Policy change in flat pensions: comparing Canada and the UK.

Beland, Daniel ; Waddan, Alex

Much has been written about pension reform in advanced industrial countries. A central aspect of this discussion concerns the assessment of the scope and nature of policy change currently witnessed within the modern public pension systems created during the postwar era (Marier 2008; Myles and Pierson 2001; Weaver 2010). The following article explores the issue of policy change in relationship to the flat-rate pensions created in the UK and Canada in the years following the end of World War II. Taking a long-term historical perspective on the development of these flat pensions and drawing on the recent policy literature on incremental and transformative change (Hacker 2004; Streeck and Thelen 2005), this article suggests that, over time, the UK flat pension (Basic State Pension) has witnessed more dramatic changes than its Canadian counterpart and that the greater mobilization of beneficiaries defending the flat pension program in Canada helps explain this key difference between the two countries. Overall, this article contributes to the ongoing debate on incremental change by exploring how subtle policy alterations may or may not lead to transformative change over time. The article also adds to the literature on pensions by suggesting that contributory pensions such as the UK Basic State Pension (BSP) are not necessarily more politically resilient than universal, non-contributory flat pensions like Canada's Old Age Security (OAS).

The article begins with a brief general discussion about the relationship between old-age pensions and policy change, before exploring the UK and Canadian cases and drawing general lessons about the evolution of flat pensions in these two countries. Of particular interest are policy changes that affect the eligibility criteria for the public pension and alterations to the level of benefit. With regard to the former, relevant changes would include modifications to the qualifying number of contributory years needed to receive a pension or a redefinition of what counts as a contributory year. In addition, raising or lowering the retirement age, either to reflect changing concepts of gender equality or expectations about life expectancy, could have important potential implications for eligibility. Alterations to the level of benefit could be straightforward decisions to raise or lower payments, but also include moves to index, or re-index, benefits to alternative measures of increases in the cost-of-living. It is also important to investigate whether and how benefits are subject to taxation.

Pension reform and policy change

Public pension policy is replete with demands that compete for priority. The obvious primary function of a state-organized pension system is to provide income support for seniors, but this does not in itself determine how generous pensions should be. For the state, providing for retirees has to be set against controlling aggregate costs. Furthermore, governments must decide how to balance the desire to provide generous benefits against concerns about the incentive effects of encouraging or discouraging people from saving for their own retirement through their working lives. A generous universal benefit might have low disincentive effects but it is likely to be expensive. A greater reliance on income or means testing will reduce overall cost but will increase disincentive effects (Bozio, Crawford and Tetlow 2010).

Given these complex issues and the inherent difficulties in satisfactorily resolving these policy trade-offs, it is unsurprising that in recent decades UK and Canadian governments, with pension systems that incorporate both flat and income- or means-tested elements, have consistently amended their pension policies in attempts to reform schemes that were initially established in the aftermath of World War II. In this context, and taking into account the balance between universality and income or means tests, and the episodes of retrenchment and expansion, the best manner in which to understand pension policy in the UK and Canada is to take a historical perspective, which allows for assessing which of the changes that have been enacted have had the greatest impact, and which have turned out to be less significant than first anticipated.

As is outlined below, both the UK and Canada created flat-rate pension benefits--the BSP in the UK and OAS in Canada--that have remained at the center of government pension provision since the early post-war era. Pension policy in these two countries has not been marked by long periods of institutional stability, but neither have there been regular and obvious moments of path departing change. Rather, policymakers have enacted apparently incremental changes, some of which have nonetheless produced significant policy transformation over time. Therefore, a thorough understanding of pension policy change requires a proper examination of changes that may appear to be un-dramatic, and which may indeed be presented as such by political actors anxious to avoid controversy, but which can have important long-term impacts (Hacker 2004; Streeck and Thelen 2005).

Tracking policy development over an extended period also allows observers to judge which features of the respective systems in the UK and Canada have proved to be the most durable and resistant to change, and which features have proved less resilient. Here, the comparison between our two cases is potentially instructive. Why was the Thatcher government able to introduce a change to the method of indexation of the BSP, whereas when Prime Minister Mulroney attempted something similar with respect to the OAS in Canada, he was forced to retreat in the face of strong opposition? In this context, we must examine which factors determine the institutional resilience of particular programs. What policy legacies have resulted from previous patterns of development and, importantly, how have these shaped the public's perception of their stake in the program? Is pension policy closely watched by groups able to mobilize protest if the beneficiaries of existing arrangements are threatened (Beland and Marier 2006; Marier 2008)?

As Andrea Campbell (2003) has suggested in her analysis of the United States, the political mobilization of retirees is a potent explanation of policy development. Interestingly, however, the Canadian case concerns mobilization over a non-contributory, universal flat pension scheme, which is different than the Bismarckian social insurance model in place in the United States, as well as in continental Europe. According to scholars like John Myles and Paul Pierson (2001), contributory pension schemes are especially resilient politically, in part because they create a sense of entitlement derived from individual contributions. In this context, since the BSP is a contributory scheme, one would expect that it would prove less vulnerable to cutbacks than OAS, which is not "earned" through payroll contributions.

A historical perspective also provides insight into which reforms have proved to have longevity. In the UK, for example, governments have at different times made apparently significant reforms aimed at introducing an earnings-related social insurance style benefit, to sit alongside the BSP. For all the effort put into establishing those schemes, however, they have failed to take root. (1)

It is also important to recognize that policy changes sometimes do not have as big of an impact as might have been expected. For example, in 1989, a policy of fiscal "clawback" was instituted in Canada to reclaim some OAS benefits through the tax system from seniors with high incomes. This was important in principle but did not prevent the vast majority of eligible Canadians from keeping their full OAS pension (Beland and Myles 2005).

Finally, taking a long-term perspective also enables us to assess whether there has been a constant direction to reform efforts. So, for example, have the universalistic aspects of state pension policy in the UK and Canada been consistently either expanded or retrenched? Or have policy initiatives been both pushed and pulled following alternative agendas, with the consequence that there have been moments of both retrenchment and expansion?

As illustrated below, the evidence suggests that, in both countries, pension policy reform efforts have followed contradictory trajectories, but overall, despite being a non-contributory scheme, OAS has proved more resilient than BSP, which is a reality that challenges the above-mentioned assumption that contributory schemes are more resistant to restructuring than non-contributory ones (Myles and Pierson 2001).

The UK

Over the last 60 years, pension provision in the UK has evolved into a complex array of state and private arrangements. (2) Underlying state provision is the BSP, established by the post-war Labour government, modelled to some extent on the recommendations contained in the 1942 Beveridge Report (Beveridge 1942). A flat-rate pension, the BSP is something of a hybrid between a universal and a contributory scheme. The fact that individuals are only eligible for the full BSP in retirement if they have an appropriate contributory record through paying National Insurance (NI) contributions out of their payroll (or performing activities recognised as a substitute for contributions) means that it is distinct from the Canadian OAS, which is granted to all on the basis of residency in the country. Contributions towards the BSP are measured in years of payment rather than the amount paid, with the original requirement being 44 years of contribution for men and 40 years for women. Thus, the BSP is not universal, as many people reach retirement age without making the requisite contributions to receive the full BSP. But the scheme is compulsory: individuals cannot opt out of paying NI, and seniors who have made some contributions throughout their working age do receive BSP benefits on a sliding scale. Also, although contributory, the BSP is distinct from a Bismarckian social insurance scheme; its flat-rate payment means it does not replace past earnings in a manner reflecting previous differentials in earnings (Barr 2004: 31).

The BSP has been reformed in ways that constitute both retrenchment and expansion of the program, but since 1948, it has been a constant feature in the changing landscape of UK pension policy. Another consistent feature of state pensions has been the reliance on means-tested top-ups for seniors left with inadequate income by the BSP and with minimal other sources of income. In addition, the UK has experimented with social insurance style pension benefits by putting in place different earnings-related schemes, but none of these has passed the test of longevity. Further to state pension benefits, the UK witnessed the growing availability of pensions in the private sector through occupational schemes and private pension plans.

Implementing Beveridge

William Beveridge (1879-1963) was the intellectual founder of post-war British social policy, but when examining how far his blueprint for welfare state development was put in place (Beveridge 1942), we should remember that many social welfare programs were competing for scarce funds in an austere post war fiscal environment (Timmins 1996). In turn, this limited how generous any pension settlement could be and left a lasting legacy affecting the development of UK pension arrangements.

In the end, the Labour government (1945-1951) only followed some of Beveridge's advice. The BSP was established as a contributory scheme based on NI contributions paid out at a flat rate. But while Beveridge had called for payment of the full pension to be delayed for twenty years to build an actuarially sound reserve fund of contributions, the government deemed that politically impossible and the BSP was immediately paid out to men aged 65 and over and women aged 60 and over. Yet, reflecting the limited fiscal resources available in the late 1940s and also that the aim was only ever to offer a parsimonious benefit, the BSP was paid at a level "below the measly estimated level of subsistence" (Blackburn 2002: 58). As a consequence, since the 1940s, non-contributory means-tested payments to supplement the BSP have played a "key role" in helping low-income seniors in the UK (Bozio, Crawford and Tetlow 2010: 51).

A significant consideration during the postwar years was the increasing number of workers with occupational pensions. In 1956, about a third of workers were enrolled in occupational pension schemes, rising to over 50% by 1967, partially reflecting a tight labour market that made employers aware of the need to attract "good" employees (Pemberton 2006). Nevertheless, many remained reliant solely on the state for income in retirement. This produced pressure to introduce a state-run earnings related supplement to the BSP for those not in occupational or private schemes. A graduated retirement benefit was introduced in 1961 but this never gained traction (Bozio, Crawford and Tetlow 2010: 9); in 1975, the State Earnings Related Pensions Scheme (SERPS), to be implemented in 1978, was established. Also in 1975, a potentially important change was made to the calculation of the BSP, when the Labour government formally linked the level of benefits with the higher of the rate of price inflation or the growth in average wages. The BSP had remained structurally unchanged prior to this time with payments raised on an ad hoc basis, although generally these increases had at least kept up with the growth in average earnings and had exceeded the rate of inflation. In July 1948, measured in 2011 sterling, the full BSP stood at 38.48 [pounds sterling] per week. This had reached 85.29 [pounds sterling] by November 1975 (Department of Work and Pensions 2012: 14). In this context, the significance of the change made in 1975 was perhaps only properly illustrated when the method of indexation was again changed by the Thatcher government.

Downsizing the state pension

In 1980, in a move that aroused surprisingly little opposition given its significance, the indexing of the BSP was altered (Blackburn 2002: 285-6). It was now only to rise in line with inflation rather than with that measure or earnings; this meant that, without any ad hoc increases in prospect, its replacement rate value would decline significantly over time. The BSP's value dropped from 24.6% of overall average earnings in November 1980 to 16.8% by the time the Conservatives left office in 1997 (Department of Work and Pensions 2012: 14). Thus, although there had been no change in eligibility for the BSP, more seniors became reliant on secondary sources of income if they were to keep up with the working-age population.

Yet, despite the significance of this change in the method of benefit indexation, the government suffered little political backlash, certainly if measured in electoral terms, in response to this change. (3) This response contrasts with events in Canada under Prime Minister Mulroney and even with the United States, where the Social Security program's reputation as the "third rail" in political life (touch it and die) was reinforced when President Reagan was forced to back down from his plans to retrench benefits in the early 1980s (Beland and Waddan 2012). For social policy scholars (Myles and Pierson 2001), however, this is a potentially counterintuitive narrative, as the BSP's status as a contributory benefit and, therefore, as an earned entitlement, should have provided it extra public support leading to popular protest when its value was downgraded. Yet this result did not happen on a large scale. The opposition parties in parliament did vote against the measure but this was of little significance given the government's large majority. There was also opposition from pensioners' groups but again this did not unduly influence the government, partially because of the fragmented nature of even the best-organized age advocacy groups in the UK at the time and also because the most voluble protest came from groups associated with the labour movement (Pratt, 1993: 124-30), which did not resonate with the Thatcher government.

In a further move in 1986, the Thatcher government reduced the future benefits that would be paid out by SERPS - the state-run earnings related scheme introduced in the 1970s--and implicitly encouraged people to opt out of SERPS altogether and start their own private pension plan (Blake 2004: 22-5). Prior to implementing this act, the government had considered more radical proposals that would have abolished SERPS and further slowed increases of the BSP with the expectation that the private sector would step in to dominate the area of pension provision. If enacted, these plans would have constituted a "radical transition" with the state reduced "to the role of pension provider for the poor" (Taylor-Gooby 2005: 121). Interestingly, however, these plans provoked opposition from the pension industry and the Confederation of British Industry and raised concerns in the Treasury about the transitional costs always present when switching from a pay-as-you-go arrangement to a funded system (Bonoli 2000: 67-72). Thus, some contemporary commentary depicted the government as backtracking in 1986, but, importantly, this retreat resulted from opposition from the government's allies and the Treasury rather than pressure from pensioners' groups, which remained hostile to the final legislation. Furthermore, the government's policies still constituted a case of "layering," which involves "the grafting of new elements onto an otherwise stable institutional framework" with the intention of changing that framework (Thelen 2004: 35). In this case, the government had made state provision less attractive, meaning that pension "expansion would be channelled into the private sector" (Pierson 1994: 59).

New Labour and pensions

Throughout the 1990s, the Labour party reconsidered much of its wider social policy outlook, which by the 1997 election included a retreat from previous promises to re-index the BSP to wages as well as inflation. But by the time Labour left office in 2010, the government's position on pension reform had evolved considerably. The initial strategy was to alleviate poverty amongst seniors through means-tested additions to the BSP, rather than by increasing the value of that flat pension. Also, a new Second State Pension (S2P), replacing SERPS, was introduced to support low paid workers and people unable to work for reasons such as caring responsibilities. The government also established regulatory structures that would better enable workers in jobs with moderate incomes to take up private pensions. These measures, however, had limited success as take-up of the new schemes remained very limited (Evandrou and Falkingham 2009; Waine 2009).

The government also established an independent Pensions Commission to analyze the options for further reform. Importantly, the reports issued by this Commission (Pensions Commission 2004 and 2005) provided new pressures for a major overhaul of the existing state pension arrangements (Brooks and Denham 2005). The Pensions Commission (2005: 2) also noted that pensioner income from occupational and private pensions was at an all-time high, but that "voluntary private provision is not growing: rather it is in serious and probably irreversible decline." In this context, Labour began its third term in office in 2005 committed to pension reform: "The pensions system we have today is rooted in the society of the 1940s. Society has moved on and, unless we act now, women and carers retiring in the next two decades will continue to suffer the effects of the system of contributions which applied during their working lives" (Department of Work and Pensions 2006: 12). The government acknowledged that the increasing reliance on means-testing, while it had particularly benefited women, was inappropriate because it acted as a disincentive to saving. In 2005-06, the full BSP for a single person was 82.05 [pounds sterling] per week, way below the minimum income of 109.45 [pounds sterling] per single person per week guaranteed to seniors (Brooks and Denham 2005: 23). Yet, closing the gap between the BSP and the income that the government deemed the minimum acceptable for seniors was daunting because of the cost implications of a universal increase in the level of BSP.

On the other hand, the BSP was brought into further disrepute by the manner it distributed benefits in a highly gender discriminatory fashion. Whereas over 80% of recently retired men were in receipt of the full level of BSP, only just over 30% of women were in this position (Department of Work and Pensions 2006), illustrating that the program fell short of true universality.

In this context, the Pensions Act of 2007 brought important change. From some point after 2012, the BSP was to be re-indexed to rises in average wages as well as inflation. Furthermore, while the government rejected the Pension Commission's idea that the BSP be based on residence rather than contributions, there were significant changes to the criteria by which individuals would be eligible for the full BSP (Glennerster 2009: 152-3). First, the qualifying number of contributory years required to receive the full BSP was reduced to 30 for everyone, both men and women, retiring after 2010. Second, the definition of a contributory year was changed so that people, primarily women, who had spent years of their working aged lives in a caring role rather than doing paid work, would find those "caring" years count as contributory years towards the BSP (see Bozio, Crawford and Tetlow 2010: 18-22). According to government projections, this would mean that 95% of men and 75% of women reaching the state pension age in 2010 would qualify for the full BSP (Pensions Policy Institute 2011: 68). Flence, while still a contributory benefit, access to the full BSP was considerably eased. The importance of increasing the amount paid to seniors through the BSP is exaggerated by the fact that many eligible households, due to the complexity or stigma involved, do not claim their full entitlement to means-tested benefit. It is estimated that in 2009-10, the take-up of Pension Credit by eligible households was somewhere between 62% to 68% (Pensions Policy Institute 2012: 7).Countering these moves, and illustrating how policy can move in divergent directions simultaneously, there was retrenchment in the form of a gradual raising of the state pension age to 68. This regressive move was done in a way to minimize protest as it did not affect current or near retirees. Soon after taking office in 2010, the Coalition government reinforced the directions of travel outlined in the 2007 Act. From April 2012, the BSP was to rise in line with highest of average wages or inflation, or by 2.5%, but simultaneously the government also elevated the state pension age for men and women to 66.

Canada

The emergence of, and the debate over, OAS

In Canada, just like in the UK, pension reform became a significant policy issue in the immediate post-war era (Bryden 1974). (4) The central aspect of this post-war pension debate concerned the elimination of the means test created as part of the 1927 Old Age Pensions Act (Orloff 1993), a law that gradually led to the creation of federally subsidized old age assistance pensions in each of the nine provinces (Newfoundland only became a Canadian province in 1949). Due to provincial resistance especially strong in Quebec and despite the increase of the level of federal funding in 1931, it took nearly a decade for all provinces to offer these means-tested pensions (Banting 2005: 119-120). After the war, like in the UK, removing the means test became a priority, but there was no consensus about how to achieve this expansion of pension coverage. Two main policy options were discussed: earnings-related pensions and flat pensions. Inside the federal government, supporters of the second option came to dominate, which led to adoption of a universal pension scheme, in which entitlements would only derive from residency and age, as people aged 70 and older meeting certain residency criteria became entitled to a federal flat pension. Consequently, in contrast to the BSP in the UK, what become known as OAS was not a contributory scheme. Yet, to compensate for the additional fiscal burden created by the new flat pension, the Liberal government of Louis St. Laurent asked civil servants in the Department of Finance to design a new taxation scheme that, as explained below, took a different form than a traditional social insurance payroll tax (Bryden 1974: 122).

After securing provincial support for a constitutional amendment that allowed the federal government to directly enter this provincial jurisdiction (instead of subsidizing provincial pension schemes as it had done since 1927), parliament passed the Old Age Security Act in 1951, and the law came to effect on January 1 of the following year. In addition to the OAS program, which offered a $40-a-month pension to all people aged 70 and older meeting the residency requirement, a second legislation, the Old Age Assistance Act, also came to effect in January 1952. This law created a federal means-tested pension for people aged 65 to 69. This provision remained in place until 1970, when the age of eligibility for the OAS flat pension reached 65 (Beland and Myles 2005). This brief description of the program stresses a crucial difference with the UK's BSP: the OAS offered a truly universal pension not derived from one's contribution record. This is a key distinction, as the existing pension literature stresses the stronger political resilience of contributory schemes (Myles and Pierson 2001), which our comparison between the UK and Canada challenges, at least as far as flat pensions are concerned.

By the late 1950s, despite the creation of voluntary registered retirement savings plans (RRSPs) in 1957 and increases in OAS benefits aimed at compensating for the negative effects of inflation, it became clear that the limited coverage offered by voluntary occupational pensions and the modest nature of the OAS flat pension could not, on their own, guarantee the economic security of many older people (Babich and Beland 2009). To address this problem, in the mid-1960s, with support from the provinces, the federal government created an earnings-related pension scheme known as the Canada Pension Plan (or Quebec Pension Plan, a nearly identical program covering Quebec workers). To this day, the Canada Pension Plan (CPP) and the Quebec Pension Plan (QPP) offer a low replacement rate of 25% of covered wages. The modest nature of this replacement rate is related to the enduring belief that voluntary occupational pensions should remain a key source--and even the main source--of income security for large segments of the population (Babich and Beland 2009).

In addition to CPP and QPP, the mid-1960s also saw the advent of the Guaranteed Income Supplement (GIS), an income-related program created in 1967 as a temporary measure originally designed to expire when CPP/QPP matured as an established pension program (Bryden 1974; Guest 1997: 144-145). Working in tandem with OAS, this income-tested program helps explain why Canada has been quite successful in fighting old age poverty, and its positive impact helps explain why GIS was made permanent (Beland and Myles 2005). For instance, using the same definition of poverty, Michael Wiseman and Martynas Yeas (2008) showed that Canada has been much more successful than the UK and the US in fighting this type of poverty, a reality they attribute in large part to the central role of GIS, as it complements OAS and CPP/QPP. According to these authors, Canada does as well as social-democratic Sweden--and more than twice as well as the UK--in poverty reduction among older people (Wiseman and Yeas 2008).

The politics of retrenchment: From the indexation debate to the fiscal clawback

Like the UK, Canada witnessed a turn to social policy retrenchment in the 1980s. Although more moderate ideologically than the Thatcher government, Progressive Conservative Prime Minister Brian Mulroney (1984-1993) pledged to fight large budget deficits, which put potential OAS cutbacks on the agenda (Myles 1988). In the spring of 1985, less than a year after the election of his Progressive Conservative government, Mulroney pushed for the partial de-indexing of OAS benefits. More specifically, according to his plan, automatic OAS benefit increases would only compensate for yearly inflation above the 3% level. As Ken Battle (1997: 530-531) puts it, if implemented, this measure would have favoured a gradual decline in the real value of the OAS pension, in a fashion similar to what happened in the UK after 1980: "If inflation were 3 per cent or higher a year, then OAS benefits would automatically lose 3 per cent of their value. Even if inflation were less than 3 per cent, benefits would decline by the amount of inflation (e.g. an inflation rate of 2% would reduce the value of OAS by 2%)." In part because it appeared to contradict an electoral pledge about the need to preserve the integrity of OAS, this policy proposal was met with the largest wave of elderly protest in Canadian history. A major turning point was a televised and widely discussed confrontation between the Prime Minister and a tiny elderly woman from Quebec, who accused him of having lied to Canadians regarding OAS. With the help of extensive and typically negative media coverage, the Mulroney government was forced to withdraw its highly controversial proposal (Beland and Myles 2005). From this perspective, it failed where the Thatcher government had (surprisingly quietly) succeeded in slowly reducing the real value of the flat pension in order to control public spending. In this context, in Canada, the social and political mobilization of beneficiaries, who successfully organized mass rallies in Ottawa and received intense and sympathetic media coverage, played a direct role in the politics of universal flat pensions.

The 1985 political setback forced the Mulroney government to control the increase in OAS costs using indirect, and less visible, means (Pierson and Weaver 1993). In a low-profile 1989 budgetary move famously described by former federal civil servant Ken Battle (1990) as "social policy by stealth," the government implemented a fiscal device known as "clawback," which used the tax system to recuperate the OAS money paid to high income seniors. As a consequence of this under-the-radar policy change, benefits for people earning more than $51,765 per year were reduced by 15 per cent for every income dollar above such a threshold. Thus, people with an annual income above about $89,000 had to repay their OAS benefit in its entirety. Importantly, the cut-off point ($51,765) where the clawback took effect was only indexed to inflation above a 3% annual rate. This meant that, over time, more and more people would be negatively affected by the cut-off. Although this threshold was fully re-indexed in 2000, during the Liberal era, the 1989 fiscal change ended universality in an indirect and politically masterful way. Yet, importantly, the clawback only affects a small fraction of the elderly population, and the vast majority of seniors keep all the OAS money they receive (Beland and Myles 2005). From this perspective, the effectiveness of the clawback as a source of fiscal savings is rather modest, which helps explain why other OAS retrenchment attempts would take place during the following decades, in a context of accelerated population aging and enduring concerns about growth in OAS spending.

Rejecting income testing

The first major attempt to control costs actually involved the proposed replacement of OAS and GIS by a new income-tested scheme known as the Seniors Benefit. Put forward in the mid-1990s by Liberal Finance Minister Paul Martin (Government of Canada 1996), this proposal would have slightly increased benefits for low-income seniors while penalizing better-off seniors, including those in the middle class (Battle 1997; Geddes 1998). For this reason, the scheme would have reduced anticipated federal pension spending without penalizing the poor, which is probably why the Liberal Party supported this approach. Yet, the scheme faced strong criticisms from a variety of sources. On one front, some economists argued that the new scheme, through the implementation of an income test, would further reduce the incentives for older people to save. On another front, feminists like the leader of the New Democratic Party (NDP) Alexa McDonough described the Seniors Benefit proposal as an attack against the financial autonomy of Canadian women: "Senior women currently receive the OAS directly, but the Seniors Benefit Program will be calculated on a couple's combined income. Older women may lose their Seniors Benefit based on the income of their spouse, threatening their financial independence" (McDonough 1997). As a result of these attacks, as well as the more-rapid-than-expected advent of federal budget surpluses, in 1998, the Liberal government of Jean Chretien withdrew the controversial Seniors Benefit proposal (Beland and Myles 2005; Martin 1998).

This late 1990s episode was the last direct attempt to end the formal universality of OAS benefits, which remain available to people aged 65 and older meeting specific residency criteria (the clawback affecting wealthier seniors is applied after the benefits have been paid out). This is true because the recent attempt to control the rise in OAS costs stemming from population aging focuses on retirement age rather than income testing or even the extension of the fiscal clawback introduced in 1989. Announced in late March 2012 as part of the federal budget, these changes to OAS concern the graduate increase of retirement age, which should climb from 65 to 67 between 2023 and 2029 (Government of Canada 2012).

Like the gradual increase of Social Security eligibility age currently taking place in the United States (Light 1995; Pierson 1994), this decision to simply raise the eligibility age of future retirees reflects a strategy to reduce the potential political backlash against government. This is true because, starting in a decade, this increase in eligibility age will not affect current retirees or people near retirement, and these were exactly the two constituencies that mobilized so strongly against the 1985 Mulroney initiative on de-indexation, which had forced the Progressive Conservative government to withdraw its controversial proposal. In 2012, the Harper Conservative government went ahead with the unpopular change without having to face the massive wave of protest that had led to the humiliating demise of the Mulroney de-indexation proposal in 1985.

Nonetheless, the lack of protest should not obscure the regressive nature of a policy change that should penalize low-income workers. This is because these workers are unlikely to save enough to retire before they receive their modest OAS pension (maximum monthly benefit of $445, as of March 2012), something that better-off Canadians can do. The importance of this reality is exacerbated by the fact that occupational pensions cover a shrinking minority of the country's working population (Boychuk and Banting 2008).

Discussion

Several patterns emerge from a direct comparison between our two cases. First, the extent of cumulative policy change is much more pronounced in the UK than in Canada. This contrast in policy outcomes challenges the broad claim that contributory pension schemes are especially sturdy politically (Myles and Pierson 2001). In fact, the non-contributory OAS has proven to be much stronger politically than BSP, suggesting that lacking a contributory basis is not necessarily a political liability, even in an era of retrenchment and restructuring.

Second, to explain this contrast between BSP and OAS, we can turn to the power of social mobilization involving current and future beneficiaries, which was striking in Canada at the beginning of the Mulroney era but much more subdued during the early Thatcher years. Our goal here is not to explain why protest was stronger in Canada than in the UK at the time but to stress its impact on the politics of flat pensions. In Canada, the 1985 wave of protest forced the Mulroney government to withdraw its partial de-indexation program and to adopt a low-profile approach through enacting a fiscal "clawback" in 1989, but this modest change has not resulted in transformative policy change over time, in part because the Liberal government stopped its incremental expansion in 2000. In 2012, the Conservative government of Stephen Harper announced a gradual increase in retirement age that would not affect current and even near future retirees. Perhaps because of this decision to not affect current and near retirees, no 1985-style elderly protest took place in the aftermath of this announcement. Overall, the Canadian case suggests strongly that social protest on the part of beneficiaries can shape the politics of social policy change, which is consistent with some of the recent literature on pension reform in advanced industrial countries (Beland and Marier 2006; Campbell 2003).

Third, our article suggests that adopting a long-term time frame is necessary to assess the issue of incremental change. For instance, in the immediate aftermath of the 1989 creation of the fiscal clawback in Canada, "social policy by stealth" (that is, enacting significant changes through less-visible means) seemed to have been a stunning success in bringing about low-profile but nonetheless transformative policy change to the field of universal pensions (Battle 1990). Yet, in light of more recent trends, our comparative study suggests that incremental change has proved more powerful and sustainable in the UK than in Canada, despite the central role of obfuscation and incremental tactics in the latter country. These remarks point to the crucial importance of the time frame we use to analyse, assess, and even explain incremental policy change and its impact on the life of beneficiaries as well as on public finance (Campbell 2004).

Finally, our analysis of the two cases suggest that the development of flat pensions in the UK and Canada is a complex and multifaceted story that is not about the unilateral decline of universality and the unstoppable rise of retrenchment. This is particularly the case in the UK, where efforts have been made since the Blair years to improve the protection offered by the most basic elements of the UK public pension systems. This complex story undermines the traditional dichotomy between expansion and retrenchment, which is supposed to characterize two distinct historical eras (Pierson 1994). The 2007 Pensions Act in the UK simultaneously expanded the universality of the BSP, while enacting retrenchment by increasing the retirement age. Our analysis shows that contemporary social policy reform is not only about retrenchment, despite the undeniable fact that cost control has become a key social policy imperative in both countries during the era of accelerated demographic aging.

References

Babich, Kristina, and Daniel Beland. 2009. "Policy change and the politics of ideas: The emergence of the Canada/Quebec pension plans." Canadian Review of Sociology 46 (3): 253-271.

Banting, Keith G. 2005. "Canada: Nation-building in a federal welfare state." In Federalism and the Welfare State, edited by Herbert Obinger, Stephan Leibfried and Frank G. Castles. Cambridge: Cambridge University Press, pp. 89-137.

Barr, Nicholas. 2004. The Economics of the Welfare State. Oxford: Oxford University Press.

Battle, Ken (under the pseudonym of Grattan Gray). 1990. "Social policy by stealth." Policy Options 11 (2): 17-29.

Battle, Ken. 1997. "Pension reform in Canada." Canadian Journal of Aging 16 (3): 519-552.

BBC News. 2012. "Pensions auto-enrolment starts for biggest firms," BBC News online, October 1. Available at http://www.bbc.co.uk/news/business-19760421

Beland, Daniel, and Patrik Marier. 2012. "Vieillissement et politiques de retraite au Canada" in Patrik Marier (ed.), Le vieillissement de la population et les politiques publiques: enjeux d'ici et d'ailleurs. Quebec City: Presses de l'Universite Laval, pp. 109-128.

Beland, Daniel, and Patrik Marier. 2006. "Protest avoidance: Labor mobilization and social policy reform in France." Mobilization 11 (3): 297-311.

Beland, Daniel, and John Myles. 2005. "Stasis amidst change: Canadian pension reform in an age of retrenchment." In Ageing and Pension Reform around the World, edited by Giuliano Bonoli and Toshimitsu Shinkawa. Cheltenham: Edward Elgar Publishing, pp. 252-272.

Beland, Daniel, and Alex Waddan. 2012. The Politics of Policy Change: Welfare, Medicare and Social Security Reform in the United States. Washington DC: Georgetown University Press.

Beveridge, William H. 1942. Social Insurance and Allied Services. Cmd6404, London: HMSO.

Blackburn, Robin. 2002. Banking on Death Or, Investing in Life: The History and Future of Pensions. London: Verso.

Blake, David. 2004. "Contracting out of the state pension system: The British experience of carrots and sticks". Rethinking the Welfare State: The Political Economy of Pension reform, edited by Martin Rein and Winfried Schmal. Cheltenham, UK: Edward Elgar, pp. 19-55.

Bonoli, Giuliano. 2000. The Politics of Pension Reform: Institutions and Policy Change in Western Europe. Cambridge University Press.

Boychuk, Gerard W., and Keith G. Banting. 2008. "The public-private divide: Health insurance and pensions in Canada." In Public and Private Social Policy: Health and Pension Policies in a New Era, edited by Daniel Beland and Brian Gran. Basingstoke: Palgrave Macmillan, pp. 92-122.

Bozio, Antoine, Rowena Crawford, and Gemma Tetlow. 2010. The History of State Pensions in the UK: 1948 to 2010. London: Institute for Fiscal Studies.

Brooks, Richard, and John Denham. 2005. The Politics of Pension Reform. London: Fabian Society.

Bryden, Kenneth. 1974. Old Age Pensions and Policy-Making in Canada. Montreal: McGill-Queen's University Press.

Campbell, Andrea L. 2003. How Policies Make Citizens: Senior Political Activism and the American Welfare State. Princeton: Princeton University Press

Campbell, John L. 2004. Institutional Change and Globalization. Princeton: Princeton University Press.

Department of Work and Pensions. 2006. Security in Retirement: Towards a New Pensions System. London. Available at http://www.dwp.gov.uk/docs/white-paper-summary.pdf

Department of Work and Pensions. 2012. The Annual Abstract of Statistics for Benefits, National Insurance Contributions, and Indices of Prices and Earnings 2011 Edition. London. Available at http://research.dwp.gov.uk/asd/asdl/abstract/abstract2011.pdf

Evandrou, Maria, and Jane Falkingham. 2009. "Pensions and income security in later life." In Towards a More Equal Society? Poverty, Inequality and Policy since 1997, edited by John Hills, Tom Sefton and Kitty Stewart. Bristol, Policy Press, pp. 157-76.

Geddes, John. 1998. "Citizen's Revolt," Maclean's, 111 (20 April): 12-13.

Glennerster, Howard. 2009. Understanding the Finance of Welfare: What welfare costs and how to pay for it. Bristol: The Policy Press

Government of Canada. 2012. Budget 2012. Available at http://www.budget.gc.ca/2012/ themes/theme3-fra.html

Government of Canada. 1996. Government Proposes New Seniors Benefit. Ottawa: Government of Canada (March 6).

Guest, Dennis. 1997 The Emergence of Social Security in Canada (third edition). Vancouver: UBC Press.

Hacker, Jacob. 2004. "Privatizing risk without privatizing the welfare state: The hidden politics of welfare state retrenchment in the United States." American Political Science Review 98: 243-260.

Light, Paul C. 1995. Still Artful Work: The Continuing Politics of Social Security Reform. New York: McGraw-Hill.

Marier, Patrik. 2008. Pension Politics: Consensus and Social Conflict in Ageing Societies. London: Routledge.

Martin, Paul. 1998. Finance Minister's Statement on the Seniors Benefit. Ottawa: Government of Canada (July 28).

McDonough, Alexa. 1997. "McDonough defends pensioners," The New Ontario Democrat. December.

Myles, John. 1988. "Social policies for the elderly in Canada." In North American Elders: U.S. and Canadian Comparisons, edited by Betty Havens and Eloise Rathbone-McCuan. New York: Greenwood Press, pp. 37-54.

Myles, John, and Paul Pierson. 2001. "The comparative political economy of pension reform. In The New Politics of the Welfare State, edited by Paul Pierson. New York: Oxford University Press, pp. 305-333.

Orloff, Ann Shola 1993. The Politics of Pensions: A Comparative Analysis of Canada, Great Britain and the United States, 1880-1940. Madison: University of Wisconsin Press.

Pemberton, Hugh. 2006. "Politics and pensions in post war Britain," History and Policy. Available at http://www.historyandpolicy.org/papers/policy-paper-41.html

Pensions Commission. 2004. Pensions: Challenges and Choices. London: Pensions Commission.

Pensions Commission. 2005. A New Pensions Settlement for the Twenty-First Century. London: Pensions Commission.

Pensions Policy Institute. 2011. An assessment of the government's options for state pension reform, pensionspolicyinstitute.org.uk

Pensions Policy Institute. 2012. Pensions Facts, September 2012. Available at http:// www.pensionspolicyinstitute.org.uk

Pierson, Paul. 1994. Dismantling the Welfare State? Reagan, Thatcher and the Politics of Retrenchment. Cambridge: Cambridge University Press.

Pierson, Paul, and R. Kent Weaver. 1993. "Imposing Losses in Pension Policy." Do Institutions Matter?: Government Capabilities in the United States and Abroad, edited by R. Kent Weaver and Rockman, Bert A. Washington, DC: Brookings Institution, pp. 110-150.

Pratt, Henry J. 1993. Gray Agendas: Interest Groups and Public Pensions in Canada, Britain and the United States. Ann Arbor: The University of Michigan Press.

Taylor-Gooby, Peter. 2005. "UK pension reform: A test case for the liberal welfare state?" In Ageing and Pension Reform Around the World: Evidence from Eleven Countries, edited by Giuliano Bonoli and Toshimitsu Shinkawa. Cheltenham: Edward Elgar.

Streeck, Wolfgang, and Kathleen Thelen, eds. 2005. Beyond Continuity: Institutional Change in Advanced Political Economies. Oxford: Oxford University Press.

Thelen, Kathleen. 2004. How Institutions Evolve: The Political Economy of Skills in Germany, Britain, the United States, and Japan. Cambridge: Cambridge University Press.

Timmins, Nicholas. 1996. The Five Giants: A Biography of the Welfare State, London: Fontana Press.

Waine, Barbara. 2009. "New Labour and pension reform: Security in retirement." Social Policy and Administration 43 (7): 754-771.

Weaver, R. Kent. 2010. "Paths and forks or chutes and ladders? Negative feedbacks and policy regime change." Journal of Public Policy 30 (2): 137-162.

Wiseman, Michael, and Martynas Yeas. 2008. "The Canadian Safety Net for the Elderly," Social Security Bulletin, 68 (2): 53-67.

Notes

(1) In October 2012, a new workplace pension scheme was introduced in the UK, with employees automatically enrolled in a pension plan unless they explicitly opt out. Starting with larger employers, this scheme will be expanded to cover all workers earning over [pounds sterling] 8,105 a year by 2018 (BBC News 2012).

(2) In the UK, the distinction between state pensions and public sector pensions is essential. Public sector pensions in the UK are occupational pensions for workers employed by central or local governments, such as teachers or doctors. The Coalition government implemented reforms to public sector pensions in 2010 that resulted in a number of strikes, but these reforms are not the subject of this paper.

(3) The Conservatives won a plurality of votes amongst voters aged over 65 in all of their general election triumphs from 1979 to 1992 (Blackburn 2002: 291).

(4) For a recent, historically-informed overview of the Canadian pension system, see Beland and Marier 2012.

Daniel Beland is Professor and Canada Research Chair in Public Policy (Tier 1), Johnson-Shoyama Graduate School of Public Policy. Alex Waddan is Senior Lecturer in American Politics in the Department of Politics and International Relations, University of Leicester. A previous draft of this article was presented at the annual Social Security History Conference in Vancouver (November 2012). The authors thank the participants at that conference as well as Greg Marchildon and the anonymous reviewers for their comments and suggestions. They also thank Tanya Andrusieczko for her help in preparing the article. Daniel Beland also acknowledges support from the Canada Research Chairs program.