Issue Brief: Volume 1, Number 11

Optimizing Outcomes on the Health Insurance Exchanges

Author: Tom Baker, Jonathan Kolstad, Amanda Starc, Kevin G.M. Volpp

On September 9, 2013, the Wharton Public Policy Initiative (PPI) organized a bipartisan discussion on the future of health care reform, featuring three of the University of Pennsylvania’s foremost academic experts on health care policy.


  • The success of the new health insurance exchanges will depend greatly on the quality of the enrollment decisions that consumers make. Choosing the wrong insurance product can translate into billions of dollars in wasteful spending at the national level.
  • Faculty at the University of Pennsylvania have contributed to several studies outlining important ways that the exchanges can be made to work better for consumers—and for the larger economy.
  • Developing simplified insurance products that make it easier for people to understand the actual cost of their health care; offering standardized plans that allow consumers to better appreciate the comparative value of different levels of insurance coverage; and building features such as cost calculators and smart defaults into the architecture of the exchange websites, are critical improvements that would enable Americans to make decisions that enhance both their own and the nation’s economic welfare.

Mark Duggan, Faculty Director of Wharton PPI, led the proceedings, with Zeke Emanuel (the Diane v.S. Levy and Robert M. Levy University Professor and formerly the special adviser for health policy to the director of the White House Office of Management and Budget during the first Obama Administration) and Dan Polsky (Executive Director of the Wharton School’s Leonard Davis Institute of Health Economics (LDI), current member of the Congressional Budget Office’s Panel of Health Advisers, and formerly the Senior Economist on health issues for the President’s Council of Economic Advisers under George W. Bush) providing commentary on the issues. A video of their discussion can be found online at What appears below is an extension of the dialog from the September 9 event, based on a broader survey of the research conducted by faculty experts at the University of Pennsylvania, and focusing specifically on the new health insurance exchanges created by the 2010 Affordable Care Act (ACA).

The potential significance of the exchanges is enormous. According to CBO projections, they will expand health care coverage to 20 million people, and in the process—if they function as proponents of the ACA expect—establish efficient, competitive marketplaces that keep down the cost of buying health insurance. Since the exchanges were launched on October 1, public attention has focused primarily on the technical failings of the federal exchange site,, and of several of the 15 state-administered exchanges. The technical issues obviously are critical; otherwise many discouraged users may opt to go without health insurance coverage. Securing ample enrollment is essential for spreading health risks, increasing the number of plan options, and making insurance on the exchanges affordable.

Once the technical problems are fully addressed, however, the success of the exchanges will hinge not only on driving a sufficient number of consumers to them; it will depend also on the quality of the choices they make when enrolling for health insurance. Choosing the wrong insurance product can translate into significant, unnecessary expenses or inefficient health care for individuals and households, which, in the aggregate, can amount to billions of dollars in wasteful spending at the national level.

The challenges of helping consumers use the exchanges to make good economic decisions regarding health insurance was emphasized by Dan Polsky at PPI’s September 9 event. As he pointed out, there has been a health insurance exchange around for many years already called It provides an open but generally unstructured marketplace: “you just click on something and there [are] thousands of plans and you have no idea what’s in them and it’s impossible to make any choice between the price versus the things that … the insurance plans [are] covering,” he said. So “you wind up making fairly random decisions.” Government-run exchanges that operate along these same lines of just throwing options at consumers without “guid[ing] people toward the right decisions” are not going to be effective in generating good, economically beneficial matches between people and plans.

This issue of helping consumers better comprehend the features of the insurance plans offered on the exchanges becomes all the more pressing in light of research indicating that the level of “health literacy” among Americans is quite low. Many people—despite feeling confident in their own knowledge about buying insurance—actually have a very limited understanding of insurance terminology and how health insurance works. A recent paper co-authored by several scholars from the Penn Center for Health Incentives and Behavioral Economics at LDI, including Jonathan Kolstad from Wharton’s Health Care Management Department and Kevin Volpp from Penn’s Perelman School of Medicine and Wharton’s Health Care Management Department, finds that only 14% of a surveyed sample group was able to correctly answer four multiple choice questions about the four basic components of health insurance plans: deductibles, copays, coinsurance, and maximum out of pocket costs. Consequently, even fewer—just 11%—could navigate the features of a traditional insurance plan to calculate accurately the cost of a 4-day stay in the hospital.[1] These are the exact concepts and calculations that federal and state governments now are expecting Americans to grapple with successfully while purchasing insurance for themselves on the exchanges.

In response to these comprehension problems, the team of which Kolstad and Volpp were part (led by George Loewenstein from Carnegie Mellon University) worked with one of the major insurers to devise a simplified, all-copay insurance plan, without deductibles and coinsurance—two of the most commonly misunderstood, yet ubiquitous, aspects of health insurance. The research found that survey participants were much more adept at calculating the actual costs of a hospital stay with the simplified insurance plan. Moreover, there was a strong preference among the survey participants for the simplified plan, especially after they went through the exercise of using it to compute health care costs.

As the authors note, simplified health insurance—by empowering consumers to more correctly understand the actual cost of their health care—can help them become more adept in selecting the insurance plan that provides the best combination of coverage and costs for their expected needs. And this, in turn, will help to keep the cost of health insurance down, which is a main reason for establishing health insurance exchanges in the first place.

For this very reason, the health care exchanges will need to do a better job of closing the knowledge gap that exists among many consumers with regard to fundamental concepts that define health insurance. In this, the requirements of the Affordable Care Act itself will not be sufficient. The authors of this study point out that “the ACA adopts a somewhat superficial approach to dealing with” this knowledge gap “that revolves around the standardization and simplified presentation of information about insurance plan features.” The goal, however, should not be “to explain inherently complex insurance plans in simple terms,” which can wind up leaving people insufficiently informed. (And in fact, the new exchanges generally do very little with respect to educating consumers about things like what coinsurance is, or how deductibles work.) Rather, “a more fundamental approach would be to (1) design health insurance products that are truly simple,”—such as the one with no deductibles or coinsurance—“and (2) require plans to offer identical features that can be directly compared.”

This last point regarding the potential utility of offering standardized insurance plans on the exchanges, with identical features that can be compared side-by-side, is explored more fully in research co-authored by Amanda Starc, Assistant Professor of Health Care Management at the Wharton School. Starc has worked with Keith Ericson from Boston University in studying a “natural experiment” on product standardization afforded by the state exchange in Massachusetts (the Massachusetts Connector). In the case of Massachusetts, they found product standardization had a significant impact in helping to drive more consumers toward higher-value insurance plans that offered better coverage.[2]

Prior to 2010, the Connector presented purchasers with 25 different plans from 6 insurance carriers; the plans were categorized by actuarial value into 3 different tiers (gold, silver, and bronze). Ericson and Starc observed that between 2007 and 2009, the majority of new enrollees (63%) bought plans on the lowest (bronze) tier of coverage, offering the least expensive but least generous level of coverage, with 20% enrolling in the cheapest plan available. As with the study co-authored by Jonathan Kolstad and Kevin Volpp cited above, the research by Ericson and Starc demonstrates that consumers do not make health coverage choices based on an accurate understanding of insurance plan features and design. Rather, the propensity of consumers in Massachusetts to select the least expensive insurance option suggests to them that many people rely on simple “rules of thumb” in making these complex health care decisions—in this case, the basic maxim of “buy the cheapest plan.”[3]

As Ericson and Starc discovered, however, enrollments in Massachusetts shifted in 2010, after new regulations went into effect that standardized the insurance options. In response to feedback from consumers, who found the range of options confusing, Massachusetts in 2010 standardized plans into seven tiers (gold, silver-high, -medium and –low, and bronze-high, -medium and -low), offered initially by the same 6 carriers. Plans on the same tier now were required to have the same deductibles, copays, and coinsurance parameters, although they still differed in terms of the breadth of their physician and hospital networks, as well as the brand name of the carrier.

The work done by Ericson and Starc reveals that standardization had a marked positive effect on insurance market outcomes in Massachusetts: “it shifted consumers into more generous products.” Instead of reflexively going for the cheapest plan, the fraction of enrollees in bronze plans dropped from 63% to 44%, as more consumers opted for plans on the silver tier that offered greater financial coverage with lower deductibles but narrower provider networks.

Standardization enhanced the welfare of consumers. In addition to giving rise to more plan options, by making financial comparisons between those plans more straightforward, standardization made it easier for consumers in Massachusetts to see—and value—the relative financial generosity of the different choices. They were better able to appreciate that they would get more by selecting a plan on a higher tier. And because the premiums are slightly higher on those silver tier plans, insurance companies reaped some of the “welfare surplus” too.

Since the exchanges now being rolled out by the Affordable Care Act are very similar in design to the Massachusetts Connector, the lessons on standardization offered by the Massachusetts example are telling. The proliferation of health insurance exchanges should allow for further research and cross-state comparisons on the potential benefits of standardization. And as health reform moves forward, regulators at the federal and state levels need to be mindful of the ways in which insurance plan standardization can help lead consumers to select plans with the best combination of price and quality given their preferences and income.

In addition to pointing out the benefits of standardization, the experience of Massachusetts also highlights the important role that site design—the way in which information about available insurance plans is presented—plays in shaping consumer decisions. As Ericson and Starc observe, the shift to standardized plans in Massachusetts in 2010 also entailed a shift in how the experience of choosing insurance was structured on the Connector website. Prior to the 2010 standardization, plans appeared simply as a list, by ascending premium order; the one with the lowest premium sat at the top. Once standardized tiers were created, however, this changed. Consumers in Massachusetts now were asked to select which tier of financial coverage they wanted, and then to choose an insurance carrier within that tier. The revised design helped lead consumers away from just picking the cheapest plan, and toward selecting a plan based on the generosity of coverage within their chosen tier.

This idea that the “choice architecture” of the health insurance exchanges will be critical to enabling consumers to make good, efficient choices has been reinforced by other research findings. A couple of months prior to the start of the new exchanges, a study co-authored by Tom Baker from the University of Pennsylvania Law School gained wide media coverage for shining light on the problems that consumers were likely to face in making well-informed choices from the range of insurance options that would confront them on the online exchanges. That study found that when left solely to their own devices, individuals perform poorly—at “near chance levels”—in choosing the most cost-effective insurance plan, paying too much attention to out-of-pocket costs and deductibles, and not enough to the overall costs incurred in paying monthly premiums. What is more, the study found that most people are not even aware that they are making sub-optimal choices.[4]

The economic consequences of consumers selecting the wrong plan for them are significant. Even taking into account the subsidies that many families will receive to purchase insurance on the exchanges, insurance premiums will amount to between 4 to 9.5 percent of the income of individuals or families whose income falls between 150 and 400 percent of the federal poverty line.5 As this represents a large share of household income, it is important to ensure that those funds are spent as efficiently as possible. But also on a macro-level, inefficiencies like the ones identified in the research cited above defeat part of the purpose of having health care exchanges in the first place. The exchanges are supposed to offer an efficient marketplace for insurance, where competition for consumers and participation by a large number of insurers puts downward pressure on the costs of buying a health plan.  But as noted by Baker and his co-authors: “If consumers cannot identify cost efficient plans, then the Exchanges will not produce competitive pressures on health plan costs, one of the main advantages of relying upon choice and markets” on the exchanges.

The good news, however, is that there are demonstrably successful interventions that dramatically enhance the ability of individuals to make better health plan selections. When participants in the study were presented with a calculator that stated the total annual cost of a given health plan, their performance improved—especially when they were given “just in time education” in the form of a tutorial that explained how the annual costs of an insurance plan should be calculated. Performance was improved further still when the calculator and tutorial were combined with “smart defaults” that preselected the most cost efficient health plan options based on a person’s anticipated health care needs. These are all relatively inexpensive interventions that can—and according to the study authors, should—be integrated into the design of the websites that drive the exchanges, in order to optimize their functionality and effectiveness.

At the moment, however, the health care exchanges rolled out on October 1 do not include these features. Both the federal exchange and many of the state exchanges do point users to the Subsidy Calculator on the Kaiser Family Foundation website. This calculator allows visitors to input basic information such as their state of residency, income, and number of adults and dependent children in their household, and the calculator computes the size of the means-based subsidy that can be expected from the government to support the purchase of insurance on the exchanges, as well as the anticipated premium that would be charged for a plan on the silver tier of coverage. The Subsidy Calculator also offers information about the comparative cost of other levels of coverage, besides the silver tier.

What has not been put into usage, however, is the kind of calculator that Baker and his co-authors recommend: a tool to help consumers understand the relative cost of different plans available to them on the specific health care exchange that they themselves are using, so they can make the most educated, cost-effective choice. Regulators of the exchanges still need to consider implementing an online total cost calculator as the Centers for Medicare and Medicaid Services eventually did for their Part D beneficiaries, to help them accurately compute the out-of-pocket costs associated with different plans, based on their expected patterns of health care utilization. The exchanges do not currently use smart defaults, either. 

The potential economic impact of such tools is significant. In a controlled experiment done as part of the study described here, a group that had access to both smart defaults and a cost calculator made an error of just $77, on average, in purchasing a health insurance product—$456 less than the average error of participants in a group that did not have these tools. At the individual level, this is an expensive differential. That $456 amount is equal to around 1% of the median household income of consumers. “But, in the aggregate, an error of $456 represents staggering sums,” the study authors note. “If 20 million individuals make choices using the exchanges, a figure suggested by Congressional Budget Office estimates, unaided choice represents a cost to consumers of $9.12 billion dollars each year.” Moreover: “Since almost all of these policies are subsidized through tax credits, good choice architecture would produce substantial savings to the federal budget and taxpayers.”

The new health insurance exchanges are going to continue evolving over the months and years ahead. On the technical side, as Zeke Emanuel envisioned at the September 9 event, long after the glitches that currently impede consumer access to the exchanges are eliminated, there surely will be many smart techies entering the market to develop more sophisticated software to make shopping for health insurance more interactive, informative, and customized, like buying goods on “And like Amazon that makes recommendations about what books you might want to have,” the exchanges hopefully will get to a place technologically where “they are going to begin to tell you” to look at particular plans, “and then you are going to be able to see how well they are rated on quality, cost,” and other key parameters. 

But as the research studies described above make clear, those charged with setting up and administering the exchanges also will have a large role to play in making them work better for consumers—and for the larger economy. By continuing to review and adjust the design of the insurance products offered on the exchanges and the way those products are explained and presented to enrollees, policymakers can make it easier for consumers to negotiate the complexities of the insurance marketplace and make purchasing decisions that enhance both individual and national economic welfare.

About the Authors

Tom Baker, JD

Professor of Law and Health Sciences,  Penn Law School; Professor of Business Economics and
Public Policy, The Wharton School 

Jonathan Kolstad, PhD

Assistant Professor of Health Care Management, The Wharton School 

Amanda Starc, PhD

Assistant Professor of Health Care Management, The Wharton School

Kevin G.M. Volpp, MD, PhD

Professor of Medicine, Perelman School of Medicine; Professor of Health Care Management, 
The Wharton School; Director, Center for Health Incentives and Behavioral Economics, Leonard Davis Institute; Director, Penn CMU Roybal P30 Center in Behavioral Economics and Health


1 George Loewenstein, Joelle Y. Friedman, Barbara McGill, Sarah Ahmad, Suzanne Linck, Stacey Sinkula, John Beshears, James Choi, Jonathan Kolstad, David Laibson, Brigitte Madrian, John List, Kevin G. Volpp, “Consumers’ Misunderstanding of Health Insurance” (2013),  Journal of Health Economics 32 (2013), 850-862. 

2 Keith M. Marzilli Ericson and Amanda Starc, “How Product Standardization Affects Choice: Evidence from the Massachusetts Health Insurance Exchange” (2013); Keith M. Marzilli Ericson and Amanda Starc, “Designing and Regulating Health Insurance Exchanges: Lessons from Massachusetts (2012). Both papers are available at

3 Keith Marzilli Ercison and Amanda Starc, “Heuristics and Heterogeneity in Health Insurance Exchanges: Evidence from the Massachusetts Connector” (2011),

4 Eric J. Johnson, Ran Hassin, Tom Baker, Allison Bajger, Galen Treurer, “Can Consumers Make Affordable Care Affordable? The Value of Choice Architecture” (2013),

5 The 2013 federal poverty line (FPL) is $11,950 for an individual, $15,510 for a family of two, and $23,550 for a family of four. ACA subsidies are available to individuals and families with incomes between 100 and 400 percent of FPL. The 4 to 9.5 percent cost figures refer to the second lowest-cost silver plan in the exchange. Those with incomes from 100 to 150 percent of FPL would face premiums from 2 to 4 percent of income.