UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Taking class seriously : alternatives to the 'income paradigm' Haworth, Byron 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


24-ubc_2016_november_haworth_byron.pdf [ 645.37kB ]
JSON: 24-1.0314218.json
JSON-LD: 24-1.0314218-ld.json
RDF/XML (Pretty): 24-1.0314218-rdf.xml
RDF/JSON: 24-1.0314218-rdf.json
Turtle: 24-1.0314218-turtle.txt
N-Triples: 24-1.0314218-rdf-ntriples.txt
Original Record: 24-1.0314218-source.json
Full Text

Full Text

TAKING CLASS SERIOUSLY: ALTERNATIVES TO THE ‘INCOME PARADIGM’   by  Byron Haworth  B.A., University of Washington, 2011   A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF   MASTER OF ARTS in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Political Science)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  September 2016   © Byron Haworth, 2016    ii  Abstract  The recent surge in literature on economic inequality in the United States has provided new insights into the political ramifications of growing economic stratification. This literature has also revealed particular methodological standards by which such inequality is measured and understood. In lieu of more robust, sociological conceptions of class, political scientists largely conceive of group-based economic inequality in terms of continuous proxy variables, such as income. This paper addresses the underlying logic and motivations of this practice, sometimes referred to as the ‘income paradigm.’ The paper is divided into two sections; in the first, prevailing income-based research strategies are evaluated in terms of their generic and specific advantages and disadvantages. I argue that income is a less productive measure in describing the nature of actual inequality than is frequently assumed, and suffers from problems of causal ambiguity in the explication of satisfying and comprehensive causal explanations. In the second section, I propose three alternative conceptualizations of class developed in the sociological tradition of class analysis. These class models, derived from Weberian, Marxist and contemporary empirical traditions, are advantageous over proxy-based standards because of their a priori theoretical robustness. I argue that political scientists should transgress the disciplinary boundaries which have implicitly prohibited the use of class-based methods.           iii  Preface   This thesis is an original, unpublished, independent work by the author, Byron Haworth.                             iv  Table of Contents   Abstract…………………………………………………………………………………ii Preface…………………………………………………………………………………..iii Table of Contents………………………………………………………………………iv List of Tables …………………………………………………………………………...v List of Figures………………………………………………………………………….vi Acknowledgements…………………………………………………………………….vii 1. Introduction………………………………………………………………………….1 2. Income and Class……………………………………………………………………4  2.1 Generic features of income measure…………………………………………….4  2.2 Causality and indeterminacy: problems of income-based explanations…..10 3. Sociological Alternatives to the Income Paradigm……………………………....16  3.1 Political science and sociology: historical cleavages……………………….16  3.2 Erikson-Goldthorpe class model……………………………………………….21  3.3 Wright’s neo-Marxist class model……………………………………………..24  3.4 Microclass model………………………………………………………………...28 4. Conclusion…………………………………………………………………………32 References……………………………………………………………………………34           v  List of Tables    Table 1: Erikson-Goldthorpe Class Typology………………………………………22                             vi  List of Figures   Figure 1: Wright Class Typology…………………………………………………….25                         vii  Acknowledgements    I would like to thank my supervisor, Dr. Lisa Sundstrom, for her guidance throughout this process. Her comments were immensely useful in shaping the paper and instilling some much needed self-confidence. I also wish to thank my examiner, Dr. Paul Quirk, for his engaged and productive commentary on the finished draft.   I offer my sincerest gratitude to the political science faculty at UBC, from whom I’ve learned so much over the past year, and to the graduate students and friends from around the world who have enriched and broadened my horizons. And especially to my family, and anyone who withstood the neuroses involved with writing this paper, and returned only love and support.                   1. Introduction  American society is complex. The cherished, idealized image of the US as a middle-class nation in which “most people live more or less the same kind of life” (Krugman 1996) has been largely abandoned by social scientists, who have recognized that capturing an adequate picture of American life requires the ability to distinguish Americans across a large variety of categories, including race, religion, and ethnic origin. In recent years, observers of American society have increasingly recognized the central importance of economic distinctions as well. It is widely accepted that the US is less economically equal than it was three decades ago, and that the American political economy is riddled with “categorical mechanisms that produce unequal distributions of material, symbolic, and emotional resources which together combine to yield extreme stratification” (Massey 2007, XVI). Economic inequality concerns every social science discipline, but contemporary assumptions recognize that inequality cannot be understood holistically through a single, unified approach. Individual disciplines within social science are now seen as functional niches, each charged with explaining a particular piece of the larger puzzle of inequality. Consequently, each field is left with the task of adopting standards of research that best complement their particular mission.  Political scientists have engaged an aggressive research agenda into the deleterious effects of rising income inequality in the United States. In 2004, the American Political Science Association published a landmark report on economic inequality and American democracy. The report saw an assemblage of top political scientists speaking with one voice, and their conclusions were dire: economic inequality in the US has stalled or reversed democratic progress, undermined the proper functioning of democratic institutions, and represents an existential threat to the democratic ideals of American society (APSA 2004, Jacobs and Skocpol 2005).  The 2008 financial crisis has instilled yet greater urgency in the eyes of researchers, and has drawn the gaze of policymakers and the public alike. Recent years have therefore seen an unprecedented spike in studies on the political effects of inequality, offering striking new insights into how economic advantage has radically affected economic policymaking (Hacker and Pierson 2010), civic and economic opportunities (Putnam 2010), political participation, and citizen  2  influence (Brady and Schlozman 2012, Baumgartner et al. 2009), and the responsiveness of representative institutions (Bartels 2010, Gilens 2012).  Though this flurry of activity has undoubtedly greatly advanced our knowledge about the political consequences of economic inequality, it has also highlighted important problems in how political scientists traditionally conceptualize such inequality. Political scientists frequently embrace methodological strategies where a privileged variable such as income serves as a vehicle to describe larger forms of structured economic inequality. This strategy conforms to the individualist and pluralist traditions of political science, but is increasingly inadequate as a means of analyzing economic inequality. Recent findings by political scientists suggest a vision of American inequality that is more structured and comprehensive than previously thought, and extends beyond what can be captured through differences between income groups. This methodological approach, labeled the “income paradigm” by some inequality scholars (Grusky and Ku 2008), is functionally suited to the domain of economists but is largely insufficient in addressing the complex questions studied by political scientists. Income measures conceal patterns of inequality which are relevant to political scientists, contain measurement complications which are frequently overlooked, and contribute to analyses find compelling relationships in data but produce very thin, causally incomplete explanations.  This paper addresses the problem of using income-based measures as the foundation of political science research into economic inequality. I argue that income-based measures are used by political scientists almost reflexively, with little thought to the underlying assumptions that accompany them. A careful examination of income reveals a number of underlying problems, both in measurement and in its tendency to privilege particular kinds of analyses over other. In place of an income-centric approach, I argue that political scientists should instead turn to sociological accounts of class. This involves transgressing disciplinary boundaries and a strong, implicit “taboo” on the use of class. However, such a step is necessary in allowing political scientists to fully engage with the comprehensive, structured nature of inequality in the US.  I begin by elaborating the problem of income-based approaches by way of illustration, using an influential recent income-based study by Gilens to demonstrate the strengths and drawbacks of the current approach. I then discuss the underlying logic of  3  income, and its generic strengths and weaknesses. I provide a brief theoretical and historical account of the differences between political science and sociological approaches to class and inequality. Lastly. I discuss three approaches to studying class developed in the sociological tradition which might serve as alternatives to income-based measures of class. These three methods, each from a distinct theoretical tradition, can serve to enhance political science’s understanding of economic inequality by respectively emphasizing how common economic life chances produce similar political interests, how classes can describe important social antagonisms, and how class operates at a micro-level to produce homogenous groupings.                         4  2. Income and Class  2.1 Generic features of income measures  In this section, I discuss the “income paradigm,” the approach currently favored by many political scientists studying class. The use of income variables by social scientists developed gradually over time, as industrialization brought an increasingly large share of individuals into modern productive relationships (Cowell 2009). To be sure, income offers a compelling means of assessing an individual’s social welfare. Social scientists revere income data for its ubiquity and reliability. In the US, debates over how to properly distribute incomes forms a decisive criterion of distinction between conservative and progressive attitudes. Income is an inarguably crucial attribute of modern society. However, the deployment of income as a generalized proxy variable to describe social stratification in general is considerably more controversial. Income is a more volatile measure than is often considered. Furthermore, as a unidimensional measure of stratification, income is not particularly descriptive. This ambiguity means income groups not only fail to adequately describe relevant differences between social groupings, but may also conceal distinctions between people with very different life chances.   A recent prominent study by Gilens (2012) is illustrative of the strengths and drawbacks of the income-based approach to class. Gilens finds that when the preferences of high income and low income Americans diverge, the former are dramatically more likely to see their preferences enacted. This suggests that democratic responsiveness to lower income Americans is almost entirely incidental, and Gilens estimates that the poor and middle classes have virtually no independent capacity to influence their representatives. Such findings have led Gilens and Page (2014) to stake out an even more aggressive conclusion: that the weight of the evidence of class bias suggests a system of sustained domination by economic elites that should overturn the pluralist consensus which has guided much of political science.    Though the Gilens-Page thesis has already attracted considerable attention from scholars, policymakers (Bernstein 2014), and the press, and has been widely criticized (e.g. Enns 2015, Branham, Soroka and Wlezien forthcoming), none have thus far taken  5  direct issue with their use of income groups to simulate social class, suggesting that income-based explanations are seen as widely legitimate within the greater scholarship on inequality. This is clearly problematic, as is evidenced upon a closer review of Gilens and Page’s assertion of elite domination.  At first blush, the authors’ findings do seemingly break through the pluralist prohibition of ‘ruling elite’ explanations as famously established by Dahl’s (1958) hypothetical test.1 Gilens and Page present clear evidence of instances where the preferences of the rich diverge around key decisions, and evidence that in those instances, the rich do tend to prevail. It is their reliance on income groups that becomes their greatest hindrance, and which ultimately causes them to fail the test. A “ruling class” of high-income earners hardly constitutes a well-defined group, and indeed, income groups drawn in any configuration can offer a compelling basis for social division. Subsequent definitional squabbles between Gilens and Soroka and Wlezien in trying to distinguish the “truly rich from the merely affluent” (Gilens 2012, 117) only reinforce the fact that the elitist framework which Gilens’ findings invite can never be sufficiently captured through income alone. What Gilens finds in his data analysis is a compelling start to a larger, more vexing explanation of class domination. Unfortunately, for both Gilens and Page, and their wider readership, the mere suggestion that the rich rule seems sufficient. I argue that it is not; what is needed is a richer explanatory frame for class than the one offered by the income paradigm.   What is the rationale of income-based, so-called “unidimensional” accounts of inequality? Curtis and Jackson (1977) describe the principles underlying the use of single variables proxy variables in analyses of inequality. Variables of interest such as income are considered proxies because they seek to describe complex and partially concealed processes of stratification and to represent crucial outcomes of the “status attainment system” (Curtis and Jackson 1977, 124).  The underlying assumption that makes proxies suitable for univariate analyses is that these variables accurately summarize the multidimensional forms of stratification underlying the social structure. Good proxy variables, therefore, should be highly correlated with other proxies describing different                                                1 Dahl’s proposed test:  “The hypothesis of the existence of a ruling elite can be strictly tested only if: 1. The hypothetical ruling elite is a well-defined group. 2. There is a fair sample of cases involving key political decisions in which the preferences of the hypothetical ruling elite run counter to those of any other likely group which might be suggested. 3. In such cases, the preferences of the elite regularly prevail” (466)  6  dimensions of stratification. Grusky and Ku (2008, 6) describe eight dimensions on which stratification can be produced: economic, power, cultural, social, honorific, civil/legal, human and physical. Strong proxies will not only correlate with the primary dimension they seek to describe, but also with other privileged variables along other dimensions. Individuals with high levels of a particular valued status asset will also be likely to hold other status assets, thus describing a consistent and crystallized underlying “status attainment system” (ibid). It is important to note that under this form of analysis, “class” refers to little more than groups of individuals with similar levels of economic (and, to a lesser extent, cultural) assets. Class groups are derived by (often arbitrarily) subdividing a continuous distribution of incomes into distinct categories, i.e. middle class, knowledge class, noble class. This is in contrast to familiar “sociological” view of class, in which classes are derived from their relation to a shared, central economic process, the latter forming “objective” class groupings distinct from their occupants. From a strictly definitional standpoint, classes derived from income are statistical abstractions and “classes” in name only. However, this form of class is frequently invoked by scholars, politicians and laymen alike, bolstering its usefulness through ubiquity alone.   By standard criteria, income represents an ideal measure for univariate analyses.  In the United States in particular, income is strongly correlated other key measures of status such as education (Gilbert 1998, Bailey and Dynarski 2011), health and mortality (Rogot, Sorlie and Johnson 1992, Case and Deaton 2015) geographic location (Oakes 2012, Acemoglu, Johnson and Robinson 2002), and even physical attractiveness (Frieze, Olson and Russell 2006). Additionally, the relatively high salaries of many ‘status’ professions in the US, such as artists (BLS 2015) and professors (Jaschik 2012), makes income a stronger predictor of cultural dimensions of stratification than in other countries.  Other practical considerations make income an attractive variable with which to study inequality. Household income data is abundant and accessible. The American Community Survey, General Social Survey and American National Elections Survey form a reliable pool connecting income to a host of political and social attitudes. The more recent Luxembourg Income Study has greatly enhanced the ease of cross-national  7  comparisons of income-based studies.  As a “global standard of well-offness” (Cowell 2009), income measures generally speak for themselves and demand little methodological justification. This makes findings based on income categories more relevant and appealing to policymakers (Oakes 2012). Furthermore, income-based studies may make political science findings more portable in an interdisciplinary sense; in particular, they may enable a more efficient conduit between political scientists and economists (Atkinson and Brandolini 2011).  The well-studied methodological influence of economics is likely a major reason for the dominance of income-centric approaches.  Political scientists are more likely than sociologists, for instance, to adopt sanitary language of economists in describing inequality  (e.g., in the discussion of elites as “top earners”). More significantly, income-centric political researchers like Gilens favor an inverted approach to inequality researcher shared by economists, where income (or expenditure) serves as the starting point, and theoretical explanations are produced post-hoc (Atkinson and Brandolini 2011, 17). Theoretical approaches to income determination favored by economists tend to cast individuals as rational actors who choose between maximizing income in the present or developing human capital to secure greater future earnings (Wright 1979, 68). Much like rational choice theories, these assumptions are well-suited to economic investigations but are controversial when utilized by social scientists, whose task involves “sort[ing] out the relative importance of and interrelationships between achieved and ascribed factors in determining various outcomes, such as income” (ibid).  Economic approaches therefore assume a close correspondence between social processes that generate distinction (collectively referred to as human capital) and their concomitant representation in income levels.  This theoretical approach, including its emphasis on post-hoc theorization, is well-suited in explaining the economic effects of income inequality (i.e. poverty), but is less effective when addressing more complex political outcomes, as evidenced by Gilens.   The advantages to income-centric approaches are numerous and well-documented, but there are also a number of serious drawbacks that have generally received less attention. For one, the definition itself is broad and sometimes ambiguous. The term generally implies direct earnings over a set period of time, though what is included in this is sometimes unclear. Forms of indirect income such as capital gains,  8  government transfers or income-in-kind may or may not be included. Because these types of income are typically most prevalent among either low or high-income earners, ambiguities in the composition of income measures can be particularly problematic when addressing income inequality. Income data-gathering agencies frequently fail to distinguish between incomes generated from capital and labor (Piketty 2014, Piketty and Saez 1999), further exacerbating this problem. Labor market trends have also affected the stability of income measures.  Female workforce participation has dramatically altered the income composition of American households, weakens the link between income and other class indicators, particularly for middle-income households (Gilbert 1998). Micklewright and Schnepf (2007) argue that traditional survey measures of household income produce distortionary effects and information loss, particularly in relation to respondent’s age, sex and parental status, and that these issues are exacerbated in low-income settings. The mismatch between household and individual incomes poses a particular concern for women as it may distort or conceal relevant gender inequities in the class structure. Wright (1997) argues that the individual class character of those who marry higher-class spouses is lost and replaced by the mediated class location of their partners.  Despite its ubiquity, measurement issues are also pose a problem for researchers looking for reliable data on income. Some research suggests that income is frequently either under or over-reported on surveys, and that response rates are relatively low (Foster and Lound 1993, Oakes 2012). Unique problems present themselves when examining the upper-extremes of the income distribution. “Top-coded” datasets generally lump pinnacle elites within the top ten percent of income earners, making even the basic identification of the ultra-wealthy a difficult task (Page, Bartels and Seawright 2013). Researchers attempting to gather data on top earners have remarked that for the wealthiest Americans, “even their gatekeepers have gatekeepers” (ibid 53). This also implies that increasing stratification among the ultra-wealthy, as is the case in the US, may have a distortionary effect on future findings based on income-measures.  These authors highlight how temporal problems are of particular concern for the use of income. Cowell (2009) notes that the annualization of income is a convention which suits its operationalization as a variable, but also conceals more than it explains.  9  Unlike educational attainment, which is fixed on average after the age of 25 (Oakes 2012), annualized income is all but assuredly volatile within an individual’s lifetime. If wealth is a “person’s total immediate command over resources,” then we can see how income’s total usefulness as a variable will diminish over time as individuals come to draw more on accumulated assets or pension rights over direct earnings (Cowell 2009, 4). Likewise, individuals with similar levels of income may face dramatically different career trajectories based on differences in some other privileged asset (Wright 1997, Carnes 2012). For instance, the incomes of law students, journeymen, and union casuals do not reflect their prospective opportunities relative to those with similar incomes and limited career prospects.  Macroeconomic trends suggest that the connection between income and class will only become more complicated in the future.  The decline of household savings and the stagnation of real wages in the United States have blurred the relationship between earnings and material welfare.  Atkinson and Brandolini (2011) suggest that savings, and not earnings, are the crucial determinant of middle class status. The increasing reliance on credit for survival makes relating income to overall life chances more challenging. For lower-income households, access to credit is unreliable, leading to a special type of underclass referred to as “the unbanked” (Mullainathan and Shafir 2010).  Welfare retrenchment on state and national levels also complicate the picture for low-income earners. As these trends continue, income groups conceal increasingly heterogeneous life experiences for individuals within them. This logic has led economic theorists such as Sen (2008 [1997], 236) to call for a conceptual push away from the income paradigm and toward an understanding of economic inequality that takes into account how specific contingencies “lead to systematic variations in the conversion of incomes into distinction functionings we can achieve.” One criticism of the income paradigm is in its tendency to view income scales as “smooth continuous distributions” (Grusky and Ku 2008, 7). In actuality, the very meaning of income itself depends on other factors, in particular, how much of it you already have. This suggests that the usefulness of income will only continue to erode as income polarization continues in the US.   Given the problems associated with income measures, researchers may choose to employ a number of strategies to try and mitigate these issues. A common tactic seen in  10  recent political inequality research is to replicate analyses with education variables in place of income (e.g. Soroka and Wlezien 2008, Gilens 2009), which may serve to bolster confidence that the data are indeed describing a class relationship. In terms of explaining stratification among multiple dimensions, educational attainment is a more convincing operationalization of class than income, but it suffers from the problem of “bunching” around a small number of conventional educational categories. In some instances, correlational researchers may simply rely on the virtues of aggregation to deal with the problems of increasingly heterogeneous within-class experiences (Gilens 2012, Page and Shapiro 1992). This strategy of “washing out” variation may yield coherent preferences among income groups, but does little to address the larger problems of causal explanation. A once-dominant strategy for overcoming the problems of univariate income measures was to construct composite scores of socioeconomic status (SES) based on multiple dimensions of inequality. This had the advantage of combining income, education, as well as subjective dimensions of status within a unidimensional framework. This practice has fallen out of favor in recent decades, however, mostly due to the additional complexity of creating weighted scores and the rise of multivariate analyses (Lareau 2008, 12; Weeden and Grusky 2005, 158). In a comprehensive review of existing SES prestige scales, Hauser and Warren (1997, 177) suggest that such measures are a poor empirical fit to modern occupations, particularly those held by women, and conclude that such measures are “scientifically obsolete.”  Simplified composite measures of SES are still used on occasion, however, such as in Schlozman et. al’s (2005) recent analysis of political participation.   2.2 Causality and indeterminacy: problems with income-based measures  For a particular methodological outlook to be a dominant approach, it should be capable of explaining outcomes across the full breadth of questions a particular discipline is tasked with answering. Yet, as Hooker and Tillery (2016, 2) allege:   “[t]he recent literature on inequality in political science … has focused almost exclusively on rising income inequality and how it affects political representation … analyses of inequality in political science focus only on certain dimensions of state action (such as political behavior or voting) while ignoring others where the  11  bulk of citizens, particularly communities of color, experience key disparities shaped by race and class, such as the welfare state, the criminal justice system, and the provision of public goods and services.”   Most recent analyses of economic inequality by political scientists have been tightly focused around formal political processes. This is no doubt in part due to the influence and momentum built up by the Task Force report, but it is also reasonable to suspect that the lacuna in the recent literature identified by Hooker and Tillery is related to the income paradigm. In the end, income groups represent little more than statistical constructs rather than fully theorized economic classes. Their usefulness is maximized when related with dependent variables that express a clear preference, such as a vote or an opinion poll. However, the disciplinary territory of political science extends far beyond simply those questions, as the authors point out. Political scientists are charged with examining the state more broadly, and in this case, evaluating the state’s influence on economic inequality. Political scientists should therefore involve themselves in topics that are only peripherally related to the political arena, such as welfare, incarceration and labor policy institutions.   Political scientists who have previously worked on the question of how policies “feed back” onto society have done so by invoking appropriately thick conceptions of social actors. When analyzing the reciprocal effects of state activity on inequality, political scientists need tools that meaningfully describe society.  Esping-Anderson’s (1990) analysis of the welfare state examined not only the financial effects of welfare policy in terms of income, but effects on the more detailed circuitry of the occupational structure. Skocpol’s (1992, 58) pioneering work on feedback effects were based on how social policies “affect the social identities, goals, and capabilities of groups that subsequently struggle or ally in politics.” As the redistributionary dimensions of welfare continue to decline, revisions to welfare policy that target specific economic populations with work and employment-focused policies will demand increasingly granular methods of analysis (Heclo 2001). Likewise, the analytic framework for healthcare policy developed by Hacker (2002) emphasizes the social antagonisms embedded in the structure of employer-provided health plans. These forms of group-based, “institutionally nested” analysis (Jacobs 2007) are at odds with dominant income-based  12  conceptualizations of economic inequality, despite having many of the same practitioners. Replacing income-based approaches with more robust conceptions of class may serve to better integrate findings and close the gap between these two worlds of research.  The blending of state and society in these analyses demands a more robust conception of social groups than as financial categories. This does not imply that all pluralist income-based researchers must radically shift their frame of analysis to accommodate the state in its entirety. But it is also incorrect to believe that political scientists should cede these sorts of questions to other disciplines. The gradual narrowing of focus over recent decades is a worrying sign that instead of developing new approaches to address the problems under their disciplinary purview, political scientists might be gravitating towards the types of problems best suited to their methodological preferences.    Income-based measures might not only keep political scientists from studying the full-range of state activities, but may also encourage them to focus on the extremes of the income distribution. Income is inherently most meaningful when describing individuals with either a little or a lot of it, but its role as a determinant of life chances becomes less clear when considering the vast population of middle-earners. Middle-income earners may command similar financial resources, but this captures homogenous interests only under certain circumstances. This likely informs the decision for Bartels (2010), for instance, to focus his analysis around a few specific policies like the Bush tax cuts, the inheritance tax, and minimum wage reform. Though the middle class is a ubiquitous concept in the rhetoric of politicians, it forms a coherent category only in a relatively small number of cases.  One is therefore more likely to encounter studies that emphasize the differences between the very poor and very rich. However, these findings are ultimately less valuable when they cannot also uniquely describe the effects of middle-income earners. As argued by Atkinson and Brandolini (2011), income does not have a uniform effect on outcomes; social structures are “lumpy” entities consisting of heterogeneous experiences that are not captured by income. When classes are created by subdividing a continuous distribution of incomes, boundary problems inevitably arise. Lubrano’s Limbo (2003) describes these  13  problems through the experience “straddlers,” middle-income earners who inhabit “two worlds” of class. These sorts of contradictory experiences are ultimately lost when class is equated with income.  Many of the problems associated with income-based analyses can be avoided through thoughtful research design. However, they contain a more fundamental flaw which cannot be addressed as easily.  Analyses based on income may produce powerful associations which are difficult to ignore, but are all but silent on how conditions of inequality are translated into political outcomes. This mechanism problem is on some level generic and endemic to all types of correlational analyses. It is a particularly grievous problem, however, when the privileged explanatory variable is a proxy for the complexities of social stratification.  Sociological critiques of univariate research designs frequently cite this mechanism problem as a core deficiency of the approach. The problem has also been well documented by political scientists. For example, Brady, Verba and Schlozman (1995, 290) note that researchers used univariate analyses to confirm and reconfirm the relationship between socioeconomic status and participation with “monotonous regularity” for decades while leaving the more involved task of theorizing causal mechanisms incomplete. Work on causal mechanisms lagged well behind and relied on fugitive measures of psychological engagement and “political interest” (ibid 271). The authors’ resource model, which was designed to elaborate the causal mechanisms connecting socioeconomic status and participation, was introduced some three decades after the first associations were published. In perhaps an analogous recent trend, inequality researchers have piled onto a new or possibly refurbished term, “political voice,” which seems to be intended to animate recent findings behind a powerful metaphor, but may also inadvertently serve to conceal their theoretical skimpiness.  In the absence of a strongly-theorized connection between cause and effect, findings based on income-group measures will inevitably tend to privilege explanations which emphasize the role of money in securing political outcomes. Few facts about American politics are less controversial than money’s outsized role in the process. Findings confirming the magnitudes-greater influence the wealthy have over other citizens risk creating the characterization of the governing process as a “mechanical cash  14  register” that “tabulates the unequal political voices of the advantaged and organized” (Bartels et. al 2005, 88). The multitude of intricate ways in which elites influence the political process, such as through interpersonal social networks, command of symbolic power, or integration with bureaucracies and media installations (Perrucci and Wyson 2003), are washed out when analyses single out the command of financial resources as the rationale for class division. There is little evidence that money translates directly into votes in Congress (Ansolabehere, De Figueiredo, and Snyder 2002), though it facilitates avenues of access to elected officials (Langbein 1986).  Given the causal complexity suggested by these findings, it is therefore crucial that practitioners of income-based analyses devote significant time attempting to provide strong causal explanations for their findings.  Unfortunately, Bartels (2010) and Gilens (2012) devote only a handful of pages of their sizable studies to speculate as to how the rich influence the political process. In the absence of strong theorization, these authors effectively leave much of the causal heavy lifting to the imaginations of their audience. In the case of Gilens’ well-publicized study, media outlets were quick to roundly summarize his findings with headlines such as “Princeton Study: U.S. No Longer an Actual Democracy” (James 2014) and “Study: US is an oligarchy, not a democracy” (BBC 2014). “The clear finding” wrote one journalist, “is that the U.S. is an oligarchy, no democratic country, at all … The U.S., in other words, is basically similar to Russia or most other dubious ‘electoral’ ‘democratic’ countries” (Zuesse 2014). The runaway mischaracterization of Gilens’ study as an “oligarchy thesis” is a consequence of his failure to contextualize findings within the limits of a more complete causal explanation. His conclusions play to the sensitivities of a public deeply concerned with the political domination of ‘the affluent,’ but do little to advance the more crucial task of explaining how that domination occurs. This misinterpretation is somewhat tragic because it diminishes the actual significance of Gilens’ findings, which are quite remarkable. But more importantly, it highlights the pressing need for researchers to be clearer on what characterizes the social groups they are analyzing.    A similar problem can occur when researchers load too much causal explanation onto a single variable. Frank’s (2004) book What’s the Matter with Kansas? is an  15  analysis of shifting class-based preferences that uses education as the privileged variable. Frank’s work, a largely qualitative, historical analysis, suggests that cultural and moral differentiation on the basis of education has caused a major shift in the sympathies of working class voters. Frank’s thesis runs aground by overvaluing his central variable in his causal narrative to the point of describing the less-educated working class as boorish, resentful and susceptible to Republican appeals to working class disdain for “bookish, all-powerful liberals who run the country but are contemptuous of the tastes and beliefs of the people who inhabit it” (Frank 2004, 13).  By radically overreaching in the causal implications of his core variable, Frank commits the opposite sin of Gilens and Bartels, but to similarly troublesome effect. In a vigorous side debate, Frank’s reasoning has been roundly criticized as chasing a singular causal explanation at the expense of contradictory evidence (e.g. Bartels 2006, Van der Waal, Achterberg and Houtman 2007, Greeley and Hout 2006). As with Gilens’ and Bartels’ quantitative studies, Frank’s argument fails to respect the complexity of the phenomena he is attempting to describe.  The limits of univariate analyses in providing strong causal explanations are well known to researchers, but are still frequently unheeding within the frontiers of political science research on inequality. With evidence that socioeconomic divisions in the US are growing more salient, political scientists should be freely developing new approaches. Bourguignon (2006) describes an analogous problem faced by economists in defining the poverty line. Strong normative and sociological critiques of income-based measures by the likes of Sen led to the development of an alternative frame for understanding poverty based on endowments and capabilities. With political scientists faced with their own challenges related to the income paradigm, I argue that a similar first step is needed in re-evaluating such measures.          16  3. Sociological Alternatives to the ‘Income Paradigm’  3.1 Political science and sociology: historical cleavages  In this section I discuss alternatives to income-based class measures that have developed within the sociological tradition of class analysis. At first glance, it is not immediately apparent which factors have historically caused sociologists to value class much more highly than their counterparts in political science. The distinctions between political science and sociology are at times inconsistent; within the macro-institutionalist tradition, for instance, disciplinary boundaries all but disappear, with political scientists and political sociologists working on very similar types of problems using similar methodologies. The two fields diverge more sharply when analysis occurs at the individual level, where political scientists have developed a more distinctive, pluralist outlook and methodology. I argue that it is this pluralist outlook which has imbued political scientists with a deep discomfort about class. Pluralist intuitions, such as a skepticism of academic Marxism and elitist explanations of politics, have greatly influenced modern political scientists. In recent years, the role of class has been further sidelined by the rise of identity politics, whose accounts of stratification are more amenable to a pluralist framework. Lastly, the proliferation of a “post-class” scholarship has directly challenged the relevance of class and has made class-based methods controversial.   The legacy of Dahl seems crucial in explaining the absence of class from the methodological toolkit of political scientists. As I have mentioned previously, Dahl argued that elitist explanations of politics were not only unsatisfactory, but potentially dangerous as a means of describing political inequality. The temptation for elitist researchers to search for a “covert group” --  a ruling class which informally occupies authority structures above actual leaders -- leads scholars down a rabbit-hole of  “quasi-metaphysical” investigation amounting to an “infinite regress of explanations” (Dahl 1958, 463). These sorts of explanations are attractive because of their elegance; the messy, complicated task of identifying differing inputs is replaced by the identification of a ruling cohort from which all true preferences emanate.   17   In their famous essay, “The Two Faces of Power,” Barach and Baratz (1962) provide analytical clarity to this distinction between plural-individualist accounts favored by Dahl and the elitist, class-centric approach of sociologists. The authors argue that it is principally through diverging accounts of how power influences social structures that the division between sociology and political science can be understood. While sociologists view power as centralized within rigid and enduring social structures, political scientists (particularly those of the pluralist variety), see power as “widely diffused” within a community (Barach and Baratz 1962, 947). From this perspective, power is defined principally as “participation in decision-making,” making the pluralist researcher disinterested “in the reputedly powerful” (ibid 948). Furthermore, such power can only be expressed during moments of “key decision,” wherein actual influence can be reconciled and measured. This conception of power reinforces the individual as the primary unit of analysis, and individual participation as the primary vehicle of political empowerment. Compared to sociological understandings of concentrated “power structures” within a community’s stratification order, the influence of elites is circumscribed within the boundaries of the decisionmaking process. Elites instead exercise power through membership in “epistemic communities” with outsized influence on decisionmaking outcomes (Jacobs and Page 2005). Ultimately, Barach and Baratz argued that neither concept of power was truly sufficient. Both sociologists and political scientists overlook a form of power which could be described sharing features of both versions. Power over the dynamics of “non-decisionmaking” refers to the capacity to determine which types of decisions are carried into the political process. This suggests that Dahl had only half of the equation; what is needed is to marry pluralist tendencies with a deeper examination of the social bases of power which are causally prior to the formal political process, but which deeply shape the trajectory of those outcomes.   Indeed, political scientists since Barach and Baratz’s writing have come to increasingly appreciate the antecedent social conditions that preempt the political process. These developments, however, have largely overlooked class in favor of other social affiliations argued to be more basic to one’s identity, such as race, gender and ethnicity. This new framework of identity politics has reframed many of the classic problems of class inequity in new light: the politics of class-based material redistribution  18  have been recast as a politics of group-based recognition and inclusion (Taylor 1994, Young 1990). From this perspective, class is seen as an essentializing category that strips away other vital forms of identity that effectively reproduces underlying forms of political and economic inequality.  The extent to which identity politics has affected political science can hardly be overstated. As the new normative core of the discipline, the last several decades have seen an explosion of theoretical and empirical work devoted to this new framework of inequality. While these developments should no doubt be applauded and celebrated in their own right, the extent to which they have contributed to the current impoverished standards for assessing strictly economic inequality has not been widely addressed. Fraser (1999) argues that identity politics pushes class further out of the frame as a relevant form of inequality. This omission not only contributes to the types of ill-defined, euphemistic operationalizations of class typical of the income paradigm, but potentially makes identity politics a “handmaiden of capitalism” by distracting the watchful eye of social scientists from newer, more pernicious forms of economic exploitation (Fraser 2013).  What might account for this shift toward ascriptive forms of identity over class as the decisive factor in stratification? One major explanation might be the growth in a “post-class” theoretical perspective, which sees a declining salience of class in response to changes in the post-industrial economic order. From this perspective, classes and class experience are seen as being replaced by a “complex mosaic of taste subcultures” (Pakulski and Waters 1996, 157). Lipset (1960), an early proponent of class-based analysis in the study of democracy, later became a prominent advocate of a post-class perspective (Lipset and Clark 1991). This debate on class is contentious, ongoing and defies brief summarization, but the relative merits of each perspective are no doubt overshadowed by the impact of the debate itself. In many quarters, the status of class is seen as being in question.  Undoubtedly, amongst political scientists, any form of implicit sanction against the use of strong forms of class is at least partially caused by the uncertainty produced by this debate. These challenges posed to class by both identity politics and post-class perspectives are useful in explaining why political scientists have shied away from class in recent decades; however, they are ultimately secondary to the  19  central problem identified in this paper, which is that class is indeed being used by contemporary political scientists, albeit in ambiguous, non-ideal terms. Lastly, one cannot overlook the relationship between class analysis and the negative legacies of Marxism. Dahl’s critique of elitist explanations was made during a time when strong Marxist currents ran through social science, and his comments can be seen as a statement against Marxian interpretations of democratic politics. Although Marxist ideas figure heavily in several prominent perspectives in political science, few would disagree that there is a strong, implicit aversion to Marxism within the discipline, particularly when compared to cognate fields of sociology, anthropology and geography. There are countless well-informed reasons for discrediting Marxist reasoning; so many as to constitute a separate paper unto itself. However, the Great Recession has seen a renewal of so-called “soft Marxism” within mainstream analyses. Piketty’s (2014) towering, bestselling study on the contemporary effects of capital accumulation largely confirms Marxist intuitions which were dismissed for decades. In US popular discourse, class rhetoric such as “99 percent” is met with charges of “inciting class warfare.” Despite its recent fashionable reformulation, there are few signs that Marxism has moved from the academic margins in any meaningful way. Thus, the strong association of class analysis as a primarily Marxist instrument will no doubt continue to be a major implicit barrier in the adoption of stronger conceptions of class among political scientists.  Though class-based analyses will always smack of Marxism on some level, few contemporary class theorists are directly descended from Marx and fewer still are concerned with explicitly furthering his intellectual project. Even those which are explicitly Marxist in character (e.g. Wright’s class model, discussed below) are interested more in extracting the unique forms of analytical leverage afforded by a Marxist approach than in connecting class struggle into a determinate historical framework. The extent to which the adoption of class-based approaches by political scientists would be predicated on first “de-Marxifying” class-analytic traditions is unclear, but two instructive points are clear. First, approaches to class analysis cannot demand a Marxist epistemological frame, or, for that matter, a reliance on “benign narratives” which view greater equality as a determinate consequence of economic development (e.g. Kuznets 1955). Likewise, to the extent that imperialist attitudes about Marxism persist within  20  political science, they should not serve as a basis through which to discredit contemporary forms of class analysis.  As “sociology’s favorite concept,” class has been the object of intense scholarly debate for centuries. The obsessive nature of these debates, often regarding the seemingly smallest minutiae of class definition, is well known even outside of the field. Furthermore, the vision of class that is produced from these debates often bears little resemblance to its everyday usage. It is of little surprise, then, that sociological conceptions of class seem to receive little attention outside of the committed gaze of diehard class specialists. However, I argue that it is a mistake to completely ignore the work of contemporary class analysis altogether. Within the broad surplus of work on class, there are a number of distinct perspectives with the prospect of enriching the study of political inequality.  In this section, I will briefly outline and discuss three alternatives to income-based operationalizations of class: Erikson and Goldthorpe’s neo-Weberian model, Wright’s neo-Marxist mode, and Weeden and Grusky’s microclass model.  Grusky and Ku (2008, 7) suggest that the contemporary rendition of a structured-class analysis should be understood as “a distinctively sociological reaction to the individualism of the income paradigm and other unidimensional approaches to measuring inequality.” As such, these theories all conceptualize class in a way that is fundamentally different than the atheoretical modes that political scientists currently favor. Conventional categories of ‘lower-middle-upper’ classes are “gradational” in that they describe groups with similar amounts of a particular asset. Class analytic approaches are “relational,” in that they describe classes based on the “relationship of people to income generating resources or assets” (Wright 2003, 2). Rather than collections of individuals with similar levels of income, classes are structural in the sense that they are  “empty places” which “exist independently of individual occupants of these positions” (Sorenson 1991, 72). However, the extent of this structure and rationale for class divisions varies significantly between class theories. These boundaries may serve diverse purposes such as establishing “common positions within status hierarchies,” or even as “conflict groups determined by their position within authority or power structures” (Wright 1974, 3-4).  Ultimately, class distinctions have analytical integrity when they genuinely describe people with common economic life chances (ibid). Class models, therefore, treat economic processes as “status  21  producing” activities, and seek to capture salient group features based on their relationships to production. Differences between specific approaches to class analysis can generally trace themselves to the theoretical canons of prominent social theorists, especially Weber, Marx and Durkheim. Weberian approaches emphasize how employment and market access affects life chances, that is, “the chances an individual has for sharing in the socially created economic or cultural ‘goods’ that typically exist in any given society” (Giddens 1973, 130). Marxist models emphasize class interdependencies and conflict mediated through mechanisms of exploitation (Wright 2005, 3), while Durkheimian approaches focus on how class locations “homogenize” individuals and produce group dispositions (ibid).   3.2 Erikson-Goldthorpe class model  The Erikson-Goldthorpe (EGP) class schema is likely the most popular and widely utilized class model today. In addition to being frequently used in class analyses by academics, the EGP scheme forms the basis of several government employment classifications, including the UK National Statistics Socioeconomic Classification (Oakes 2012). This makes data availability a less acute problem for the EGP scheme, as data are frequently either collected or formatted into EGP categories. Erikson and Goldthorpe (1992, 37) describe the aim of the model is “to differentiate positions within labour markets and production units or, more specifically … to differentiate such positions in terms of the employment relations that they entail” (authors’ emphasis). Within employment relations, Erikson and Goldthorpe seek to describe how the nature of an individual’s relationship with their employer affects broader life chances. The nature of this employer-employee relationship can be summarized through an important dichotomy “between positions that are regulated under a labor contract, and those that are regulated by a ‘service’ relationship with the employer” (Breen 2005, 37). In contrast to traditional, routine exchanges of labor for wages under close supervision, skilled employees who enjoy a “service relationship” operate with more discretion and delegated authority. Service relationships are produced when employers attempt to overcome monitoring problems caused by the high “asset specificity” of specialized workers. In lieu of external  22  sanctions, employers seek incentive structures that build “moral commitment” in both parties, establishing a link “between employees’ commitment to and effective pursuit of organized goals and their career success and lifetime material well-being” (Goldthorpe 2000, 220). Thus, service relationships are built on prospective incentives: “salary increments on an established scale, assurances of security…pension rights…. [and] well defined career opportunities” (qtd. in Breen 2005, 38).   Table 1 - Erikson-Goldthorpe Class Typology  Class Occupation Employment relation I + II professionals, administrators, managers, skilled technicians (service class) service III routine non-manual workers intermediate IV small proprietors and artisans (petty bourgeoisie) self-employed V lower-grade technicians and supervisors intermediate VI skilled manual workers intermediate VIIa non-skilled manual workers (non-agricultural) labor contract VIIb agricultural workers labor contract Source: Erikson and Goldthorpe (1992)   For Erikson and Goldthorpe, the crucial determinant of class boundaries is the type of employment relation which characterizes those occupations. This distinction between service relationship and labor contract elaborates a more detailed causal explanation connecting education, skills and expertise to material and status outcomes. The “salariat” of skilled workers enjoy improved life chances not only by virtue of their superior incomes and educations, but because their skills insulate them from the precariousness which otherwise dominates working life. The prospective incentive structures of the service relationship mean that skilled workers see long-term security, Adapted from Erikson and Goldthorpe (1993)   23  prosperity and career advancement in place of uncertainty. In other words, whereas traditional accounts of class would consider improved life chances as byproduct of increased income, the EGP model sees increased income as merely epiphenomenal of processes which secure improved life chances.  Breen (2005, 36) argues that the “single defining characteristic of Weber-inspired class analysis is that classes are of interest only insofar as they shape life chances,” which in principle limits its power to predict other types of outcomes. However, “as an empirical fact, it emerges that such schemata do often prove to be good predictors of behaviors, actions, attitudes, preferences and so forth” (ibid). A particular strength of utilizing the Erikson-Goldthorpe scheme over unidimensional proxy analyses is the ability to conduct longitudinal studies of class. EGP categories have been repeatedly shown to be excellent predictors of intergenerational mobility, which is particularly problematic for gradational accounts of class (Ganzeboom, De Graaf and Treiman 1992, 4). This has made EGP the method of choice for studying political realignment (e.g. Hout, Brooks and Manza 1999, Goldthorpe 1999, Evans 2000). Associations uncovered by these analyses have advantageous over income-based analyses because they enable more satisfying, detailed causal explanations. Hout, Brooks and Manza (1999, 90) illustrate this, in describing a distinction uncovered between class I professionals and class II managers: The most important feature of this trend has been the bifurcation of the salariat. Managers and professionals have followed different political paths since the 1960s. Managers’ interests cling closely to the low-tax, deregulation agenda of the Republican Party. Professionals have less of a stake in these policies for several reasons. Although their incomes have been as high as managers’ in recent years and were actually slightly higher in the 1970s, professionals are not as worried about government spending or regulation as managers are. Strikes and wage demands seldom affect professionals’ work.  The causal intuitions about politics extracted from a sociological measure of class are more detailed and compelling because effort is made to identify causally relevant distinctions at the measurement level, rather than after the data are collected. A neo-Weberian emphasis on life chances yields a more compelling image of the differentiated interests of groups that would otherwise be assumed to want virtually the same thing. Attempts to explain this nuanced picture through the inclusion or substitution of  24  additional proxy variables (e.g. education in Frank 2004), will also likely miss such fine-grain distinctions.    Though it yields a much richer and more useful picture of class than one defined by income alone, the EGP schema has attracted considerable criticism as well. As an occupational aggregate, the scheme has been criticized in its ability to predict the “wider cultural and social activities and identities which do not appear to be closely linked to people’s class position” (Savage et. al 2013, 222).  A strong feminist objection to the scheme has been made on the grounds that its categories were initially generated based on a predominantly male workforce (Stanworth 1984). Goldthorpe (1983, 468) defended this convention by suggesting that female participation in the workforce “typically forms part of a family strategy … in which the husband’s employment remains the dominant factor.” More problematic still is how the underlying rationale for class distinction, the employment relation, has little relevance for those who do not have traditional employment. The class V category of petty bourgeoisie and artisans is bloated and surely describes individuals with very different life chances. More critically, the regular 7-part scheme omits employers and economic elites (Savage and Williams 2008), who are of particular interest to the current wake of political science researchers. These issues limit the usefulness of the EGP in addressing certain questions.   3.3 Wright neo-Marxist class model  Perhaps no scholar has been a more successful advocate for the relevance of contemporary class analysis than Erik Wright. His neo-Marxist model is the result of a decades-long research project to reconstitute a Marxian class analysis within the framework of an empiricist, positivist social science (Wright 1985, 1). This has entailed an abandonment of many problematic features of classical Marxist analysis: the labor theory of value, the denial of middle classes, the teleological implications of “true class interest,” and the skepticism of democratic politics. Prior strategies by Marxists to explain how advanced capitalism “survived” Marx’s predicted downfall (e.g. Poulantzas 1973), or to empirically “prove” Marxist assumptions through the identification of elitist  25  structures of domination (e.g. Mills 1956, Domhoff 1967) serve only to entrench critics of Marxist analysis and further marginalize its sympathizers. Instead, Wright looks to theorists such as Dahrendorf (1959), whose work can be seen as a more compelling attempt to rescue the useful elements of a Marxist analysis from the weight of its own theoretical implications. Wright therefore embraces the ‘middle-class problem’ of the post-industrial labor market with an expanded version of a Marxist class typology. Wright stops short, however, of embracing a fully positivist orientation which equates theorization with the process of identifying and categorizing observable regularities (Wright 1985, 3).Wright’s model therefore attempts to capture the empirical complexity of the post-industrial labor market while retaining a strongly-theorized structure of class relations.  The result is a class map of structured, “empty space” class locations based around newly important post-industrial considerations. Though his scheme offers a template of class structure which is more amenable to an audience of class skeptics, it remains a purpose-built instrument for examining contemporary Marxist questions about class. For this reason, “deductive” character has led Sorenson (2005, 124) to remark that its “deductive” character makes it the most theoretically ambitious of the contemporary class models in wide use today.  Figure 1 - Wright Class Typology   Source: Wright (1997)   26  Like the EGP schema, Wright’s model observes the close correspondence between expertise and life chances in the modern labor market. However, as a reconstruction of the traditional Marxist class framework, Wright’s model treats class categories as locations in a rigid class structure, rather than aggregations of occupations. Wright’s model is a “2.5” dimensional typology consisting of twelve class locations, three for employers, and nine for employees. This yields an immediate advantage over the standard 7-part EGP classification, which focuses strictly on the experiences of employees. From this basic capitalist-worker dichotomy, Wright subdivides employers into large, small and self-employed categories, with small employers being defined as having ten or fewer employees.2 Following Weber and Dahrendorf, Wright further subdivides employees along skill and authority dimensions, yielding nine class locations for employees.  For Wright, the principal cargo worth salvaging from Marxist analysis is the concept of exploitation. Exploitation is the mechanism which gives Marxist operationalizations of class explanatory leverage over their counterparts. Wright’s definition of exploitation strips away some of the detritus of previous deployments of the term.  For Wright, exploitation describes situations characterized by a particular form of social interdependence which satisfies three criteria: (1) the “inverse interdependent welfare principle,” where the material welfare of exploiters causally depends on the material deprivations of the exploited, (2) the “exclusion principle,” where this inverse interdependence is dependent on the exclusion of the exploited from access to certain productive resources, and (3), the “appropriation principle,” where exclusion generates material advantage by enabling the appropriation of a material surplus (Wright 2005, 23). Exploitation is significant because it defines a scope of antagonistic action between social actors. Furthermore, exploitation does not simply occur across the “property line” between employers and employees. Forms of exploitation also pattern the relationship between different kinds of employees. Wright argues that in post-industrial labor markets, specialized knowledge and expertise also form the basis of exploitation through their status as rent-producing assets. These assets are created and sustained through                                                2 Wright acknowledges that this scheme fails to capture the true dynamics of capitalist wealth holding, but yields a reasonable empirical estimate of capitalists as constituting approximately 2% of the labor force (Wright 1997, 44).   27  credentialing institutions and political regulations. Predictions that postmodernization would result in the “de-skilling” of workers and the heightening of exploitation around the property line (e.g. Braverman 1974) have not come to pass, but neither have predictions of post-class social integration.   In exploitation, Wright has a strongly-theorized mechanism which is sorely missing from proxy-based investigations of class. Instead of having a vaguely defined middle class of medium-affluent individuals, shifts in the composition of the labor market can be seen as opening up new sites of exploitation and antagonism between classes. Income-based analyses tend to reiterate many of the same policy tests of class interest, such as tax policies and minimum wage laws (Bartels 2010), Wright’s model demonstrates how these tests constitute only a fraction of relevant issues in which interests may form on the basis of class location. Divergent class interests instead take shape in the form of skill and loyalty rents, credentialing, and access to social and cultural capital (Wright 1997, 18-19). Gallagher (1986, 251-252) argues that social scientists too frequently ignore the historical and social context of how groups develop around specialized knowledge, instead relying on a functionalist belief that expertise is deployed altruistically. This critical perspective, which has been referred to as the “politics of professionalism,” (Provine 1986) emphasizes how organized professionals embed themselves within political and bureaucratic structures to capture and control markets. Credentialing, for instance, serves as a guarantee of “risks incurred through specialized training [which] takes the form of a monopoly, or at least of protection by public authority” (Larson 1977, 15). These processes often play out in the margins of the political arena, but over time have tremendous consequences for class structure. Moreover, these forms of political struggle are overlooked by political scientists who consider only the direct, financial dimensions of economic inequality. Wright’s operationalization of class categories, though itself far from perfect, more ably captures these forms of divergent class interest which escape the income paradigm. These forms of social closure should be of great interest to political scientists, who have themselves frequently debated the “contradictory” nature of post-Fordist class politics. While explanations pointing to threshold effects (Inglehart 1990), cross-cutting cleavages (Lipset and Clark 1991, Gilens 1999) and moral values (Frank 2004) are valuable in their  28  own right, they risk painting over the subtle antagonisms within class politics which likely co-exist with ideological motivations.  Wright contends that class analytic approaches like his cannot serve as a satisfactory method to settle entrenched debates about class because they cannot adjudicate between perspectives that treat class as a “realist concept” and an “arbitrary convention” (Wright 1997, 256).  The enormous debates around the alleged decline in class voting, for instance, have remained largely unsettled due to contradictory evidence produced by different class measures (e.g. Gorz 1982, Lipset and Clark 1991, Pakulski and Waters 1996, Evans 1999, Van der Waal, Achterberg and Houtman 2007). Inglehart (1997, 256), for instance, admonishes class proponents for their use of complex class measures instead of the “theoretically clear and easy to operationalize” manual-non-manual distinction. The signal sent by this stalled debate is clear: rather than attempting to achieve consensus on the ambiguous question of whether class-based voting “exists” in the modern democracies, researchers should instead engage the more productive, but much more ambitious question of how class interests have evolved in response to socioeconomic change. An investigation of this sort places classes themselves as the central subject of interest, rather than as instruments tasked with justifying their own existence.   3.4 Microclass model  Despite Wright’s best effort in bringing Marxist class analysis up to the standards of empirical social science, it remains unlikely to be embraced by political scientists. Within class analysis, there are similar concerns with the indebtedness of class models to a particular theoretical outlook: “Most often, scholars choose a measurement paradigm not on the basis of scientific criteria, but rather as a matter of faith or as a symbolic badge of affiliation with a discipline, subfield, or favored scholar” (Grusky and Weeden 2008, 65). The impetus for Weeden and Grusky’s (2005, 141) microclass model is to produce a class model which favors empirical veracity over stylized renditions of “big class” maps that conform to the standards of a theoretical forbearer. The authors contend that while such class models have been traditionally adequate in explaining macrolevel outcomes  29  such as revolutions and larger accounts of collective action, they are poorly suited for explaining the microprocesses of interest to contemporary social scientists.  In simplified terms, the microclass approach means that class is operationalized at the level of occupation rather than some larger aggregation. Weeden and Grusky’s goal is to operationalize class in a way that dispenses with the usual preoccupation with class formation and focus their efforts of a model which has maximum utility in explaining outcomes. They do, however, retain an understanding that the defining advantage of a class analytic approach over a proxy-driven strategy is in the articulation of stronger causal mechanisms. Thus, the authors goal is to provide an “information-rich class map” that identifies “internally homogenized” categories at the site of production which are institutionalized by the labor market itself (Weeden and Grusky 2005, 143). In contrast to “big class” models, the microclass approach is “not very demanding; after all, classes that are fine-tuned to the micro-level agenda do not need to embody antagonistic interests, act collectively on behalf of these interests, or bring about fundamental macrolevel change” (ibid).   Weeden and Grusky’s underlying rationale for their class map is based on homogenizing mechanisms that occur on the micro-level. This makes their model particularly attractive to scholars outside of sociology who are skeptical of procedures used to create a “big class” typology. The authors outline three types of mechanisms: allocation, social conditioning, and institutionalization of conditions. Allocation mechanisms describe processes of self-selection where individuals sort themselves into occupations by matching their personal dispositional and demographic characteristics to particular professions. In turn, firms partially filter candidates based on their dispositions (and in some cases, demographics) in order to find employees who fit the culture in place. A second category of mechanisms, social conditioning, describes how homogeneity crystalizes out of social interaction. Training regimes instill explicit codes of conduct, and forms of “interactional closure” impose implicit standards on interpersonal communication. These mechanisms expose employees to the beliefs and preferences of their colleagues and also establish which pre-existing attitudes are acceptable. The last type of mechanism, institutionalization of conditions, describes how occupational incumbents face similar working conditions and daily patterns. The authors  30  argue that the growing variety of working conditions makes capturing these distinctions in aggregate, as Erikson and Goldthorpe attempt, increasingly untenable.  Even in a “disaggregated” class model like Weeden and Grusky’s, a class typology must be constructed based on some underlying protocol. The microclass model features 126 occupational categories, making it difficult to summarize as a traditional class map. The authors used occupation-level data on forms of social closure (e.g., credentialing, certification, associations, licensing) “that detailed occupations have pursued and realized … to identify the institutionalized boundaries that generate intracategory homogenization” (Weeden and Grusky 2005, 156). This approach is reminiscent of Wright’s in its focus on social closure as a decisive feature of class differentiation. What is produced is a model which the authors claim is far more effective in predicting microlevel outcomes. When submitted to rigorous comparative testing, they find that the microclass model is significantly better than both “big class” and traditional composite SES measures at explaining outcomes across a variety of domains. Interestingly, they find that microclass is surpassed only by religion as a predictor of political beliefs (ibid 176). This finding reinforces their intuition that occupations forge homogenous beliefs based on interpersonal social mechanisms. . When subjected to the same tests, “big class” maps also show statistically significant correlations between class and political beliefs but to a much lesser extent. Weeden and Grusky’s analysis confirms the general poverty of proxy-based approaches relative to the class analytic approach., but also  that the occupation-level operationalization of class, though less parsimonious, would be of greatest interest to correlational researchers.   Weeden and Grusky’s ambition to produce a class model suited for use by the broader social science community shows signs of success. Carnes, a political scientist studying inequality, has recently broken with income-based conventions and embraced Weeden and Grusky’s microclass perspective. In his analyses of class bias within US political institutions, Carnes has turned to class analytic methods to great effect. In his small-n study on the effects of class background in congressional decisionmaking, Carnes (2012, 2013) finds that a legislator’s occupational background significantly affects their economic policy preferences. Re-running the same analysis using income and educational measures, Carnes finds no effect. A more recent paper on the shortage of working-class  31  candidates for office (Carnes 2015) highlights the existence of “demand-side” class biases among institutionalized party gatekeepers. Carnes’ aggressive research agenda based on stronger conceptions of class is peerless in recent scholarship. His intentional eschewing of income-based measures (Carnes 2012, 8) has afforded him unique leverage on the study of economic inequality. Carnes’ findings provide evidence of sociological intuitions that class experiences forge deep interests, and that interclass antagonisms form a salient feature of the American political system. Carnes’ findings, however promising, are only preliminary, but his research offers a glimpse into the potential benefits of adopting a sociological approach to class.   Carnes’ studies represent a promising utilization of class-based methods, but do so within the narrow confines of a traditional quantitative research agenda. By focusing on the background of a relatively small dataset of legislative elites, Carnes can utilize widely-available biographical information as the basis of his class investigation. This sidesteps the otherwise inevitable confrontation with data availability problems which complicate class-based analyses. Gathering reliable data on the class backgrounds of non-elites and voters is a difficult and costly proposition, especially if the resulting data may not yield much additional leverage over the use of proxy measures. Ideally, then, microclass measures would be utilized to address questions where the underlying logic of income-based proxies can be seriously questioned. For instance, investigations into the composition of middle-income political preferences, which are difficult to decisively answer using income alone, are promising avenues of research that have been largely overlooked by political scientists thus far. The considerable resource investment demanded by these projects means that class-sympathetic researchers must be clear and forceful in their acknowledgement of the advantages of class-based techniques over prevailing standards.         32  4. Conclusion  In an era of unprecedented economic inequality, the methods of political scientists fall short of what is necessary in capturing a satisfying representation of the nature of economic inequality. The intent of this paper is to highlight this deficiency in preliminary terms, but leaves the more challenging task of detailing precisely what these developments may entail to future work. It is a telling fact that within a field as renowned for its methodological innovation as political science, nothing approaching an authoritative text has been written on the subject of class. The last decade of research on economic inequality by political scientists has seemed to invite the specter of class, yet it remains at best an ambiguous concept.   Embracing the types of sociological conceptions of class detailed in this paper would present unique challenges of their own. The conventions of data collection and analysis are very much tailored to current standards of research. Datasets that are reliably mined by political scientists do not presently lend themselves class-based analyses easily. Lareau (2008, 8) notes that “[f]ew data include key variables that correspond to our conceptual models. For example, we seldom have access to detailed measures of authority relations in the workplace, the amount of job-specific expertise required by particular jobs, the informal cultural knowledge used in the workplace, or current conditions in the regional labor market.” Those political scientists who are interested in class-based approaches, such as Carnes, may first face the challenge of producing proprietary datasets. The class models presented in this paper are designed to be utilized in quantative, aggregate analyses, but class approaches may also require significant investments in historical or ethnographic methods. This poses a serious practical concern in academic settings that value prolificacy and high-volume scholarship.   These challenges reflect the high costs researchers would have to incur in order to adopt new standards, but also suggest the promise of such approaches in bringing some much-needed clarity to the study of economic inequality. Political scientists on the frontiers of research may find additional avenues to studying class that are less costly; for instance, through investigations of the subjective and cognitive dimensions of class. Analyses of this sort would fit comfortably with existing cognitive-affective models of  33  politics (e.g. Conover 1988), and could yield important insights into how class influences political behavior (e.g. Sadin forthcoming, Carnes and Sadin 2015) and how it intersects with other politically relevant forms of identity.   The opportunities by a class-based approach to political science are countless. However, taboos against class are embedded in the very foundation of the discipline. In an era that has been characterized as a “New Gilded Age,” (Bartels 2010), political scientists should feel comfortable transgressing old disciplinary boundaries to test and develop new methods with the goal of producing stronger, more complete causal explanations. Class is destined to remain a deeply contested concept within social science, but this should not dissuade political scientists from seeking out the best available practices for capturing it.                      34  References  Ansolabehere, de Figueiredo, and Snyder. 2003. Why Is There So Little Money in US Politics? Journal of Economic Perspectives 17:105-130. APSA Task Force on Inequality and American Democracy. 2004. “American Democracy in an Age of Rising Inequality.” Perspectives on Politics 2 (December): 651-70.  Atkinson, Anthony and Andrea Brandolini. 2011. “On the identification of ‘middle class.’” Society for the Study of Economic Inequality Working Paper Series.  Bachrach, Peter and Morton Baratz. 1962. “The Two Faces of Power.” The American Political Science Review 56 (December): 947-952.  Bailey, Martha J. and Susan Dynarski. 2011. “Inequality in Postsecondary Education.” In Whither Opportunity, eds. Greg Duncan and Richard Murnane. New York: Russell Sage Foundation: 117-32.  Bartels, Larry, Hugh Heclo, Rodney Hero and Lawrence Jacobs.  2005. “Inequality and American Governance.” In Inequality and American Democracy , eds. Jacobs, Lawrence and Theda Skocpol. New York: Russell Sage Foundation. Bartels, Larry. 2010. Unequal Democracy: The Political Economy of the New Gilded Age. New York: Russell Sage Foundation.   Bashir, Omar S. “Testing Inferences about American Politics: A Review of the ‘Oligarchy Result’.” Research & Politics 2 (September): 1-7.  Baumgartner, Frank R., Jeffery Berry, Marie Hojnacki, David Kimball, and Beth Leech. Lobbying and Policy Change: Who Wins, Who Loses, and Why. Chicago: University of Chicago Press. BBC. 2014. “Study: US is an oligarchy, not a democracy.” BBC News. 17 April.  Brady, Henry, Sidney Verba and Kay Lehman Schlozman. “Beyond SES: A Resource Model of Political Participation.” The American Political Science Review 89 (June): 271-294.  Branham, Alexander, Stuart Soroka and Christopher Wlezien, forthcoming. “When do the Rich Win?” Working paper. Braverman, Larry. 1974. Labor and Monopoly Capital. New York: Monthly Review Press.  35  Breen, Richard. 2005. “Foundations of a neo-Weberian class analysis.” In Approaches to Class Analysis, ed. Erik Olin Wright. Cambridge: Cambridge University Press. Bureau of Labor Statistics. 2015. “Occupation Employment and Wages, May 2015.”  http://www.bls.gov/oes/current/oes271013.htm (13 August 2016). Carnes, Nicholas and Meredith Sadin. 2015. “The ‘Mill-Worker’s Son’ Heuristic: How Voters Perceive Politicians from Working-Class Families—and How They Really Behave in Office.” The Journal of Politics 77 (January): 285-298.  Carnes, Nicholas. 2012. “Does the Numerical Underrepresentation of the Working Class in Congress Matter?” Legislative Studies Quarterly 37 (February): 5-34.  Carnes, Nicholas. 2013. White Collar Government: The Hidden Role of Class in Economic Policymaking.  Case, Anne and Angus Deaton. “Rising morbidity and mortality in midlife among white non-Hispanic Americans in the 21st Century.”  Proceedings of the National Academy of Sciences 112 (December): 15078-15083.  Conover, Pamela J.  1988. “The Role of Social Groups in Political Thinking.” British Journal of Political Science 18 (January): 51-76. Cowell, Frank. 2009. Measuring Inequality. 3rd ed. Oxford: Oxford University Press. Curtis, Richard and Elton Jackson. 1977. Inequality in American Communities. New York: Academic Press. Dahl, Robert. 1958. “A Critique of the Ruling Elite Model.” The American Political Science Review 52 (June):463-469. Domhoff, William G. 1967. Who Rules America? Saddle River: Prentice Hall.  Enns, Peter K. “Relative Policy Support and Coincidental Representation.” Perspectives on Politics 13 (December): 1053-1064. Erikson, Robert and John Goldthorpe. 1992. The Constant Flux. Oxford: Clarendon. Esping-Anderson, Gosta. 1990. “Post-industrial Class Structures: An Analytical Framework.” In Changing Classes: Stratification and Mobility in Post-Industrial Societies, ed. Gosta Esping-Andersen.  London: SAGE Publishing. Evans, Geoffrey. 2000. “The Continued Significance of Class Voting.” Annual Review of Political Science 3 (June): 401-417.   36  Foster, Kate, and Charles Lound.  1993. "A Comparison of Questions for Classifying Income." Survey Methodology 32: 1-7. Frank, Thomas. 2004. What’s the Matter with Kansas? New York: Holt. Fraser, Nancy. 1998. “From Redistribution to Recognition?” Dilemmas of Justice in a ‘Post-Socialist’ Age.” In Feminism and Politics, ed. Anne Phillips. Oxford: Oxford University Press. Frieze, Irene, Josephine Olson, and Jane Russell, J. 1991. “Attractiveness and income for men and women in management. “ Journal of Applied Social Psychology, 21: 1039 –1057. Gallagher, William. 1988. “Review of Judging Credentials: Non-lawyer Judges and the Politics of Professionalism.” Legal Studies Forum: 251-261.  Ganzeboom, Harry B., Paul M. De Graaf, and Donald Treiman. 1991. “A Standard International Socio-Economic Index of Occupational Status.” Social Science Research 21: 1-56.  Giddens, Anthony. 1973. The Class Structure of Advanced Societies. London: Cambridge University Press. Gilbert, Dennis. 1998. The American Class Structure in an Age of Growing Inequality. Belmont: Wadsworth.  Gilens, Martin and Benjamin Page. 2014. “Testing Theories of American Elites, Interest Groups, and Average Citizens.” Perspectives on Politics 12: 564-581. Gilens, Martin. 1999. Why Americans Hate Welfare. Chicago: University of Chicago Press. Gilens, Martin. 2009. “Preference Gaps and Inequality in Representation.” Political Science and Politics (April): 335-341.  Gilens, Martin. 2012. Affluence and Influence. Princeton: Princeton University Press.  Goldthorpe, John. 1999. “Modeling the Pattern of Class Voting in British Elections.” In The End of Class Politics?, ed. Geoffrey Evans. Oxford: Oxford University Press. Goldthorpe, John. 2000. On Sociology: Numbers, Narratives and Integration of Research and Theory. Oxford: Oxford University Press.  Goldthorpe, John.. 1983. “Women and Class Analysis.” Sociology 17 (4): 465-488.  37  Gorz, Andre. 1982. Farewell to the Working Class. Trans. Michael Sonenscher. London: Pluto Press. Grusky, David and Kim Weeden. 2005. “Are There Social Classes? A Framework for Testing Sociology’s Favorite Concept.” In Social Class: How Does it Work?, eds. Annette Lareau and Dalton Conley. New York: Russell Sage Foundation. Grusky, David and Manwai Ku. 2008. “Gloom, Doom and Inequality.” In Social Stratification.3rd ed. ed. David Grusky. Philadelphia: Westview Press: 2-27. Hacker, Jacob S. and Paul Pierson. 2010. Winner Take All Politics. New York: Simon and Schuster.  Hauser, Robert and John Robert Warren. 1997. “Socioeconomic Indexes of Occupational Status: A Review, Update, and Critique.” In Sociological Methodology 27: 177-298. Hooker, Juliet and Alvin B. Tillery Jr. 2016. “Executive Summary.” In The Double Bind: The Politics of Racial and Class Inequalities in the Americas. Hout, Jeff, Michael Brooks and Clem Manza. 1999. “Classes, Unions and the Realignment of US Presidential Voting.” In The End of Class Politics?, ed. Geoffrey Evans. Oxford: Oxford University Press. Inglehart, Ronald. 1990. Culture Shift in Advanced Industrial Democracy. Princeton: Princeton University Press. Inglehart, Ronald. 1997. Modernization and Postmodernization in 43 Countries. Princeton: Princeton University Press.  Jacobs, Lawrence and Theda Skocpol, eds. 2005. Inequality and American Democracy. New York: Russell Sage Foundation. James, Brendan. 2014. “Princeton Study: US No Longer an Actual Democracy.” Talking Points Memo. April 18.  Jaschik, Scott. 2012. “Faculty Pay, Around the World.” March 22.  https://www.insidehighered.com/news/2012/03/22/new-study-analyzes-how-faculty-pay-compares-worldwide (13 August 2016).  Krugman, Paul. 1996. “The Spiral of Inequality.” Mother Jones. November-December.  Langbein, Laura. 1986. “Money and Access: Some Empirical Evidence.” The Journal of Politics 48 (4): 1052-1062.  38  Lareau, Annette. 2008. “Taking Stock of Class.” In Social Class: How Does it Work?, eds. Annette Lareau and Dalton Conley. New York: Russell Sage Foundation.  Larson, Magali Sarfatti. 1977. The Rise of Professionalism: A Sociological Perspective. Berkeley: University of California Press. Lipset, Seymour Martin and Taylor Nichols Clark. 1991. “Are Social Classes Dying?” International Sociology 6 (4): 397-410.  Lipset, Seymour Martin. 1960. Political Man. New York: Doubleday. Lubrano, Alfred. 2004. Limbo. Hoboken: John Wiley and Sons. Massey, Douglas. 2007. Categorically Unequal: The American Stratification System. New York: Russell Sage Foundation. Micklewright, John and Sylke V. Schnepf. 2007.  “How Reliable Are Income Data Collected with a Single Question? #3177.” Institute of the Study of Labor Working Paper Series. Mills, C.W. 1956. The Power Elite. Oxford: Oxford University Press. Mullainathan and Shafir. 2010. “Savings Policy and Decisionmaking in Low-Income Households.” In Insufficient Funds: Savings, Assets, Credit, and Banking Amongst Low-Income Households, eds. Rebecca Blank and Michael Barr. New York: Russell Sage.  Oakes, Michael. 2012. “The Measurement of SES.” NIH OBSSR Online Textbook Research Methods for Health Research.  Page, Benjamin, Larry Bartels, and Jason Seawright. 2013. “Democracy and the Policy Preferences of Wealthy Americans.” Perspectives on Politics 11 (March): 51-73. Pakulski, Jan and Malcolm Waters. 1996. The Death of Class. London: SAGE Publishing. Perruci, Robert and Earl Wysong. 2007. New Class Society: Goodbye American Dream?. Lanham: Rowman and Littlefield. Piketty, Thomas and Emmanuel Saez, 1999. “The Evolution of Top Incomes: A Historical and International Perspective.”  American Economic Review 96: 200-205.  Piketty, Thomas. 2014. Capital in the 21st Century. Cambridge: Belknap.  Poulantzas, Nicos. 1973. Political Power and Social Classes. London: NLB.   39  Provine, Doris Marie. 1986. Judging Credentials. Chicago: University of Chicago Press. Rogot, Eugene, Paul D. Sorlie, and Norman J. Johnson. 1992. "Life Expectancy by Employment Status, Income, and Education in the National Longitudinal Mortality Study." Public Health Reports :457. Sadin, Meredith. Forthcoming. “Campaigning with Class: The Effect of Candidate Social Class on Voter Evaluations.” Working paper. Savage, Mike and Karel Williams. 2008. “Elites: Remembered in Capitalism and Forgotten by Social Sciences.” The Sociological Review 56 (May): 1-24. Savage, Mike, Fiona Devine, Niall Cunningham, Mark Taylor, Yaojun Li, Johs. Hjellbrekke, Brigitte Le Roux, Sam Friedman and Andrew Miles. “A New Model of Social Class? Findings from the BBC’s Great British Class Survey Experiment.” Sociology 47 (2): 219-250.  Schlozman, Kay Lehman, Benjamin Page, Sidney Verba, and Morris Fiorina. 2005. “Inequalities in Political Voice.” In Inequality and American Democracy , eds. Jacobs, Lawrence and Theda Skocpol. New York: Russell Sage Foundation. Schlozman, Kay Lehman, Sidney Verba and Henry E. Brady. 2012. The Unheavenly Chorus: Unequal Political Voice and the Broken Promise of American Democracy. Princeton: Princeton University Press.  Sen, Amartya. 2008 [1997]. “ From Income Inequality to Economic Inequality.” Southern Economic Journal 64: 383-401.  Skocpol, Theda. 1992. Protecting Soldiers and Mothers. Cambridge: Harvard University Press. Sorenson, Aage B. “Foundations of a Rent-based Class Analysis.” In Approaches to Class Analysis, ed. Erik Olin Wright. Cambridge: Cambridge University Press. Sorenson, Aage B. “On the Usefulness of Class Analysis in Research on Social Mobility and Socioeconomic Inequality.”  Acta Sociologica 34 (April): 71-87. Soroka, Stuart and Christopher Wlezien. 2008. "On the Limits to Inequality in Representation.” PS: Political Science and Politics (April): 319-27. Stanworth, Michelle. 1984. “Women and Class Analysis: A Reply to John Goldthorpe.” Sociology 18 (2): 159-170.  Taylor, Charles. 1994. Multiculturalism. Princeton: Princeton University Press.   40  Wright, Erik Olin. “Foundations of a neo-Marxist Class Analysis.” In Approaches to Class Analysis, ed. Erik Olin Wright. Cambridge: Cambridge University Press. Wright, Erik Olin. 1979. Class Structure and Income Determination. New York: Academic Press.  Wright, Erik Olin. 1985. Class, Crisis and the State. New York: Verso. Wright, Erik Olin. 1997. Class Counts. Cambridge: Cambridge University Press. Wright, Erik Olin. 2003. “Social Class.” In Encyclopedia of Social Theory, ed. George Ritzer. New York: SAGE Publications. Young, Iris Marion. 1990. Justice and the Politics of Difference. Princeton: Princeton University Press. Zuesse, Eric. 2014. “US is an Oligarchy  Not a Democracy, says Scientific Study.” Common Dreams. April 14.                      


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items