The Rise and Fall of Adult Literacy: Policy Lessons from Canada

There was a period of time, from the late 1980s until the early/mid-2000s, when interest in adult literacy in Canada was strong among the public, in the media, and with policymakers, and a policy window opened for the mainstreaming of literacy. Against this background, it is surprising that the Canadian literacy infrastructure was subsequently largely dismantled. Drawing on theories of policy formation, and recent and previous research, including interviews with key stakeholders, we argue that mainstreaming literacy has failed and explore the reasons for this failure. The paper is structured in three sections. First, we report on the construction of an adult literacy infrastructure in Canada over two phases: i) the period from the 1970s up until the launch of the International Adult Literacy Survey (IALS) in 1994; ii) the story of IALS and changes occurring up until around 2005. Second, we examine the reasons for the failure of the mainstreaming of literacy in Canada. We conclude with further reflections on the present situation in which adult literacy has been largely reduced to employability skills which are under-supported.1


Introduction
Beginning in the late 1980s, adult literacy emerged as a policy issue in several industrialised countries including Canada (Barton & Hamilton, 1990). This is in large part due to a changing definition and understanding of literacy as no longer a binary construct of literates versus illiterates, but as existing on a continuum, as a contextual social practice (Street, 1984), and as related to surviving and thriving within a changing workplace and society. The first adult literacy surveys in Canada, which included the Southam newspaper report Broken Words, published in 1987, followed by the report En toutes lettres et en français, published by the Institut canadien d'éducation des adultes (ICÉA) for the francophone population, pointed to poor literacy skills among the adult population in the country and provoked public debates and policy responses. Canada is a particularly interesting case, as, at one point, interest in literacy was strong among the public, in the media, and with policy-makers. Both Conservative and Liberal federal governments have been supportive of adult literacy over the years. Canada was a driving force behind the 1994 International Adult Literacy Survey (IALS), conducted in collaboration with the Organisation for Economic Co-operation and Development (OECD), which furthered the adult literacy agenda in the country. Between the 1970s and 1990s, considerable infrastructure was built up for literacy in the form of national and provincial organisations. However, the Canadian adult literacy infrastructure has since been largely dismantled.
This paper provides new insights into the Canadian literacy story. It is based on recent and previous research, including nine semi-structured interviews with key stakeholders who played a part in the story, comprising government officials, experts and academics, and NGO representatives involved in literacy-related policy-making, research, and advocacy work in Canada. The purpose of these interviews was to better understand both the creation of Canada's role in the IALS as well as the history of adult literacy in the country. We also engaged in analysis of secondary data which laid out the historical development of Canadian adult literacy (e.g., Darville, 1992). Thirdly, we draw on data from our previous research on Canadian adult literacy policy (Rubenson & Walker, 2011;Walker, 2008;Walker & Rubenson, 2014). Conceptually, the article is framed by the concept of "mainstreaming" literacy, which will be further explained below, and by theories of policy formation, in particular Kingdon's (2003) theories of "policy windows" and "policy entrepreneurs" (see also Béland & Howlett, 2016). We argue that mainstreaming literacy has failed in Canada and explore the reasons for this failure in three sections. First, we report on the construction of an adult literacy infrastructure in Canada over two phases: i) the period from the 1970s up until the launch of the International Adult Literacy Survey (IALS) in 1994; ii) the story of IALS and changes occurring up until around 2005. Second, we examine the reasons for the failure of the mainstreaming of literacy in Canada. Finally, we reflect on the present situation, in which adult literacy has been largely reduced to employability skills which are undersupported.
Canada's experience with adult literacy is of interest to other countries. Canada is important as a pioneer of adult literacy policies and measurement, especially given its crucial role in the creation of the OECD's IALS and subsequent surveys, which includes the current Programme for International Assessment of Adult Competencies (PIAAC). As we show in this paper, Canada has been unable to support a robust system of adult literacy programming and policy for a number of reasons, which offer valuable lessons about policy-making in adult literacy and more generally, especially for federated states.

Understanding mainstreaming adult literacy in Canada
Throughout the industrialised world, adult education has long been lamented as the "poor cousin" (Newman, 1979) to compulsory schooling, and has clamoured to capture government attention as a policy issue. In Canada, as in other Anglo-Saxon liberal welfare states, such as the UK, Australia or the United States, adult education has long been associated with basic education for poor people with low literacy skills. Accordingly, as veteran Canadian adult literacy researcher and practitioner, Allan Quigley (1990), has long noted, it has been stigmatised. Literacy practitioners have been associated with volunteer do-gooder grannies in cardigans, rather than professional teachers, and adult literacy has, by and large, existed outside the mainstream of education and its learners outside what is generally understood as the mainstream of society.
"Mainstreaming" literacy is about bringing literacy in from the periphery to the centre of both education and social policy. We are using the term here in three related ways. First, it refers to bringing literacy into the mainstream of social policy, so it occupies a similar space in public policy to compulsory schooling or other social policies (i.e., it is no longer an afterthought, the poor cousin, or lying outside principal government concerns). Second, it means institutionalising adult literacy policy so it links with, and infuses into, other policies and connects with other related government bodies (e.g., in the areas of housing, health, correctional services, employment etc.). Third, the term also refers to bringing adult literacy more squarely into, or embedding it in, existing vocational, language, and skills curricula in a contextualised manner (Conway, Lopez, & Casey, 2007). The hope is that adult literacy, and adult education more broadly, will no longer be thought of as existing for a small and marginalised subset of society, or as separate from other educational programming, and that its connection to all other social policy questions will be recognised. This idea of "mainstreaming" was expressed by several of our interviewees: So, the question of literacy training at the end of the day isn't or shouldn't be independent of a whole range of essential skills. And it shouldn't be independent of teamwork skills. It shouldn't be independent of language training for immigrants…The correct policy answer… is to move it [literacy] into the mainstream. (Interview with EI 2 , Former Assistant Deputy Minister in the Canadian Federal Government) I would say that [mainstreaming literacy means] it needs to be not off on its own, sort of a renegade. It has to find its place within a suite of activities and look like the other programmes and act like the other programmes. (Interview with RA, former programme director in the National Literacy Secretariat, Canadian Federal Government) So it's very important that [literacy] be institutionalised, to be connected with everything. So it doesn't fall between the cracks. (Interview with AR, former high-level officer in Human Resources and Skills Development, Canadian Federal Government) On the one hand, mainstreaming could be a positive direction in terms of integrating literacy into policies that support learning and social development. For example, mainstreaming is part of the strategy of the United Nations to enable the implementation of the Sustainable Development Goals, where it refers to helping governments 'to land and contextualise the agenda 3 at national and local levels; ultimately reflecting the agenda in national plans, strategies and budgets' (UNDP, 2017). At the same time, mainstreaming literacy has also led to moving literacy away from the community or regional bodies, resulting in a narrowing of literacy in content and purpose, with attendant deleterious effects. For example, previous research on New Zealand's success in developing a national literacy strategy showed how mainstreaming literacy resulted in the creation of a workplace, employment agenda, while undervalorising community literacy organisations (Walker, 2011). Furthermore, in the case of New Zealand, such mainstreaming was accompanied by an intensification of administrative work and bureaucracy, increasing competition for funds, and a disregarding of difficult-to-quantify literacy outcomes. One of our interviewees used the term "institutionalise" in terms of establishing greater bureaucratic control over literacy. In our view, mainstreaming can be a double-edged sword.
A policy window for mainstreaming literacy in Canada?
From the late 1980s until the early/mid-2000s, a "policy window" existed for mainstreaming literacy in Canada. According to Kingdon (2003), policy windows open when the separate streams of problems, policies, and politics 'are joined at critical junctures' (p. 227). The "problem stream" relates to the public perception of a problem that demands a policy response. With heightened attention to a problem, the "policy stream" can then open up, in which experts examine problems and propose policy solutions. Finally, the "political stream" connects to political will and 'comprises factors that influence the body politic, such as swings in national mood, executive or legislative turnover, and interest group advocacy campaigns' (Béland & Howlett, 2016, p. 222). Indeed, with these three streams flowing freely into one another, adult literacy had the ear of the government, national bodies existed, and literacy practitioners were optimistic that a pan-Canadian literacy initiative was possible.
Nonetheless, unlike places like New Zealand, Canada never really succeeded at mainstreaming, professionalising, or institutionalising adult literacy. This is particularly curious given the tradition of adult education in Canada and its global influence (see Draper & English, 2013), the high standard of its formal education system, and the chief role Canada played in the creation of the OECD's major international adult literacy surveys -specifically, the IALS and Adult Literacy and Lifeskills (ALL) surveys, on which the PISA (Programme for International Student Assessment) and PIAAC surveys build. Yet, since 2006, most of the literacy organisations and research institutes have become defunct, core funding for programming has diminished, and the literacy community remains disillusioned (Smythe, 2018).
Canadian adult literacy practitioners, researchers, and literacy advocates likely had reservations at what mainstreaming literacy could entail; there is, as Addey (2018) explained, a danger of a single story in monolithic and hegemonic interpretations of literacy emanating from the OECD, and in a culture of measuring literacy and comparatively ranking countries in their achievements. At the same time, raising the profile of literacy has been a continuing desire expressed by many practitioners and supportive government officials alike. After presenting a brief chronology of adult literacy in the country from the 1970s until today, we put forward theories for why the promise of mainstreaming literacy in Canada has not been realised.

Adult literacy in Canada from the 1970s to 2006
Canada is well known by adult education scholars throughout the world for its pioneering adult literacy achievements. The Antigonish Movement in Nova Scotia remains an inspiration for the development of cooperatives in its asset-based approach to community development and teaching of literacy (Selman & Dampier, 1991). Frontier College is another exemplary model of adult literacy provision: originating in the 1800s in university extension by sending students to remote communities to help teach literacy to lumber workers, the college continues to offer numerous literacy and language programmes to Canadians, new immigrants, and refugees (p. 56). For the purposes of this paper, however, we start by examining the national state of adult literacy from the 1970s up until this day. This is because there was arguably no Canadian "literacy movement" or much discussion of problems with adult literacy prior to the 1970s (Atkinson, 2019;Hautecoeur, 2001).
From the 1970s to pre-IALS: Building an infrastructure for adult literacy In the mid to late 1970s, an infrastructure around adult literacy began to be built and the question of adult literacy started to gain federal government attention. We would argue that this constitutes the beginnings of attempts to mainstream literacy; though at this time, the focus was much more on community literacy than anything else. An incipient network was formed between practitioners and the few researchers publishing on adult literacy, which ultimately became the Movement for Canadian Literacy (MCL), established in 1977 (Draper & Carere, 1998, p. 69). MCL subsequently developed a coalition of 10 literacy organisations across the country, called the Canadian Alliance for Literacy, that later released the high profile publication, A Call to Action on Literacy, which was disseminated throughout national media (Shohet, 2001). As Hautecoeur (2001) writes, 'the Movement for Canadian Literacy acquired an almost monopolistic legitimacy in the provinces and with the Federal Government' (p. 413). In addition, the national government started to commission research on adult basic education for the labour force, and the Canadian UNESCO Commission convened, for the first time, a working group to examine literacy in Canada. Québec was arguably key in growing an adult literacy sector and interest. This is, in some ways, not surprising given the historical and ongoing concern in the province to retain its culture and language, as well as its autonomy within -if not separation from -the Canadian state. In particular, the Institut canadien d'éducation des adultes (ICÉA), formed in 1946 to bring together civil society French language adult education organisations in Canada, started to examine questions of adult literacy from a Freirean perspective; its 'ALPHA' publication series on literacy and basic education research was launched in 1978, and was committed to supporting "literacy awareness" in French and Creole across the world (Hautecoeur, 2001).
By the 1980s, many policymakers, researchers, and literacy organisations were no longer satisfied in equating years of schooling with skills and literacy (Jones, 1990). In 1986 (results released in 1987), Canada conducted its first national survey of literacy skills. The Southam Survey, commissioned by the eponymous newspaper chain, examined Canadians' ability to complete "everyday literacy tasks", defining literacy in a way that was adopted almost verbatim in IALS: 'using printed and written information to function in society, to achieve one's goals, and to develop one's knowledge and potential' (Darville, 1992, p. 13). The survey found that 38% of Canadians were below the literacy level deemed adequate for succeeding in society (Calamai, 1987). According to one of our interviewees, 'it caused a political problem for the government. They said 24% of the Canadian adult population can't be functionally illiterate' (Interview with CU, a methodological expert involved in IALS). In response to the Southam Survey, in 1989, the federal government commissioned Statistics Canada to carry out the LSUDA survey-Literacy Used in Daily Activities. The precursor to IALS, LSUDA measured Canadians' reading, writing, numeracy and information processing skills across five levels.
During the mid-late 1980s, the federal government started to pay serious attention to literacy. Thanks to lobbying by MCL and other literacy advocates, and to the worrying results of the Southam survey (see Hautecoeur, 2001;Rubenson & Walker, 2011), a funding commitment to literacy was made in 1987 by the Conservative Prime Minister, Brian Mulroney, to create the National Literacy Secretariat (NLS) which then formed part of his re-election platform. The NLS was created 'with a mandate to work with provinces, the private sector, and voluntary organizations to develop resources to ensure that Canadians had access to the required literacy skills' (Hayes, 2013, p. 4). As Darville (1992) noted, 'in 1988…for the first time, the platforms of the political parties included substantial plans to mobilise governmental programmes to respond to the literacy issue' (p. 7). At this time, the link between the economy and literacy was being made strongly in media and by politicians (Walker & Rubenson, 2014), which coincided with the time around which the OECD began publishing reports on the topic (Atkinson, 2019). Following the establishment of the NLS, federal monies were put towards the creation of three additional national bodies of adult literacy: The National Adult Literacy Database (NALD) (1989), ABC Literacy Canada (1989); and Fédération canadienne pour l'alphabétisation en français (FCAF) (1989). Furthermore, 'UNESCO's International Literacy Year of 1990 gave literacy a new visibility in Canada and, in part, prompted the government to lend greater financial and moral support to [literacy]' (Rubenson & Walker, 2011, p. 3). By the beginning of the 1990s, there were six national adult literacy organisations, all but one created between 1977-1989, as shown in the table below. From all appearances, and for the first time ever, literacy advocates, researchers, and practitioners were no longer outside the mainstream.

IALS in Canada: A "focusing event"
We would argue that IALS could be considered a "focusing event" that opened and sustained a "window of opportunity" (Kingdon, 2003) (Draper & Carere, 1998). Based on the results of those earlier studies, the 1987 Southam newspaper survey, and the LSUDA study of 1989, Canadian NGOs, as well as other public advocates such as the Canadian journalist and broadcaster Peter Gzowski, were lobbying the federal Department of the Secretary of State to invest in a broader study. The department was interested, and sought to involve the OECD which 'could provide substantive analytical oversight and international credibility' (Interview with DI, senior official in the Department of the Secretary of State). Another reason to involve the OECD was to avoid working with the provinces, which hold jurisdiction for education in Canada: at the time, the provinces resisted publicising any findings on literacy problems in Canada (Interviews with AO, former director of a leading Canadian adult education NGO; and with CU, a methodological expert involved in IALS). The IALS study was then conducted cooperatively between Statistics Canada and the OECD. The expertise for the study came from Canada and the American Educational Testing Service (ETS), building on the same team of statisticians that had already worked on the LSUDA study.
The OECD was in charge of the overall coordination, recruiting countries, and planning and framing the reports and products that came out of IALS. The first IALS survey, published in 1994, was conducted in Canada, France, Germany, the Netherlands, Poland, Sweden, Switzerland, and the United States. The funding for the study was shared by these countries, with the United States being the most important financial contributor to the development of the methodology (Interview with CU). IALS examined literacy (broken into sub-components of prose and document) and numeracy. Later versions of the study (i.e., ALL and PIAAC) also looked at additional areas, such as life skills and problem-solving using technology. According to IALS, 42.2 percent of Canadians were estimated to be in the two lowest levels of the prose scale (out of 5 levels) (OECD & Statistics Canada, 1995). Level 3 was considered the minimum for a person to be able to function adequately in society. The results were widely debated in the media and policy circles, contrary to other countries such as Germany, where the IALS results were equally alarming, but never discussed, or France, which rejected the results and withdrew from the study (Thorn, 2009). For the next decade, IALS greatly contributed to advancing the literacy agenda in Canada. The funding available through the NLS (that had also funded the Canadian contribution to IALS) helped to build up a literacy infrastructure, with provincial organisations being created across the country (Interviews with RA, former programme director in the National Literacy Secretariat). In the years after the publication of the IALS study, the budget of the NLS was increased (Shohet, 2001). According to a former staff member, the NLS was "really golden" at that time: 'We had all our international literacy activity, we had all the start up stuff, things were starting to snow ball and then it rolled into IALS and then that added momentum' (Interview with RA). Prominent literacy advocates, who acted as "policy entrepreneurs", defined by Kingdon (2003) as 'people willing to invest their resources in return for future policies they favor' (p. 204), added to that momentum. For example, the creation of the NLS was enabled by two Ministers, David Crombie and Lucien Bouchard, and two media personalities, Peter Gzowski and Peter Calamai, who had personal reasons to become champions of literacy, such as experiences of illiteracy in the family (Interview with DI, senior official in the Department of the Secretary of State). Kingdon (2003) describes how 'subjects become prominent agenda items…because important policy makers have personal experiences that bring the subject to their attention' (p. 96). Another key advocate for literacy was Senator Joyce Fairbairn. When IALS was published, she was leader of the government in the Senate and Minister with Special Responsibility for Literacy. She had been one of the driving forces behind the creation of the NLS and used her extensive political influence to lobby for literacy. Her influence cannot be underestimated: 'What triggered the strong government [response to IALS] was Senator Fairbairn; she was a force of nature' (Interview with former federal government official EO). In 2003, a follow-up study to IALS was conducted, the Adult Literacy and Life Skills Survey (ALL). The goal of the survey, in which 12 countries participated -seven in the first round 2002-2003, five in the second in 2006 (Thorn, 2009) -was to measure progress since IALS. The ALL results showed very little difference compared to IALS (Rubenson & Walker, 2011). The Canadian data that were published in 2005 underpinned the launch of the "literacy and essential skills" agenda that was part of the Canadian government's Workplace Skills Strategy announced in December 2004 (Jackson, 2005). The shift to "essential skills" signalled a move away from the citizenship and collective dimensions of literacy and largely reduced literacy to individual skills required for the job market.
In 2005, Claudette Bradshaw, who was Minister of State (Human Resources Development) and, like Fairbairn, an advocate for literacy, set up the federal Advisory Committee on Literacy and Essential Skills (the Bradshaw Committee), which put forth a broad vision for a national literacy strategy that was backed up by a commitment on the part of key actors in the federal government to increase its dedicated $28 million in annual spending on literacy by $30 million over three years (Hayes, 2013). But nothing came of the momentum nor of the activities set in motion by the literacy advocates in the NLS and the federal government. In the years from around 2005 onwards, we can see a clear shift in the policy approach to literacy.

to present: The dismantling of the Canadian literacy infrastructure
It is fair to conclude that the election of the Conservative government in 2006 obliterated the progress towards mainstreaming literacy in Canada. Months after its election, the Harper government announced it would cut $17.7 million in funding to adult literacy, effectively dismantling the NLS. In spite of a budget surplus, Conservative MP John Baird lent support to his government's decision, announcing 'I think if we're spending $20 million and we have one out of seven folks in the country that are functionally illiterate, we've got to fix the ground floor problem and not be trying to do repair work after the fact' (quoted in Delacourt, 2006). The shift in adult literacy policy can only partially be attributed to the newly elected Conservative government, however (Hayes, 2013). Indeed, it had already started in the 1990s with institutional changes in the federal government (Smythe, 2018), which prevented the building of solid foundations necessary for the mainstreaming of literacy. In 1993 the NLS, originally housed in the Department of the Secretary of State (which had responsibility for citizenship), was transferred to Human Resources Development Canada (HRDC), tying literacy to the labour market (Hayes, 2013 (Hayes, 2009). While the literacy strategy pursued by the NLS was communitybased and inspired by a view of literacy as a driver of social and personal development, the HR(S)DC's interest in literacy focused exclusively on employment (St. Clair, 2016): In about 2002…all of a sudden you know we're literacy and essential skills and we're having to partner with more of the employment side of the department. Where do you put family literacy activities and where do you put community and that started to be a big rub in that didn't fit with the mandate. (Interview with RA) Even before a Conservative government was elected in 2006, there was a group of civil servants in the HRDC that actively worked against greater investments in literacy. As EO, former federal government official said in our interview, 'the late 90s I would say is when there might have been this shift from the bureaucratic side…to institutionalise literacy.' There was tension between the people working at the NLS and the "bureaucrats" in the HRDC: 'Because of their [people working in the NLS] style of doing business within the government, HRDC people…were pulling their hair out' (Interview with EO). When it came to implementing the recommendations of the Bradshaw Committee, 'the bureaucrats stalled and they stalled and they stalled until there was an election that brought in the Conservatives' (Interview with EO). According to a former NLS staff member, the NLS had been distributing funding to a range of non-governmental organisations, provincial organisations, and community groups, in a rather unbureaucratic and collaborative manner. As former NLS programme director RA said in our interview, 'at the time, the NLS was all about partnerships.' That changed in 1999 with the scandal that ensued after an HRDC audit that condemned the management of grants and contribution funds. The audit of the NLS was part of a larger, albeit likely poorly executed, audit of all government grants and contributions (Sutherland, 2001). Several of the NLS files were deemed problematic by the auditors, who criticised poor documentation and claimed missing funds (Hayes, 2009). In some ways, the problems plaguing the NLS were not entirely external. "Bureaucratic sloppiness," appears to have been an issue, with awards made without applications and a lack of faithful recording of evidence of consultation (Sutherland, 2001). Nonetheless, the uproar that followed was likely overblown, and the changes that occurred were arguably not warranted (Sutherland, 2001): As it turned out there wasn't any real money missing by the end of it, it was just not all the paper was in the file, but by the time everybody figured that out it was too late and so the whole mechanism changed. (Interview with RA) The audit led to New Public Management reforms, a management philosophy in which governments seek to make the public sector more efficient and cost-effective by increasing privatisation and instituting market-oriented reforms (Boston, Martin, Pallot, & Walsh, 1996). In the case of Canada, the Liberal government at the time introduced greater accountability measures, and tightened bureaucratic processes, such as the use of computer systems (Interview with RA). Before the audit scandal, the NLS had engaged in collaborative development of proposals with literacy organisations across the country. This way of working became much more difficult after the audit, when competitive calls for proposals were introduced. According to RA: we [the NLS] had to have calls for proposals and we weren't allowed to talk to anybody if the proposals came through and…the whole atmosphere changed…the leadership at the NLS at that point shifted to just really managing grants and contributions.
As discussed by Elfert and Rubenson (2013), these new bureaucratic arrangements 'resulted in a transformation of the relationship between the federal government and the provinces, in which the provinces [were] no longer partners but clients' (p. 225).
Although previous developments had weakened literacy's foundations, the Conservative Harper government arguably gave literacy the final blow as it was at odds with its social policy ideology. Not only was its 'social policy…based on an ideology and practice of limiting and restructuring the role of the federal government in social affairs' (Prince, 2015, p. 57), but the Harper government upheld an 'institutional-punitive conception of social policy' (p. 54), involving 'the discipline and punishment of deviant individuals' (p. 60). The dismantling of the NLS in 2007 and whittling away of support for literacy reached a culmination in 2014 and 2015 when all national literacy organisations were defunded. Jason Kenney, Minister of Employment and Skills Development (as the HRSDC was renamed), declared in 2014: Our government is committed to ensuring that federal funding for literacy is no longer spent on administration and countless research papers, but instead is invested in projects that result in Canadians receiving the literacy skills they need to obtain jobs. (cited in Smythe, 2015, p. 16) This reduction of literacy to skills for the job market is represented by the shift from literacy to essential skills. The HRSDC's nine essential skill areas, subdivided in five levels of complexity that can be tested through a workplace skills test called TOWES, derive from the IALS and ALL methodology (Jackson, 2005;Pinsent-Johnson, 2011;Smythe, 2015). OLES made applying for grants more competitive and bureaucratically cumbersome, which has had a devastating effect on community organisations in Canada. Furthermore, OLES has failed to apportion the little funding it has (Hayes, 2018); as noted by one of our interviewees, only 50% of OLES allocated funding was spent in 2017.

The failure of mainstreaming literacy in Canada
As shown above, the streams of problems, policies, and politics crossed to some extent in Canada between the mid-late 1980s and approximately 2005. The IALS survey was greatly analysed and followed up upon in Canada through investments and the creation of institutions promoting literacy as a policy issue and delivering literacy programmes. "Policy entrepreneurs", including high-level politicians, had created a favourable political climate for literacy.
However, the policy interest was short-lived. In our view, the policy window was only ever partially open and with a moderate breeze, easily blew shut. Indeed, the notion of mainstreaming adult literacy in Canada was always fragile and subject to the vicissitudes of government. Despite a strong adult education tradition in the country, there has never been adequate long-standing infrastructure. Bégin, Eggleston and MacDonald (2009), cited in Smythe (2018, p. 141), describe Canada as 'a country of perpetual pilot projects.' Adult literacy lies under the shadow of international surveys and measurements-which presents a contradiction to the sociocultural understanding of literacy as a plural and dynamic social and cultural practice 'with different literacies according to the different domains of life and defined by the individual and wider community goals and cultural practices' (Addey, 2018, p. 317, drawing on Barton, Hamilton, & Ivanic, 2000). Starting with the Southam newspaper survey, then continuing with LSUDA, leading eventually to IALS, and reinforced by the media, literacy has been constructed as a measurable and standardised skill that a person either has or does not have. Level 3 was determined as the threshold, dividing those who are literate and those who are not. It was used to frame literacy in a way that it 'was no longer about citizenship, empowerment, motivated training and being learner-centred. It was about moving people to "Level 3"' (Hayes, 2009, p. 22).
The Southam survey, IALS and ALL were double-edged swords: They helped bring literacy to the attention of media and policy makers but at a cost. The Southam survey constructed literacy as a national crisis, based on criteria to sort the literate from the illiterate that were quite arbitrary: 'In arriving at suitable figures for Canada's literacy crisis, the Creative Research Group…set a single test score against the accomplishments of an education and a career' (Willinsky, 1990, p. 9). The media that covered the Southam survey, 'across the front pages of 26 Canadian newspapers' (p. 1), identified 'Newfoundlanders, immigrants and workers' (p. 6) as the groups that Prince (2015) refers to as "deviant" or "dangerous" in social policy. French Canadians and the Indigenous population were also more likely to have literacy scores below level 3. The IALS survey has been widely criticised as contributing to the construction of literacy as the "single story" (Addey, 2018), a "project of social ordering" (Hamilton, 2001), serving literacy as a "competitiveness project" (Darville, 1999) and making illiteracy "a national sickness" (Hautecoeur, 2001, p. 411). For some of the experts who worked on the IALS, "level 3" constituted 'a line in the data that's absolutely clear' in so far as 'that group of people [levels 1 and 2] face a disproportionate share of the risk' (Interview with CU). At the same time, others criticised level 3 as it negated the very notion of literacy existing as a social practice. According to another statistician involved in IALS, the construction of "level 3" as the "watershed" of functional literacy put 'a label on [people], as inadequate' and did not appropriately reflect the competences of people, noting that 'level two people are a lot more capable than the rhetoric around the results would suggest they were' (Interview with TO, a statistical expert involved in IALS and previous Canadian literacy surveys; see also St. Clair, 2012).
Ultimately, the "single story" damaged literacy in Canada as it contributed to stigmatising people. Literacy learners were "framed" by placing them in categories and referred to as '"level ones", "level twos"' (Smythe, 2015, p. 9). Funding was invested towards raising people to level 3, neglecting those with the lowest literacy levels. A report by the Conference Board of Canada (2013), a Canadian think tank, argued: Moving this group [those currently at level 2] up to a solid level 3 -considered to be the minimum 'job standard' level that enables employees to cope with the demands of workwould be less expensive and involve fewer resources, per capita, than moving the group of employees with extremely rudimentary level 1 literacy skills up to level 3. (cited in Smythe, 2015, p. 11) As the follow-up studies to IALS did not yield significant measurable improvements, there was no clear solution for the literacy "problem" at hand. According to Kingdon (2003, p. 178), when the "alternative" as a solution is not coupled to a problem, the policy window closes. Furthermore, the political climate was shifting towards an accountability and outcome-oriented approach to policy-making. Rather than 'situating literacy within the context of full citizenship' (Hayes, 2009, p. 19), literacy efforts were increasingly measured against 'tangible delivery outcomes such as the number of jobs created and the number of people employed ' (p. 20). Another effect of the focus on measureable results was that many of the interesting findings of the IALS and ALL surveys were hardly discussed, such as the enormous difference in literacy skills found in Québec between the generation prior and after the quiet revolution of the 1960s, a period of intense social and political transformation in Québec when the Catholic church lost its tight grip on the population which was becoming more educated. One of our research participants who is very familiar with the data was particularly excited about the results from Québec: You could see in the data when the change took place, because there was such a difference in the people that had been educated before 1960, and those that had been educated after 1960…if you want proof that the education system in Québec got so much better, we have it there, in the results in the study. So things like that is what gives you confidence that you've actually tapped into something that's real. (Interview with TO) The studies also revealed new information about how adults learn that was largely ignored because it was not deemed politically interesting. In the words of TO: there's a lot more information in the results that anybody ever really made use of because most of the rhetoric and policy that came out of it focused on how many people are in a particular level. Which meant a lot of…educationally significant information never really got into the public policy.
The focus on the economic argument for literacy led to its downfall as a political issue, as evidence for economic benefits of adult literacy are difficult to quantify within the timeframe of an election cycle. Moreover, there are outcomes of adult literacy education that cannot be accounted for in narrow assessments, such as people developing the ability to make phone calls, to show up for work on time, etc., which are not considered relevant. In other words, IALS could have been used to mainstream literacy in a way which allows for multiple narratives, but Canada chose a single story. Understandably, the NGOs and community organisations used the IALS numbers to lobby for more funding. At the same time, the literacy community became disillusioned with the effects of the big data from IALS. As former NLS programme director RA said: 'What started as…an awareness, a population level glimpse at what people could do, became…all pervasive…by the time PIAAC came along things had really gone sideways in terms of how people in Canada viewed IALS.' The increasing disconnects between certain "bureaucrats" and adult literacy learners and practitioners that some of our interviewees referred to were exacerbated by institutional reforms in public management. While the period of the "high time" of the NLS was characterised by partnership-oriented and collaborative relationships between the federal government and literacy organisations and stakeholders, the bureaucratic reorganisation of government structures furthered the separation between the policy level and on-the-ground literacy learners and practitioners. There has been a trend of increasingly professionalised career bureaucrats who move from one unit to another without any expertise and no background in education. Somewhat at odds with Kingdon's (2003) analysis that 'elected officials and their appointees turn out to be more important than career civil servants' (p. 199), we found that the "hidden participants", as Kingdon calls them -middle-level civil servants -had significant leeway in resisting particular policy attention to literacy, even against the will of elected politicians.
The federated nature of Canada constitutes another challenge to integrating literacy into the mainstream of education. Policy processes in the field of adult education differ from those in relation to schooling. While the provinces have a clear mandate for schools, the responsibility for adult education is spread across sectors, more complex in its delivery and linked to labour market policies (Elfert & Rubenson, 2013). Québec is the only province that has mainstreamed literacy to some extent. Since 2001, Québec has a Government Policy on Adult Education and Continuing Education and Training (Gouvernement du Québec, 2002), which differs from those in other provinces in that it emphasises a rights-based approach to adult education and the responsibility of the state in providing adult education opportunities. Although this policy has long been neglected due to changes of government, Québec has set a system in place in which the school board offers adult education provision. This integration of literacy in the formal education structure is unique in Canada, as in the other provinces literacy and basic education provision tends to be more ad hoc and diffuse and divided between different ministries. Québec also has a unique policy in place, the Bill 90 of 1995, the "Act to Foster the Development of Manpower Training", also referred to as "the law of 1 %", which aims at improving workers' skills through training and development, to promote employment in general, and to foster employment adaptation and employment integration. The law stipulates that every employer whose total payroll for a calendar year exceeds $1,000,000 CAD is required to participate in workforce skill development by allotting an amount representing at least 1% of his or her total payroll to eligible training expenditures (Charest, 2007).
As previously mentioned, another reason why adult literacy has failed to be mainstreamed in Canada is that literacy is associated with poverty, stigmatised groups, and with adults who "made poor choices" (Quigley, 1990). In other words, people outside the "mainstream" of society. Unlike children, adults have no appeal to innocence and so are blamed for their educational "failures", particularly in Western liberal societies dominated by current neoliberal frameworks. As one of our interviewees said, discussing a (successful) pilot project with single mothers that was never expanded, 'single moms on welfare are lazy, undeserving citizens. That's the rhetorical structure' (Interview with CU). Given the cuts and policies enacted under the Harper government, one might conclude that adult literacy learners were viewed as undeserving recipients of social policy.

Further exploring the present situation
Adult literacy policies and programmes have fallen out of favour in Canada. It is important to note that the most recent study of adults' literacy skills, the 2013 PIAAC, which built on the IALS and ALL surveys, has abandoned the concept of "literacy" -as have most public policies -in favour of "skills" and "competencies". Most institutional structures of adult literacy in Canada have now been dismantled-so, even with a slightly more sympathetic government as the current Liberal government may very well be, the effort it would take to mainstream literacy would be more than they care to invest. Smythe (2018), in a chapter about the closure of the National Adult Literacy Database, writes about the infrastructure that cannot be easily replaced once it is gone. As one of her interviewees, Sue Emson, said, 'I don't know if the knowledge from the field is still out there. This is the problem of the infrastructure that has been lost' (p. 188).
Against this background, it is not surprising that PIAAC, according to St. Clair (2016), has had no policy impact in Canada so far. One of our interviewees stated that 'PIAAC died in Canada four days after it was released' (Interview with RA). This is not unlike what has happened elsewhere, such as Denmark (Cort & Larson, 2015). The reasons, according to one of our interviewees, are "political" -news about low adult literacy skills is not favourable to "getting re-elected" (Interview with CU). Another interviewee pointed to the lack of federal leadership and "policy entrepreneurs": 'There was nobody out there promoting it [PIAAC]…nobody was the flag bearer on the file' (Interview with RA). Priorities have also shifted towards compulsory education for the Indigenous population: '…the federal government in Canada is under a lot more pressure to put its educational interest into Indigenous education…for children, which is just a disaster in this country' (Interview with TO). As Steiner-Khamsi (2004) points out, 'the most frequent response to OECD-type studies is indifference. In fact, in most countries, comparative and international studies pass unnoticed by politicians or the general public and cause little excitement-positive or negative' (p. 208). This is not to say that PIAAC might not have a more indirect policy influence down the track, as one of our interviewees suggested: In many countries, it is seen as one of the major data sources that you can use when you want to justify certain directions that you go, it will kind of play out differently in different countries, depending on the policy context" (Interview with JU, an academic who was involved in IALS).
IALS fell on fertile ground because it capitalised on a "window of opportunity". 'There was money' and 'a number of people from civil society, a few public servants and a few people in leadership roles felt that this was something to do' (Interview with DI). "Policy entrepreneurs", such as politicians, the media, and the public alike were interested at the time to get to the bottom of the literacy problem and there was hope that the data could be used to implement reforms that would benefit employers and workers. As Steiner-Khamsi (2004) argued: 'The potential of influencing educational reform depends on whether a controversy over educational reforms already exists-attractive if at that particular time policymakers are in need of additional external support for an already existent agenda' (p. 208). UNESCO's International Literacy Year in 1990, a general drive for data, and a more unbureaucratic way of governing were part of this policy window. Actors among the Canadian NGOs, and in the federal government, Statistics Canada, the OECD, and academia, formed a policy network that pushed for IALS, albeit with different motivations. It is interesting to note that the initial drive for more data about literacy came from the Canadian NGOs with 'the educationally disadvantaged adult' (Interview with AO, former director of a leading Canadian adult education NGO) in mind. They lobbied for IALS, but then the file moved to the federal government and the OECD. PIAAC has now been taken over solely by the OECD as part of the 'PISA engine' (Interview with CU, methodological expert involved in IALS). Ultimately, IALS was used by the federal government to underpin the employment-oriented Essential Skills agenda, which is 'arguably not in relation to instruction and learning at all, but rather in relation to assessment and screening' (Elfert & Rubenson, 2013, p. 227). The same happened with PISA, which is built on the empirical design of IALS. As stated by Lundgren (2011), 'the outcomes of PISA we hoped could stimulate a debate on learning outcomes not only from an educational perspective but also a broad cultural and social perspective. Rarely has a pious hope been so dashed ' (p. 27). In that respect, IALS paved the way for PISA, as it offered many insights into social issues in Canada, as illustrated by the data from Québec, which were not taken up (St. Clair, 2012).
After the "golden years", when the data showed no quick improvements and the political economy changed towards bureaucratisation of governance and less NGO and civil society influence, the lack of a robust literacy infrastructure enabled the dismantling of literacy in Canada in a relatively short time. As one of our interviewees said: 'We went from the real high of being totally engaged down to like nobody even knew it happened' (Interview with RA). We might think about adult literacy in Canada as a straw house that was easily blown down by a few acts of government rather than solid infrastructure that could withstand changing political whims. Although certain policy entrepreneurs were important, their achievements were not sustainable.
What we can learn from the Canadian case is that only a robust infrastructure can survive the vicissitudes of governments and bureaucracies. It is also needed to avoid the constant reinventing of the wheel in educational policy-making (Smythe, 2018). A new promising development in Canada in the last couple of years has been the creation of a federal Future Skills research and advocacy centre which was formed as a response to the challenges facing us in the "Fourth Industrial Revolution" with the rapid developments in artificial intelligence and ensuing job automation and labour market upheaval (Schwab, 2016). But if literacy had been mainstreamed to a greater extent, this new initiative, such as other new initiatives that are bound to be launched by future governments, could build on past achievements, even if new themes and new priorities come along.