Process matters not just for diagnosing the causes of inequality, but also for how policy is shaped. The dominant paradigms for policy-making – neoliberalism, neo-Keynesianism, and neopaternalism – largely address inequality via “outcome-policies” that manipulate the levers of government and, more recently, draw on randomized trials and “nudges” to change behavior, in a manner that is not only easy to measure, but also easy to reverse. This commentary draws on the essays in this special issue of Dædalus to make the case for “reflectivism,” which shifts structural inequalities in agency, power, social structure, empathy, and aspiration in an incremental manner that is more uncertain and difficult to measure, but that can result in more lasting change.

The essays in this special issue of Dœdalus represent a major attempt to move the diagnosis of inequality from a static to a process view. Several important themes emerge from this lens: 1) the interaction between economic, social, and cultural processes in generating inequality and the centrality of the need for interdisciplinary analysis; 2) four approaches through which inequality-generating processes might work – evaluation, quantification, commodification, and policy drift; 3) the linkages between micro-, meso-, and macrolevels of analysis; and 4) the importance, in thinking of inequality-generating processes, of taking a longue durée view. For the most part, however, the emphasis of the essays in this collection is on diagnosing the causes of inequality, rather than finding solutions.1 Moreover, they are based on an analysis of conditions in North America and Western Europe, where the concern is much more on the shift, over the last few decades, from equality-generating to inequality-generating processes. This is different from the conditions in countries that are home to most of the world's poor, whose rise in inequality has been coupled with large reductions in poverty and relative expansions of social safety nets over the last two decades. But what is common to both the global North and South is the threat to democratic institutions and processes. In this commentary, I address some of these gaps and reflect on how interdisciplinary process-thinking may result in a paradigm shift in how we think about policy. While I will draw on experiences and evidence from the developing world, these points may be of broader relevance.

The manner in which we translate the desire to create a more equal world with less poverty into action is shaped by how we organize and think about the world, how we approach causality, and how we make abstract ideas legible by methods of measurement and categorization. Over the last century, the policy options available have tended to focus on what I call outcome-policies, as opposed to process-policies.

An outcome-policy focuses on manipulating the levers of government – taxes, expenditures, regulations, systems of implementation – to achieve certain outcomes. The success of the policy is judged by those outcomes. Most policies to combat inequality and poverty fall under this rubric. Think, for example, of conditional cash transfers (whose success is measured by the extent to which the transfers encourage people to satisfy the “condition,” such as keeping children in school), increasing estate taxes (measured by the extent to which they increase intergenerational social mobility), or providing health insurance (measured by effects on public health and economic hardship). Outcome-policies are thus closely tied to metrics, which also inform debates over their efficacy. If the political environment is favorable, outcome-based policies are relatively easy to institute by legislative or executive action. They are also easy to reverse.

A process-policy attempts to shift the trajectory of change. Its effects can be more difficult to measure because its proximate impact is often subtle. It focuses on equalizing power relationships by shifting the process of decision-making in favor of the less privileged, and on the incremental change of one step building on the last. The full impact of a process-policy generally takes a much longer time to reveal itself but can be longer-lasting because it reduces the inequalities and imperfections in how decisions are made. Examples include systems of deliberative decision-making that have constitutional sanction, political reservations for discriminated minorities and women, and adaptive systems of project implementation.

Policies for human welfare, growth, inequality, and poverty are broadly conditioned by three paradigms that largely derive from economics. Like all ideal types, these categories are not mutually exclusive but intersect in a variety of ways and share many elements. They are for the most part outcome-policy oriented, though they also include some process-policies. To these, I propose adding a fourth paradigm that is more shaped by political and social theory, intersecting with and also complementing the three dominant approaches. My goal is not to advocate for the “best” way of approaching the problem – I do not believe there is any such thing – but to give more weight to an approach that has been relatively neglected.

First, the neoliberal paradigm that grows from deep skepticism about the capacity of social science to understand complex human interactions, and thus the ability of policy to engineer a better world. This approach emphasizes the key principles of laissez-faire markets, free trade, human rights, and electoral democracy. The idea is that free markets result in high rates of economic growth, while electoral democracy ensures that governments are held accountable and create the conditions for growth.2 This approach underlies the “Washington consensus” that prevailed in the development policy world through the 1980s and 1990s and that over the long term, recent evidence has shown, has led to higher rates of growth.3

The neoliberal approach to poverty reduction argues that maximizing economic growth and improving and equalizing access to human capital allows every individual to benefit from the growth and be liberated from poverty. There is some evidence in favor of this,4 but also evidence showing that “the growth incidence of poverty” – the effect that a 1 percent increase in economic growth has on relative reductions in poverty – is conditioned by inequality.5 There is also strong reason to believe that neoliberal policies increase inequality and reduce social support, for instance, by creating “opportunity markets” that commodify and sharply restrict access to basic needs such as education and housing.6

The driving discipline behind neoliberalism is rational-choice economics, which provides a consistent framework to think about growth, welfare, poverty, and equality of opportunity. It derives from a worldview that is methodologically individualistic with the central goal of ensuring that markets function as efficiently as possible. Individual freedoms are central to the approach, and democratic accountability through elections is a crucial counterpoint to market efficiency. Rational-choice methodological individualism is not easily reconciled with notions of social exclusion or cultural and political inequality. Thus, any argument for socially driven policy-making – to promote gender equality or social cohesion, for instance – within a neoliberal paradigm has to be filtered through a growth prism. Hence the slew of papers from the 1990s on the positive effects of women's education and social capital on growth.7

Welfare is measured through either an income metric or consumption metric. This has spawned a large industry measuring individual- and household-level income and consumption through household surveys, which are then used to calculate poverty rates via a poverty line defined in terms of dollars and cents. Note that the neoliberal paradigm, by focusing on core principles rather than policy proposals, is primarily process-driven, rather than driven by outcomes. However, the processes are largely devoted to creating an environment for prosperity rather than economic equality.

The second approach, which I will broadly term neo-Keynesian, is a counter to the perceived limitations of neoliberalism. It is much more cognizant of market failure and convinced of the ability of the social sciences to inform policy solutions that deal with market failure. With policies informed by growth models that emphasize investments in physical and human capital, growth is still a priority, but with an awareness of the central role played by ideas and information. There is acute awareness of the inability of markets to deliver basic services to the poor and of systemic discrimination. Poverty is still a welfarist metric, but along with data on income and consumption, household surveys now collect data on gender, religion, race, and caste to enable analyses of discrimination. There is broad recognition of the need for equality of opportunity and for the necessity of legal interventions to ensure it. Strong support for affirmative action in employment and government is a key element.

All this provides the justification for more interventionist governments, as shown in the important slate of recommendations to tackle inequality made by the economist Tony Atkinson.8 Atkinson's recommendations include directing technological change to favor more employment, increased minimum wages, a capital endowment fund that ensures that everyone has a minimum inheritance, and progressive property taxes. Governments are expected to invest in infrastructure, education, and public access to better health, nutrition, and sanitation. Governments are expected to ensure effective safety nets so the vulnerable are not subject to acute distress. The enhanced role for the government also results in an emphasis on “good governance” and thus greater awareness of the importance of equalizing access to information and improving public service delivery. This leads to a greater appreciation of government failure and, in particular, of clientelism and difficulties in ensuring the interests of women and minorities. Neo-Keynesianism is thus the archetype of outcome-policy-making.

The third approach, though it is arguably a more stringent extension of the second, is called neopaternalism.9 Neopaternalism, which has exploded in influence over the last decade, starts from exactly the other end of neoliberalism in its abiding belief in the power of social science and data, particularly behavioral experiments and randomized trials, to give direction to policy, and the use of “big data” for diagnosis and surveillance. It takes its cues from public health with a firm commitment to “evidence-based” policy-making. It draws on research from behavioral science on “scarcity” to argue that the poor face large constraints on their time and attention that direct them to make suboptimal choices.10 Thus, the freedom to choose is not a meaningful option when the ability to choose is itself severely constrained. This provides the justification for a top-down push to shift people away from perceived harmful actions, norms, and practices.11

This then requires policy-makers – governments and ngos – to find ways of doing things that the poor cannot do for themselves. What those things are, however, depends on whether the policy option has been vetted by “rigorous” evidence. What follows is a synergistic relationship between research and policy-making: an industry of social scientists testing the efficacy of various interventions around the world with surveys and experiments. One advantage of this is that it generates vast amounts of data that shed light on a wide variety of outcomes that affect the poor. While income and consumption definitions of poverty are still prevalent, there is now a broader understanding of the conditions faced by the poor. Another advantage is that it creates a culture in which policy has to be justified with evidence rather than hope or idealism. And a third advantage is that it focuses attention on the efficacy of intervention design, weeding out good designs from bad ones.

The disadvantages, however, have been well cataloged by scholars.12 Here, I want to focus on two that are central to the case for a fourth paradigm. Neopaternalism has, within it, the danger of overreach. Policy can be driven by the narrow demands of scientific technique, rather than scientific technique being driven by the needs of policy: a case of the tail wagging the dog. If all evidence is to be passed through the test of a behavioral experiment or a randomized trial, then policy interventions that are not amenable to experimentation and randomization will be gradually eliminated from consideration. Moreover, a good randomized trial requires good implementation because imperfect implementation would contaminate the design of the experiment; it would be difficult to untangle the effect of the design from the effect of the implementation. This means that trials, in essence, assume away the challenge of implementation: the complicated process by which policy ideas are converted into action on the ground. This deemphasizes what scholars since Albert Hirschman have argued is the critical problem of policy-making, particularly in poor countries.13

Behavioral experiments and randomized trials, moreover, have trouble dealing with high degrees of contextual variation. An experiment or intervention that works in one region will not necessarily work in another: the well-known problem of external validity. Advocates for randomization have dealt with this by conducting trials across a variety of countries to test the efficacy of a design, or by conducting randomized trials at a very large scale.14 But this solution gets much more difficult with “complex” interventions, such as those tailored for and, in particular, tailored by communities at the microlevel. Such designs are very difficult to evaluate with standard techniques because wide variations in design make sample sizes with adequate statistical power untenable for reasons of cost and manageability. Neopaternalism, therefore, intrinsically favors interventions that are simple, do not require much adaptation on the ground, and have predictable outcomes.

Neopaternalism, while cognizant of social norms and the contribution of structural constraints to mobility and the reproduction of inequality, is far more comfortable with certainty than it is with reflexivity. Nonrational action is adduced via behavioral experiments, with cultural “traits” and social norms seen as good or bad. There is, implicitly, little patience for contextual complexity, participant observation, or participant engagement to discover the complex interactions between cultural and social process, economic conditions, and politics that determine persistent inequality. The goal is to discover broad “truths” to inform policies to design “nudges” to move people out of what are seen as harmful or dysfunctional behaviors. The analysis of culture and social interactions is, therefore, part of the scientific apparatus of the expert, to nudge people toward improvements in welfare, as the experts define it.

It is important to note that all three paradigms rely on the checks, balances, and accountability mechanisms of electoral democracy. Neoliberalism and neo-Keynesianism are both closely intertwined with the notion that universal franchise will be enough to keep government actions aligned with the needs and interests of citizens. Neopaternalism takes this reliance one step further. What legal scholar Cass Sunstein calls “liberal paternalism” requires that governments do not use the power of behavioral nudges to impinge on the fundamental freedoms of citizens.15

But the limitations of the electoral mechanism as a way of governing large, complex societies have increasingly become apparent, with challenges like elite capture and clientelism taking influence throughout the world.16 This weakening of electoral democracy as a mechanism to check elite dominance is partly responsible for unleashing processes of commensuration, commodification, and policy drift that have reinforced inequality-generating processes.17 Moreover, we live in an age when big data is increasingly being used toward a new form of “surveillance capitalism,” causing widespread concern that without urgent regulatory measures, the ability to track and monitor people with extremely precise data will increasingly lead to the delegitimization of democratic processes.18

Moreover, democracy “has come to mean much more than free and fair elections, though nothing less.”19 Throughout the world, at every level – micro, meso, and macro – people have organized themselves into bodies to monitor those in power, in what political theorist John Keane has called “monitory democracy.” To mention just a few, such institutions include citizen committees, self-help groups, minipublics, environmental groups, think tanks, and organizations such as the Bretton-Woods Project, which monitors the imf and World Bank. Moreover, such institutions can exist even in more authoritarian settings like China, where citizens groups, often with the active concordance of civil servants, have created “accountability without democracy” working within systems of “authoritarian deliberation.”20 Such forums for citizen engagement and participation do not always emerge organically but are induced by policy interventions, with development organizations spending hundreds of billions of dollars on such projects with the hope that they will alleviate poverty and empower the excluded.21

It is clear that a new paradigm of action has emerged that relies much more on process-policy than on outcome-policy, particularly in comparison with the three dominant paradigms. It may be time to give this paradigm a label and sketch its basic characteristics.

Economists think about inequality largely through the lens of equality of opportunity. A process perspective suggests that this is not sufficient to deal with the relational aspects of deprivation. As Michèle Lamont and Paul Pierson argue in their contribution to this issue of Dœdalus, the processes that reproduce, intensify, or deepen inequality are often distinct from those that generated it, and are fundamentally influenced by social relationships.22 Furthermore, in arguing that human beings are primarily held back by exogenous obstacles and not endogenous processes, equality of opportunity implicitly assumes a distinction between preferences and constraints. Equality of opportunity, consequently, needs to be supplemented by an “equality of agency” that takes into account the impact of the relationality of individuals; the political, social, and cultural contexts within which they operate; and the impact of these processes on power differences, inequality, and poverty.23

The reflective paradigm for policy-making follows from this. It recognizes that even if someone is lifted above an externally defined poverty line, they may still be subject to vulnerability, discrimination, and exclusion because they lack voice, agency, and what anthropologist Arjun Appadurai has called the “capacity to aspire,” the ability to envision a future different from what they see around them.24 It is fundamentally about redrawing boundaries and shifting “norms of interaction.”25 It allows for the fact that discrimination is not just material, it is also “epistemic” in the sense that the capacity to speak, and be listened to, is also unequally distributed.26

The term reflective builds on political theorist Robert Goodin's notion of “reflective democracy,” the idea that an effective democracy needs to inculcate the capacity for individuals to “deliberate within.”27 Deliberating within creates the space for what sociologist Mario Luis Small has called “cognitive empathy”: “the ability to understand another person's predicament as they understand it.”28 But it is a fundamental challenge to do this at the scale of entire communities and countries, to change processes of decision-making so that the powerful and the less powerful – politicians, government officials, and citizens alike – all develop horizontal and vertical cognitive empathy.

Many social movements, governance innovations, and policy interventions around the world are attempting to turn this paradigm into practice. While there are large variations in the specifics of their actions and objectives, some important processes are worth noting:

  • Filling in the blank space between elections by fostering continuous dialogue between citizens and governments. This requires the creation of systems of deliberative decision-making.29 The “system” usually has some kind of officially sanctioned civic space – a forum or a regularly scheduled public meeting – where the average citizen is given a chance to influence directly public decisions that have a bearing on their lives. But forums alone are not enough; they need to be embedded within a culture of dialogue, debate, and discussion in which the goal is to make the act of speaking and listening an everyday practice.

  • Equalizing power in decision-making by giving voice to disadvantaged groups and, more radically, by reserving a percentage of seats in legislatures for representatives from such groups.

  • Creating feedback loops within governments, and between governments and citizens, in which decision-making becomes adaptive and incremental.30 This helps governments deliver better public services and respond to the needs of citizens.

  • Using technology in a way that gives people control over their own data to strengthen democratic processes, facilitate collective action, and equalize access to public services.31

Implementing policies that do this at scale is neither easy nor predictable; indeed, it is not always measurable. As Jane Jenson, Francesca Polletta, and Paige Raibmon show in their essay in this volume, to get this right requires an understanding of policy not as a one-shot deal, but as a process of constant adaptation, through which successes and failures provide lessons for incremental change.32 Processes take a long time to show “hard” results because they shift trajectories of change in unpredictable ways. Thus, it is important to analyze shifts in processes like how decisions are made, in power relations and in narratives and discourse. These are difficult to measure in conventional ways and require us to draw on qualitative methods and narrative analysis, including natural language processing methods.33 It also requires coordination across quantitative and qualitative methods, for instance by conducting mixed-methods evaluations.34 The process of adaptive, citizen-engaged policy-making also helps avoid the problem of “policy drift,” when policies are fixed in the past rather than adapted to changing times.

What are the disadvantages? The reflective paradigm requires that policymakers take a very long time horizon. It requires a tolerance of mess because it attempts to reverse inequalities in power and agency, which is, at best, a contentious and uncertain process. Perhaps the most significant challenge is that governments (and other quasigovernment actors like the World Bank and donor agencies) are not set up to work with process-policies.35 The logic of bureaucracies and the political environments in which they function make them much more focused on interventions that can induce a quick, measurable impact that does not threaten the political status quo.

The essays in this collection make an important contribution toward rethinking the diagnosis of inequality to help reverse inequality-generating processes. I have tried in this commentary to draw on some of the ideas in these essays, along with other literatures, to make the case for a new reflective paradigm for policy-making that focuses on process-policies, rather than outcome-policies, that can supplement neoliberal, neo-Keynesian, and neopaternalistic paradigms that have dominated how we think about policy. In practice, policies at any given time in any given country will draw on all four paradigms; and it is not clear that any one paradigm is clearly superior to any other. They have different goals with different methods of execution. However, the reflective paradigm tends to be neglected in policy circles precisely because it is not easy to measure.

Process-policies usually come about as the result of years of efforts by activists working organically, usually at a small scale, within social movements or nongovernmental organizations. Given contemporary concerns about unstable democratic institutions and inequality, it may be time to think about how to bring them to the mainstream.

AUTHOR'S NOTE I am grateful to Monica Biradavolu, Michèle Lamont, Paul Pierson, J. P. Singh, and Mike Walton for their valuable comments on this essay.


David B. Grusky, Peter A. Hall, and Hazel Rose Markus, “The Rise of Opportunity Markets: How Did It Happen & What Can We Do?” Dœdalus 148 (3) (Summer 2019); and Jane Jenson, Francesca Polletta, and Paige Raibmon, “The Difficulties of Combating Inequality in Time,” Dœdalus 148 (3) (Summer 2019) are, to some extent, exceptions.


A view that now has some empirical support; see Daron Acemoglu, Suresh Naidu, Pascual Restrepo, and James Robinson, “Democracy Does Cause Growth,” Journal of Political Economy 127 (1) (2019): 47–100.


William Easterly, “In Search of Reforms for Growth: New Stylized Facts on Policies and Growth Outcomes” (2018),


David Dollar and Aart Kraay, “Growth is Good for the Poor,” Journal of Economic Growth 7 (3) (2002): 195–225.


Martin Ravallion, “Growth, Inequality and Poverty: Looking Beyond Averages,” World Development 29 (11) (2001): 1803–1815.


Peter A. Hall and Michèle Lamont, eds., Social Resilience in the Neoliberal Era (New York: Cambridge University Press, 2013). On opportunity markets, see Grusky et al., “The Rise of Opportunity Markets”; and Patrick Le Galès and Paul Pierson, “‘Superstar Cities’ & the Generation of Durable Inequality,” Dœdalus 148 (3) (Summer 2019).


For example, Lawrence H. Summers, “Investing in All the People,” World Bank Policy Research Working Paper No. 905 (Washington, D.C.: The World Bank, 1992); and Stephen Knack and Philip Keefer, “Does Social Capital Have an Economic Payoff? A Cross-Country Investigation,” Quarterly Journal of Economics 112 (4) (1997): 1251–1288.


Anthony B. Atkinson, Inequality: What Can be Done? (Cambridge, Mass.: Harvard University Press, 2015).


Cass R. Sunstein, Why Nudge? The Politics of Libertarian Paternalism (New Haven, Conn.: Yale University Press, 2014).


Sendhil Mullainathan and Eldar Shafir, Scarcity: Why Having Too Little Means So Much (New York: Times Books, 2013).


Richard H. Thaler and Cass R. Sunstein, Nudge: Improving Decisions about Health, Wealth, and Happiness (New Haven, Conn.: Yale University Press, 2008); and The World Bank, World Development Report 2015: Mind, Society, and Behavior (Washington, D.C.: The World Bank, 2015).


For example, Angus Deaton and Nancy Cartwright, “Understanding and Misunderstanding Randomized Controlled Trials,” Social Science & Medicine 210 (2018): 2–21; and Martin Ravallion, “Should the Randomistas (Continue to) Rule,” Center for Global Development Working Paper No. 492 (Washington, D.C.: Center for Global Development, 2018).


Albert O. Hirschman, Development Projects Observed (Washington, D.C.: Brookings Institution Press, 1967).


Karthik Muralidharan and Paul Niehaus, “Experimentation at Scale,” Journal of Economic Perspectives 31 (4) (2017): 103–124.


Sunstein, Why Nudge?


Jacob S. Hacker and Paul Pierson, Winner-Take-All Politics: How Washington Made the Rich Richer – And Turned Its Back on the Middle Class (New York: Simon and Schuster, 2010); and Pranab Bardhan and Dilip Mookherjee, “Clientelistic Politics and Economic Development: An Overview,” Economic Development and Institutions Working Paper Series No. WP16/10.III.5 (Oxford: Economic Development and Institutions, 2016).


Michèle Lamont and Paul Pierson, “Inequality Generation & Persistence as Multidimensional Processes: An Interdisciplinary Agenda,” Dœdalus 148 (3) (Summer 2019).


Shoshana Zuboff, “Big Other: Surveillance Capitalism and the Prospects of an Information Civilization,” Journal of Information Technology 30 (1) (2015): 75–89; and Dirk Helbing, Bruno S. Frey, Gerd Gigerenzer, et al., “Will Democracy Survive Big Data and Artificial Intelligence,” Scientific American, February 25, 2017.


John Keane, Power and Humility: The Future of Monitory Democracy (New York: Cambridge University Press, 2018), 105.


Lily Tsai, Accountability Without Democracy: Solidary Groups and Public Goods Provision in Rural China (New York: Cambridge University Press, 2007); and Baogang He and Mark E. Warren, “Authoritarian Deliberation in China,” Dœdalus 146 (3) (Summer 2017): 155–166.


Ghazala Mansuri and Vijayendra Rao, Localizing Development: Does Participation Work (Washington, D.C.: The World Bank, 2012).


Lamont and Pierson, “Inequality Generation & Persistence as Multidimensional Processes.”


Vijayendra Rao and Michael Walton, “Culture and Public Action: Relationality, Equality of Agency and Development,” in Culture and Public Action, ed. Vijayendra Rao and Michael Walton (Stanford, Calif.: Stanford University Press, 2004).


Arjun Appadurai, “The Capacity to Aspire: Culture and the Terms of Recognition,” in Culture and Public Action.


Irene Bloemraad, Will Kymlicka, Michèle Lamont, and Leanne S. Son Hing, “Membership without Social Citizenship? Deservingness & Redistribution as Grounds for Equality,” Dœdalus 148 (3) (Summer 2019).


Miranda Fricker, Epistemic Injustice: Power and the Ethics of Knowing (Oxford: Oxford University Press, 2007).


Robert E. Goodin, Reflective Democracy (Oxford: Oxford University Press, 2003).


Mario Luis Small, “Rhetoric and Evidence in a Polarized Society,” Coming to Terms with a Polarized Society Lecture Series, Institute for Social and Economic Research and Policy, Columbia University, New York, March 1, 2018.


John Parkinson and Jane Mansbridge, eds., Deliberative Systems: Deliberative Democracy at the Large Scale (New York: Cambridge University Press, 2012).


Matt Andrews, Lant Pritchett, and Michael Woolcock, Building State Capability: Evidence, Analysis, Action (Oxford: Oxford University Press, 2017).


For more on this, see Vijayendra Rao, “Reflectivism: A Process Paradigm for Addressing Poverty and Inequality,” mimeo, Word Bank. For videos, data, and visualizations that show an application of “citizen-generated data,” see The World Bank, “Democratizing Data,”


Jenson et al., “The Difficulties of Combating Inequality in Time.”


Justin Grimmer and Brandon M. Stewart, “Text as Data: The Promise and Pitfalls of Automatic Content Analysis Methods for Political Text,” Political Analysis 21 (3) (2013): 267–297.


Vijayendra Rao, Kripa Ananthpur, and Kabir Malik, “The Anatomy of Failure: An Ethnography of a Randomized Trial to Deepen Democracy in Rural India,” World Development 99 (2017): 481–497.


Mansuri and Rao, Localizing Development.

This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. For a full description of the license, please visit