can be used, it is far from clear that the cure—not-indexing benefits—would not be worse than the disease.

Rules of thumb for when and how to index: a democratic perspective

The decision regarding whether and how to index particular social benefits should rightfully take into account and be informed by economic and political arguments and considerations. But there are also democratic norms and values at stake in the decision, as the discussion above has highlighted. These boil down, I want to suggest, to two rules of thumb, or guiding principles, for when and how to index social benefits:

Rule of thumb 1: Everything else being equal, benefits should be indexed in a way that is status quo preserving. By status quo preserving, I mean preserves or maintains the publicly articulated aim or purpose of the social benefit and the program of which it is a part. This means (1) that social benefits should, as a general rule, be indexed. This would make it harder to make cuts to the real value of benefits without publicly justifying the decision to do so. And (2), it means that the indexation rules should be designed to maintain (rather than undermine) whatever the publicly articulated aim or purpose of the social benefit and the program of which it is a part is. In some contexts, for some social benefits, that might mean tying its real value to changes in prices, since what matters is the purchasing power of the benefit. In other contexts, for other social benefits, it might mean indexing the benefit to track changes in average wages, since what matters is the relative economic standing the benefit is meant to confer on the recipient. This, too, would help prevent meaningful changes being made to social welfare programs without them being publicly justified.

Rule of thumb 2: Give the government some discretion over automatic adjustments—but force it to publicly justify any decision to deviate from the automatic adjustment proscribed by the index to which the benefit is tied. Giving the government some (though not unlimited) discretionary power over automatic adjustments—e.g. allowing them to be paused for one year, let us say, or allowing the government to slightly alter the adjustment up or down—could help prevent the kind of negative consequences some fear could result from automatic adjustments (in the form of fiscal policy being too expansionary when inflation needs to be reined in, for example), by empowering the government to govern.⁴ At the same time, it would be important that this discretionary power was accompanied by a formal requirement that any decision to pause or alter the automatic adjustment be publicly justified in writing. Though there would still be reasons to fear that the government would use its discretionary power to hide meaningful changes to benefit programs, a requirement that deviations be publicly justified would at least help ensure some measure of transparency and public accountability.

Conclusion

This paper has examined the democratic stakes of indexing government benefits and considered the reasons for and against indexing benefits from a democratic perspective. Given the reality of the money illusion and the way in which non-indexation can be used as a tool to cut benefits with low visibility, it has argued, there are good reasons to think that democratic accountability would be enhanced if benefits were indexed. If benefits were indexed in a way that preserves their publicly articulated purpose, this would force policymakers wishing to make substantive changes to own and publicly justify any such decision. Indexing benefits would also protect them from being eroded due to political gridlock or neglect. However, the risk that indexing benefits could skew budget priorities and reduce the ability of elected policymakers to govern responsibly and effectively, it has further argued, provides a reason to design indexing rules in ways that afford governments some discretionary power.