Comparative,Feminist,and Qualitative: An Uncommon Perspective on Cross-National Social Policy Research

Defining cross-national social policy research

There are various types of comparative research such as comparisons within countries, across time, between genders and age categories, and across states within the same nation (Clasen, 1999; May, 1998). Cross-national comparative research can happen at either the micro or the macro level. At the micro level, individual programs are compared usually within nations with similar political and social structures. Macro comparisons are attempts to examine whole “systems” over a range of countries and across time. Deacon, Hulse, and Stubbs (1997) have maintained that social policy research should focus on supranational or global rather than comparative contexts. However comparative studies that acknowledge global complexities, for example governance at multiple levels, are highly relevant and valid: “[It is] only through analysis incorporating a number of interrelating levels, macro and micro, [that] an understanding of the complex processes taking place can be understood and the diversity of differences between nation-states can be highlighted” (Kennett, 2001, p. 38).

There are four approaches to cross-national research. First, theoretical studies “attempt to explore, and to explain, the differences between the different welfare systems of different countries and to assess the extent to which they are the result of internal policy making or external dynamics” (Alcock, 2001, p. 5). The second approach focuses on particular sectors across nations, such as social security, child care policy, housing, and health; the third, on evaluating policy effectiveness across states. Finally, the fourth approach centers on country comparisons, involving comparisons of welfare provisions in selected countries. There are also three categories of cross-national research. The first category entails hypothesis testing on a large scale using quantitative methods across many countries, although in such studies, depth might be sacrificed for breadth. Micro studies, the second approach, rely on qualitative data gathering techniques and “emphasise cultural sensitivity and specificity, agency and reflexivity in the policy research process” (Kennett, 2001, p. 7). The final approach focuses on regime theory and makes use of welfare state typologies.

Rationale for undertaking cross-national social policy research

Social policy research has tended to focus on the policy and welfare contexts of individual countries. This is a reflection of the reality that welfare policies have been developed in particular national contexts (Alcock & Craig, 2001). However this situation is changing, as “international forces are increasingly shaping the policy agendas of national governments, so that welfare provision is less and less the product of national policy debates or political considerations” (Alcock, 2001, p. 4). Pressures on policy making need to be viewed within the context of globalization and its accompanying burgeoning capitalist economy, whereby international agencies are seeking to control social policy in individual countries. Examples of these agencies and organizations are the European Union, the World Bank, the International Monetary Fund and transnational companies (Alcock, 2001, Kennett, 2001; May, 1998). Countries are also proactively engaged in examining the policies of other nations with the possible intention of transferring policies into their own contexts: “Social, cultural and economic manifestations are imported and exported across borders” (Kennett, 2001, pp. 1–2). These comparisons are also opportunities to learn from the mistakes of others.

There is a need for large-scale comparative research as well as micro-level cross-national studies that are contextually rich. Countries are growing more interdependent, and the challenges they face are similar, although their social policy responses have not necessarily been the same; “the need for better understanding of common features and crucial differences between not only individual welfare states but particular policy programs in order to unravel why and how welfare needs, or demands are being transformed into social policy” (Clasen, 1999, p. 4). Comparative cross-national research provides a tool for analyzing the impact of differing policy regimes (Fincher & Saunders, 2001).

Recently there has been a renewed feminist focus on welfare state research in a comparative context: “Crucial to this reorientation have been feminist critiques of mainstream analysis of welfare states and the combining of feminist and comparative perspectives” (Sainsbury, 1996, p. 1). The focus of feminist scholarship has moved from the complexity of particular welfare systems to scholarship interested in variation between countries. Dominelli (1991) has also critiqued single-state policy research, arguing that nation states are not autonomous entities and noting the tendency to ignore issues of class, gender, and race. It is important to recognize that comparative social policy research has tended to be gender blind and that “feminists have begun to rework orthodox comparative typologies” (Poole, 2000, p. 186).

Methodological issues and challenges

The methodological issues and challenges of cross-national research are worth examining. This type of research does not have the same methodological implications as other types of social research: Challenges are compounded when research is undertaken across borders. Oyen (cited in Kennett, 2001), asserted, “to advance our knowledge about cross-national research it is necessary to raise questions about the distinctive characteristics of comparative studies” (p. 5).

Central to this type of research are issues of definition and concept. A major challenge for cross-national comparative policy research is conceptualizing the welfare state and/or social policy contexts of different nations. These conceptualizations are highly contested within the comparative social policy field. Defining what you are comparing is fundamental to effective comparative research and is especially challenging in the social policy context. Esping-Andersen (1990), in The Three Worlds of Welfare Capitalism, argued that not all welfare states are the same and defined, at a broad level, three types of welfare state regimes: liberal, corporatist, and social democratic. Daly (2000), a feminist sociologist, defined social policy regimes as “the systematic set of arrangements governing the relations between politics and markets” (p. 5) and critiqued regime theory as gender blind, maintaining that the private sphere has been marginalized. Welfare families is a preferred term, acknowledging a more complex conceptualization of the welfare state and arguing her preference for a move away from an overreliance on ideal types and “from large typologies and binary matrices in favour of research less concerned with neatness of fit and more with the messy and stubborn practices encountered in social reality” (p. 12). Contemporary authors have also claimed that welfare “regimes” are little more than illusions and that the complexities of welfare provision are not captured in such broad categorizations (Kasza, 2002).

The emphasis should be on the importance of the ideologies concerned with families and gender, and creating gender regimes: “A gender policy regime entails a logic based on the rules and norms about gender relations that influences the construction of policies” (Sainsbury, 1999, p. 5). This provides the feminist researcher with a conceptual framework for critiquing welfare policy that places gender relations at the centre of policy analysis. The nexus between state, family, and market is crucial to feminist theorizing in a cross-national context.

At a micro level, defining what you are comparing is also critical. One of the greatest challenges of cross-national comparative research “is to understand the idiosyncrasies of national conditions [and] the conceptual frameworks of the actors” (Mabbett & Boldersen, 1999, p. 52). Social policy interventions are interpreted and understood differently in different countries and are informed by distinctive intellectual traditions. The underlying assumptions about “social need” tend to be highly contested; cultural assumptions might not be the same. An issue of major policy debate in one country might be of no policy significance in another. Specific words can even have different meanings and connotations. National data might, therefore, differ because they reflect different ways of conceptualizing problems and constructing social issues; different categories are used for generating statistics based on different assumptions, definitions, and government priorities (Kennett, 2001; Mabbett & Bolderson, 1999; May, 1998). Empirical comparability relies on comparing social indicator data across countries; “there is no guarantee that such data has been gathered and published on a similar basis in each country—and indeed [it is] a much greater likelihood that it has not” (Alcock & Craig, 2001, p. 6). Gathering your own data is therefore advantageous, although the costs involved in generating your own qualitative or quantitative data mean that researchers often have to rely on official sources of data.

Comparison in the cross-national context is not a controlled experiment, as it might be in the natural sciences. However, equivalence of concepts is important if you are to have “confidence that the components and their properties being compared are the same or indicate something equivalent” (Kennett, 2001, p. 44). Cross-national comparative research should be designed to have comparative conceptualizations and methods; concepts form the basis of relating one country to another. A potential problem in this regard is the researcher's lack of familiarity with a country other than his or her own, which might lead to confusion and misrepresentation of important features. The researcher should strive for linguistic, measurement, and conceptual equivalence, all of which require a deep knowledge of the compared countries' social policy contexts.

Doing feminist, qualitative, cross-national, comparative research

What defined my research experience was a set of circumstances that came together to provide an opportunity to undertake research in my second study site, California (my first was Australia). These circumstances included an opportunity to travel to the research site and my past work experience in the Californian child care field, which aided in access to informants, study sites, and respondents. Mabbett and Boldersen (1999) have described the process of going into another country to collect data as a “research safari” (p. 52). I undertook such a safari when I traveled to California to undertake part of my doctoral research. The prospect of returning to a place where I had previously worked in child care service planning, and pursuing research in this area, was an exciting opportunity.

A two-country framework has merit. The smaller comparative framework provides greater detail and focuses “attention on the complexities of small-scale as well as large scale variations” (Daly, 2000, p. 12). In addition, there is less need to rely on ideal types, and micro-level outcomes can be explored. The two-country approach also provided the basis for a more manageable qualitative research project. It is also worth reminding the cross-national researcher “on-site visits and face-to-face interviews … are especially resource intensive in the cross-national context” (Mabbitt & Boldersen, 1999, p. 51).

The first task in the cross-national research process is to identify the type of comparative research being undertaken. Comparative research on child care policies is micro-level comparative research, defined as individual program comparison in nations with similar social structures (May, 1998). My study focused on a particular sector (child care) as opposed to a broader theoretical or country comparison study (Alcock, 2001). This study also fitted well with Kennett's (2001) second approach to cross-national research, which is a micro study relying on qualitative data-gathering techniques.

Valid comparison is at the heart of cross-national comparative research. Identifying what you are comparing is a fundamental challenge. Initially, this could involve identifying the type of welfare state regimes America and Australia are: Both are liberal welfare regimes (Esping-Andersen, 1990). Daly (2000) and Kasza (2002) have argued, however, that this regime labeling ignores the complexities of welfare state provision. Therefore, consistent with Kasza's recommendation “that policy-specific comparisons may be a more promising avenue for comparative research” (p. 271), my focus was on defining and comparing two child care subsidy policies. Both countries have formulated policies and programs that aim to subsidize the cost of child care. It was most useful for me to define the Californian subsidy services initially as residual, because subsidies are narrowly targeted to people with the lowest incomes. I also initially defined Australian child care subsidy as a semi-universal service, as the subsidy is available to all users of approved child care on a sliding scale (Graycar & Jamrozik, 1993).

Cross-national comparative research also requires methodological equivalence (Kennett, 2001). Selecting study sites that can be meaningfully compared is central to establishing this methodological equivalence. Child care service delivery systems vary greatly between California and Australia. In Australia the federal government administers child care subsidy payments in a standardized way to all eligible parents. In contrast Meyers, Heintze, and Wolf (2002) have defined the Californian state-funded child care subsidy system as complex: “Assistance was provided through a variety of mechanisms, including direct service and vendor agreements with non-profit centres” (p 167). Determining study sites in these vastly differing service delivery contexts, with the need for methodological equivalence in mind, meant defining the elements of subsidy delivery system that would provide the basis of comparison. I chose the following site levels; service receiving, service delivery, and policy making. These sites were applicable to both the Australian and Californian child care systems. I drew my respondents from each of these sites. I then selected a data collection strategy that was identical in both California and Australia. I chose semistructured interviews as my primary data-gathering method in both locations.

To summarize, as a qualitative cross-national researcher, I had as my research design tasks to identify the type of comparative research I was undertaking (micro level), define the social policy contexts I was comparing (residual and semi-universal), and develop a research design that met the requirements of methodological equivalency regarding research site, respondent selection, and data-gathering method chosen.

Outcomes and benefits of feminist, cross-national social policy research

My decision to compare child care policies in California and Australia was the result of my work experiences in both locations. This particular cross-national comparison also made sense because child care policies are a site of policy difference in otherwise similar “welfare regime” contexts (although similar gender regimes). In addition, the cross-national construction of the study was consistent with the arguments of feminist theorists, who have referred to the renewed focus on welfare state research in a comparative context: “Crucial to this reorientation have been feminist critiques of mainstream of welfare states and the combining of feminist and comparative perspectives” (Sainsbury, 1996, p. 1). Furthermore, this type of research is important in a context where countries are proactively engaged in examining the social policies of other nations with the intention of transferring them into their own context. For example, much of Australian current welfare reform initiatives are modeled on the United States' Welfare Reform programs.

My research led me to conclude that qualitative cross-national comparative research is a powerful tool for identifying policy impact in specific policy contexts. Comparing one policy context with another highlighted what was unique about a particular child care policy. For example, the comparative process foregrounded the Californian women's experiences of stigma and unworthiness associated with their use of a narrowly targeted residual subsidy service. In comparison the Australian women did not report any stigma or sense of unworthiness attached to their use of the semi-universal Australian subsidy service. It therefore made sense to conclude that the differences in women's experiences related to the differences in their respective child care subsidy programs rather than the attributes of individual respondents. Comparative research can therefore challenge conservative assumptions about “normal” social patterns that are ascribed to the qualities of individuals and highlight disadvantage linked to specific types of policy formulation.

Feminist theorizing was also central to my cross-national comparative study. Sainsbury (1999) has maintained, “the underlying concern in feminist comparative research has been the conceptualization of gender-relevant dimensions of variation” (p. 4). In this study, I used a qualitative methodology to place women's experiences and activities at the center of my analysis—placing “gender relations centre stage” (pp. 5–6). The cross-national comparison, centralizing the experiences and activities of women, highlighted the necessity of a strong feminist engagement with child care policy. Furthermore, my study emphasized the importance of critical engagement with foreign social policy initiatives intended for cross-national transfer.

However, my learning from this uncommon research endeavor was not only theoretical but also practical. Undertaking qualitative cross-national research was, indeed, a resource-intensive process that required a sustained and in-depth engagement with each research site. Recognition of the substantial commitment is, I believe, fundamental to the success of this approach. Kennett (2001) advocated a deep understanding of culture and context. What is unsaid in Kennett's discussion is the value and necessity of meaningful sustained engagement with the field to gain this understanding of culture and context. The implications of combining a sound qualitative approach with cross-national comparative research should be articulated. The reward of culturally and context rich data, gathered from in-depth conversations with women, time spent in agencies and with policy makers in Australia and the United States, would not have been possible without a commitment to this sustained engagement.

In this article I have defined cross-national comparative research and provided a rationale for undertaking this uncommon qualitative research endeavor. I have identified concept definition as a primary methodological challenge, particularly related to the theoretical conceptualization of welfare states and social policy contexts. The challenge of micro-level cross-national comparative research is to understand the idiosyncrasies of particular policy environments (Mabbett & Bolderson, 1999). Mazer and Parry (1998) discussed the benefits of undertaking qualitative research in cross-national contexts, arguing that qualitative analysis can reflect the “complexities of cultural contexts” (p. 384), although they lack, according to Mazer and Perry, the generalizability advantage provided by quantitative designs. However, I argue that the contextual richness provided by qualitative cross-national social policy research adds significantly to the benefits of a generalizable quantitative approach, and powerfully to our understanding of the complex ways social policies (across the globe) shape women's lives.