Blogging the APPAM Conference

The APPAM conference is the annual gathering of public policy academics. It brings together professor, researcher, and think tank types. This year the meeting was held last week in Boston, where it was rainy and cold. Good thing, there were many excellent sessions to attend, including an entire track of panels devoted to asset building topics. So, basically there were scores of interesting presentations and hallway conversations. I thought I would share a few of my observations along the way.

First stop was a session on “evidence-based policymaking,” where I listen to Jeff Liebman talk about randomized controlled social policy experiments. He recently returned to Harvard after a two year stint as the number 2 at OMB, but was a leading force behind the Obama administration’s efforts to promote evidenced-based policymaking. This is another one of those below the radar but excellent initiatives that team Obama has led. Essentially, they have proposed creating a pool for money available to agencies with plans to rigorously evaluate new policy ideas. Of course, there remain challenges in getting these resources appropriated by Congress, but it is supporting a major culture change in government. I’m a big proponent of both research and evidence-based policymaking. That said, often the challenge is not to get pilots off the ground, but to use their findings to leverage large-scale policy change.

The other observation I’d make is that policy is often implemented without any observation or research basis.  Sometimes, that can be okay. Government should be able to act on promising but unproven ideas. Still, it is better when there is a research basis for expecting that some interventions can work. Gordon Berlin, President of MDRC was also part of this session; they are our partner on the AutoSave pilot.

Between sessions I wander the halls. Of course, this is where much of the action and interaction takes place. I used to come to this conference hoping to meet the tenured professors whose work I had read in graduate school. For example, I had a nice exchange with Tim Smeeding, who has written extensively on a variety of social policy topics, including how to measure and define asset poverty. We spent a good deal of time talking about all of the data limits researchers currently face and ideas on how to fill in the gaps. Anyone with ideas on this, send me an email offline, as it is an area I plan to dig into next year.

Also in the halls, I ran into Julian Le Grand, who was a senior advisor to Tony Blair and a major proponent of Labour Government’s ground breaking asset-based welfare policy initiatives. We lamented that the new coalition government is dismantling many of these efforts, but they are setting up a natural experiment where we might examine whether the participating cohort is better off than their generational neighbors.

Like any good conference, there are a set of choices to be made. APPAM often pays no mind to my session preferences when scheduling the sessions. I’m passing up a panel on the foreclosure crisis, featuring a paper by Michael Collins, and settling into a panel on behavioral economics and tax policy. Yes, two topics I love.

Sendil Mullainathan, of Harvard and the Ideas 42 policy shop, begins the behavioral economics panel by arguing that arguing that when thinking about behavioral economics it is worth remembering that economists are traditionally not that savvy about psychology. They like simplified models.  But there are many things going on in people’s minds. When looking at the impact of taxes, we try to hone in on how choice is altered through price. Another set of policies don’t control the price but they set defaults that change the error rate. Mechanism design becomes much more important. Adam Looney of Brookings focused his presentation on the observation that people don’t respond to taxes that they don’t understand.  Visibility and salience are important as drivers of behavior. Does a tax impact demand if it is presented differently? Taxes delivered at the register are potentially distinct if they are embedded in the price. The posted price of the good seems to matter a great deal.

Finally, Brigitte Madrian, whose work on automatic enrollment in savings plans I love, presented three primary implications of applying behavioral economics to tax policy. First, behavioral economics shows us that inertia is powerful. People respond slowly to changes in tax rates. Short-run elasticities are different from long-run elasticities, even though our methodologies tend to look at the short term. Second, salience matters. If you have easy pass on the highways, people often don’t think about tolls. If people don’t know what they are paying, they are not responsive to tax changes. If it is not salient, behavioral responses won’t be large. Sometimes, we want salient prices, so they can respond. Third, simplicity matters as well. When things are easy to so, people will be more likely to do it. Automatic enrollment is simple and drives numbers up. But even when forms are simplified, numbers also go up (although not as much). Simplified forms for student loans actually increase college attendance. She notes the problem of low-income people with simple taxes returns but who are fearful of what they don’t knowing going to paid tax preparers even they should be able to file on their own. Complexity drives down participation.

I got so inspired by the discussion that I decided to stay in the behavioral mindset and see my colleague David John of Heritage Foundation and the of the Retirement Security Project discuss his AutoIRA proposal on a session on retirement savings organized by Teresa Ghilarducci (whose work on retirement security is also excellent). That’s enough name dropping for now…