Higher Ed Cop Out #2

December 28, 2008 | Blog


As promised– an ongoing series!

Cop-Out #2: Conditional financial aid

There’s been plenty written about the unfortunate shift to merit aid, a shift that disproportionately benefits the middle-class. But I’ve got my eye on another, related yet distinct trend– “conditional” aid. Last week’s New York Times magazine included an article on conditional cash transfer programs that provides nice background here. But in a nutshell, this is aid allocated based on performance– maintaining full-time enrollment, a certain GPA, etc. The best example, to date, is the program MDRC is running with a lot of support from the Bill & Melinda Gates Foundation– after demonstrating (very, very small) effects at two New Orleans community colleges, MDRC has ramped up their demonstration in multiple states.

Here’s the theory of action: Low-income kids need aid, but also need some accountability requirements in order to do well in college. This is akin to the new paternalism guiding the 1990s welfare reform– it equates purely need-based aid with a “nanny state” and aid conditional on performance with a “daddy state.” If successful, simply because it works it threatens to take over, replacing need-based aid entirely.

So what’s the problem– isn’t a successful program a successful program?

I’d argue NO. The move to conditional aid is premature, because we do not yet know whether strictly need-based aid– given on need, not performance– is effective at increasing attainment. Moreover, we do not know if current levels of performance are reflective of the underfunding of need-based aid– the Pell has never been funded at the intended levels, and analysts have been far too quick to jump to the conclusion that it is therefore ineffective.

I’m not surprised at MDRC’s involvement in this one– after all, they’re also testing NYC’s conditional cash transfer program, and their evaluations were strong impetus for the welfare reforms. But I have the privilege, as an academic, of considering both theory and evidence– and I’d say the theory thus far — that simply providing the aid necessary to pay for college– has not been disproven. Why, then, take a paternalistic approach to requiring things of our low-income college kids that we do not require of the middle-class kids?

Full disclosure: I’m currently co-directing a random assignment evaluation of a need-based aid program. But trust me, this is not simply because the opportunity presented itself, but because this is a strategy I think needs to be tested BEFORE we move to conditional aid “reforms.” From an empirical standpoint, not doing so doesn’t constitute putting the cart before the horse, but I think from a political and ethical standpoint it unnecessarily opens up a can of worms that need not be opened just yet.

At this point, conditional aid appears to be a cop-out– a way, perhaps, to gain more middle class support for aid. But might we be just as persuaded by hard-core, gold-standard evidence of large effects of need-based aid? I say, let’s start there.

Addendum:

My original post wasn’t that well done, I admit. So let me clarify: I call conditional financial aid a cop out not because the current strategy (need-based aid) hasn’t yet been shown to be cost effective but because it follows on, and is part of, a potentially dangerous trend whose origins seem to be well-intentioned. This is a trend towards removing social supports and replacing them with the similar (or lesser) supports combined with accountability. Yes, it’s true (as Lashawn points out in the comments) that MDRC’s program awards new aid attached to performance requirements on top of existing need-based aid (to the extent that existing aid isn’t reduced because of aid rules). My fear is that as we’ve seen in the past, programs like these are too easily misinterpreted–even before results appear. Opening this can of worms makes it so much more likely that the results of a positive evaluation will lead schools, states, and feds to simply attach conditions to the existing aid students get, rather than ADD new aid with conditions. That means aid becomes performance-based funding, not performance based incentive funding, and I’m opposed to that as a mechanism for enhancing the performance of disadvantaged students and schools.

Some other forms of conditional cash transfer provide money that one would otherwise not receive–without taking away existing resources. For example, Oportunidades pays poor moms a stipend if they take their kids to the doctor. I’ve got no problem with that. Similarly if welfare reform had involved giving the poor more financial support with conditions, instead of adding additional requirements to their existing aid, I would’ve been for it.

So in sum– my concern is with the signals sent by these conditional aid programs, and evaluations of them– which I think, in 5-10 years we’ll see resulting in opening a gateway for policymakers to believe they have hard evidence to treat low-income college students as they do welfare recipients, requiring performance in exchange for aid. I do not find fault with MDRC’s well-designed evaluation or their goal of building on their past findings (and their own theories of action), nor do I fault the Gates Foundation for funding what appears to many to be a promising program. I highlighted this program to draw attention now to the broader implications of these conditional aid programs, to note how similar programs have been used and misused in the past, and therefore issue a warning that researchers and funders alike need to pay close attention to the assumptions supported by their work.

I realize my original posting wasn’t quite this nuanced or that well-articulated, and for that I apologize– I hope that clarifies my point.

2 Comments

  1. Reply

    Lashawn Richburg-Hayes, Senior Research Associate, MDRC

    December 30, 2008

    Recently Dr. Sara Goldrick-Rab noted that conditional cash transfers were an educational cop-out, threatening to take-over and replace need-based aid entirely. Unfortunately, Dr. Goldrick-Rab poorly characterized the conditional grant strategy and failed to note the origins of the strategy. Most importantly, she did not discuss how the strategy seeks to address the problem of poverty from a different perspective.

    But, first, a point of agreement with Dr. Goldrick-Rab. The researchers at MDRC—a nonprofit social policy research organization committed to understanding and disseminating what works for low-income people—believe that much more research needs to be done to understand how to improve college success. The Wisconsin study (http://www.finaidstudy.org/) is one such effort that we support. This study seeks to answer whether persistence and success can be attained by providing need-based funding to the maximum amount specified in the Higher Education Act (HEA). This is a laudable research goal and one whose answers will help push policy towards more need-based aid rather than less. As mentioned by Dr. Goldrick-Rab, this effort is based on the notion that students fail to continue and succeed because of inadequate funding. That may or may not be true, and this research will help fill in the gaps. This research comes at the issue of postsecondary success from another angle, which is good for research. It’s not clear whether such an angle is “right” or “wrong” or “misguided.” Such characterizations are counterproductive and do nothing to serve the students that we care so much about.

    The performance-based scholarship was an idea that descended from research on work-supports, the same research that suggested that supplementing the incomes of welfare-recipients would go a long way to helping them leave the welfare rolls more quickly. Work by MDRC (http://www.mdrc.org/subarea_index_16.html) found that while such transfers increased employment, it did not increase income—leaving families in poverty. In early 2003, officials from Louisiana were interested in knowing whether a variation of this idea—supplementing college-going—could both increase leaving from welfare and increase family income as a result of the human capital investment. The strategy of offering a need-based, contingent grant to low-income parents was piloted in two Louisiana community colleges. Results were positive, showing that upwards of 15 percentage points more (or a 30 percent increase) scholarship students returned the next semester. In addition, more students registered full-time, earned more credits, and attained grades above “C”—the very reasonable “performance” benchmark set for scholarship receipt. While students gained slightly more than 4 credits over the follow-up period, those credits contributed to the students accumulating close to a full semester’s course load. This is indeed small; however, the students in the study tended to work full-time (they did not have to be on welfare and few were), over half had two or more children, and they were single mothers. These are really high barriers to overcome, so a 30 percent increase in persistence is large and notable when taken in context.

    It’s important to note that this program operated in addition to traditional financial aid. That is, the scholarships were paid on top of Pell Grants and any other financial aid that students were qualified for. This means that the results cannot be extrapolated to suggest that such a strategy would work in lieu of traditional need-based aid such as Pell Grants.

    In short, all of the research being conducted on persistence and success in higher education is sorely needed and it will take more than one study to open the black box of barriers preventing such success. Researchers working in this area should work collaboratively, as there is a lot to be done and policy makers need to understand whether and why a particular strategy may work or not work and for whom.

  2. Reply

    Dr. Sara Goldrick-Rab

    December 30, 2008

    Hi Lashawn,

    I am delighted to know that you are reading our blog, and welcome your considered, thoughtful response.

    Thank you in particular for confirming the relationship between the performance-based scholarships and MDRC's research on work-supports. I am concerned about their relationship to not only the work supports experiments but also to the welfare experiments, which introduced new conditions to the receipt of welfare. The welfare experiments helped to redefine what welfare means in this country. As is the case with performance-based scholarships, the very existence of MDRC's welfare experiments was in many ways more important than their results. As my colleagues and I argued in our 2006 book, Putting Poor People to Work (Russell Sage), "by providing concrete instances of new welfare ideas being implemented, welfare experiments in the years leading up to welfare reform redefined what was possible politically and made more extreme ideas viable in the eyes of the public." Similarly, the very fact that the Gates Foundation has taken as its centerpiece financial aid project in the new agenda one that focuses on adding new conditions to aid is powerful and troubling.

    I also want to thank you for providing more details on the outcomes of the initial results from Opening Doors. As your colleague Tom Brock well knows, I'm very eager to read the full report on that experiment's methodology and findings (not yet provided on MDRC's elegant and informative website). Having attended many presentations about the results, I'm very familiar with the framing of a 4 credit increase in attainment as "notable when taken in context." But I'd argue instead that the statement reflects low expectations for poor women, and only serves to excuse those who would rather not more fully invest in their education. Please don't get me wrong- I am not saying this is your intention, or that of your organization. Rather, it's an unintended, subtle consequence that as a sociologist I think deserves close attention.

    Moreover, I sense there is much disagreement among researchers associated with the project as to the size of your effects. For example, on Princeton's and UCLA's websites I found statements made by health economists who had intended to leverage your experiment to learn about the causal effects of a college education on health. Statements indicated that such analysis could not occur because Opening Doors had "no effect" on attainment.

    I am in full agreement that we need to know much more about how to increase attainment of low-income adults, and I am more than delighted to see MDRC's growth in this area, and the Gates foundation support of it. But we must also think about the political and social contexts and ramifications of our evaluations, the messages they send, and what they truly achieve. Supporting-- and challenging each other-- will help us find a spirited research-based approach to a solution.


Would you like to share your thoughts?

Would you like to share your thoughts?

Leave a Reply

© 2013 The EduOptimists. All Rights Reserved.