So often in public policy debates we find ourselves sorted into political tribes that simply talk past each other. Welfare reform provides one of the strongest examples.

Earlier this year a well known poverty and welfare policy analyst was fired from his job after insulting the president and CEO of an important and respected progressive think tank. The attack incited hoards of his Twitter fans to join in on the harassment of the CEO and her defenders. The CEO’s sin? An unsubstantiated report that over 20 years ago she may (may) have favored and/or worked on President Bill Clinton’s 1996 welfare reform. This tainted her as insufficiently progressive, and thus worthy of effigy.

But rather than dwell on the negative, I’m here to highlight an example from the poverty and welfare space of what public discourse can and should be.

Debating Poverty After Welfare Reform

Last month Scott Winship of the Manhattan Institute (and Niskanen Center advisor) published a magisterial report titled “Poverty after Welfare Reform” analyzing trends in poverty, particularly child poverty, following the creation of TANF. His report is in large part a direct response to $2 a Day by H. Luke Shaefer and Kathryn J. Edin, a blockbuster book that claims extreme poverty—which they define as living on less than $2 a day—has seen a dramatic rise in the United States since the passage of welfare reform.

Winship argues convincingly that the metrics used by Shaefer and Edin omit a number of major sources of cash and cash-like income, fail to adjust for income under-reporting, and don’t sufficiently incorporate the value of publicly provided health care. After making adjustments to the official poverty rate for these and other factors (each of which Winship goes to great pains to justify), he shows that the percent of people in poverty is at an all-time low, even among children, and that extreme poverty has shown no clear upward trend.

welfare reform civil discourse

From Scott Winship, “Poverty After Welfare Reform”

Shortly after the publication of Winship’s report, Jordan Weissmann of Slate wrote a column in reply titled “The Odd Conservative Argument That Food Stamps and Medicaid Saved the Poor From Welfare Reform.” Weissmann, a staunch critic of the 1996 welfare reform, seemed to be largely persuaded of Winship’s argument about the trend in headline poverty, but disputed his narrative around cause and effect. Poverty fell, Weissmann contends, because other entitlement programs expanded to fill the gap created by TANF, not because TANF worked miracles.

To be fair to Winship, his report is almost exclusively a descriptive analysis of trends, and does not make strong causal claims about why poverty fell the way it did. Nonetheless, Winship took up Weissmann’s challenge, and wrote a defense of TANF on his blog at Forbes and for the National Review. There Winship argued that TANF had a direct link to falling poverty rates because it promoted work and independence, thereby increasing the number of families earning an income.


Genius lets you annotate articles, line by line, promoting civil discourse like this.

Shawn Fremstad, a Senior Fellow at the Center for American Progress, read the follow-up National Review piece and decided to do a line-by-line rebuttal using the annotation service, Genius. After engaging each other on Twitter, Winship agreed to reply to each rebuttal in-line, as well.

What resulted defies belief: Two passionate welfare analysts on opposites sides of an intensely tribal issue having one of the most substantive and polite welfare reform policy discussions I’ve ever read—and on a platform originally built for interpreting rap lyrics!

I have my own criticisms of TANF from a libertarian and conservative perspective, and am not fully convinced that extreme poverty has fallen in this country (although Winship has convinced me to be more skeptical of strong statements either way). But in this debate, as a partial bystander, who is ultimately right or wrong on the specifics is not what I see as important.

Rather, I want to celebrate Winship’s and Fremstad’s Genius dialog on welfare reform as a pinnacle of polite and substantive debate. Let them be an example to us all, and let’s hope they start a trend.

The Value of Civility

Civility in discourse is not “tone policing” or stacking the deck against minority points of view, as some have tried to argue. There are good reasons to feel anger and passion about public policy issues, but it’s important to consider how always being in that “hot” state of mind can contribute to confirmation bias and talking past, rather than empathizing with, your interlocutor.

Norms of civility are simply essential for taming our tribal predispositions, listening to opposing points of view, persuading others, and arriving at rational consensus. This is why courtrooms enforce decorum and turn taking in argumentation, and why committees and legislative houses have rules of order. Civility does not come natural to us immodest apes, so we employ what psychologists, borrowing from computer science, refer to as “kludges”: Inelegant solutions to underlying imperfections, like the social scaffolding (rules, norms, procedures, and systems) we use to keep our monkeys brains in check.

In the days of letter writing, for example, the inherent constraints of the medium forced interlocutors to pause and reflect on each other’s argument. For centuries, the Catholic Church made use of a “Devil’s Advocate” as a check against group-think when debating canonization. Other debate formats make use of time constraints and neutral moderators to force arguments to be both concise and on topic.

When the medium does away with these forms of scaffolding altogether the results are predictably regressive. The incentives of Twitter, for example, seem if anything to discourage civility and amplify our sweet tooth for self-righteous indignation. The annotation service, Genius, in contrast, allows for long form, sentence clause by sentence clause deconstruction of arguments.

Aumann’s agreement theorem states that two people acting rationally and with common knowledge of each other’s beliefs should never agree to disagree. That is, they should, through reflexive and open discourse, converge on the same belief. Yet this rarely if ever is the case in public policy debate.

In the internet age persistent disagreement cannot easily be explained by access to different information. But nor can the failure of the agreement theorem be explained by arguing for our inherent irrationality. Instead, we must recognize how rationality is in large measure not in our heads, but in the structures, incentives and norms that govern our conversations with one and other, and push disagreement in a productive direction.

We will never reach agreement through vitriolic bursts of 140 characters. Indeed, quite often the resultant name calling and misreadings cascade into permanently burnt bridges.

But if we truly value truth, and by extension justice, then we should follow Winship, Fremstad, and others in embracing technologies and mediums that promote inter-ideological discourse, and celebrate it whenever we can.