Reading list
Center for Studying Health System Change: (1) publications on insurance coverage and costs, (2) publications on health care markets, (3) all publications. That should be enough to keep you busy for a while.
Despite widely documented variations in health care outcomes by insurance status, few nationally representative studies have examined such disparities in the inpatient setting. [Our objective is to] determine whether there are insurance-related differences in hospital care for 3 common medical conditions. … For each diagnosis, we compared in-hospital mortality, length of stay (LOS), and cost per hospitalization for Medicaid and uninsured patients with the privately insured. Compared with the privately insured, in-hospital mortality among AMI and stroke patients was significantly higher for the uninsured (adjusted odds ratio [OR] 1.52, 95% confidence interval [CI] [1.24-1.85] for AMI and 1.49 [1.29-1.72] for stroke) and among pneumonia patients was significantly higher for Medicaid recipients (1.21 [1.01-1.45]). Excluding patients who died during hospitalization, LOS was consistently longer for Medicaid recipients for all 3 conditions (adjusted ratio 1.07, 95% CI [1.05-1.09] for AMI, 1.17 [1.14-1.20] for stroke, and 1.04 [1.03-1.06] for pneumonia), although costs were significantly higher for Medicaid recipients for only 2 of the 3 conditions (adjusted ratio 1.06, 95% CI [1.04-1.09] for stroke and 1.05 [1.04-1.07] for pneumonia). … Americans hospitalized for 3 common medical conditions, significantly lower in-hospital mortality was noted for privately insured patients compared with the uninsured or Medicaid recipients. Interventions to reduce insurance-related gaps in inpatient quality of care should be investigated.
Regression Discontinuity Designs in Economics, by David S. Lee and Thomas Lemieux
This paper provides an introduction and “user guide” to Regression Discontinuity (RD) designs for empirical researchers. It presents the basic theory behind the research design, details when RD is likely to be valid or invalid given economic incentives, explains why it is considered a “quasi-experimental” design, and summarizes different ways (with their advantages and disadvantages) of estimating RD designs and the limitations of interpreting these estimates. Concepts are discussed using examples drawn from the growing body of empirical research using RD.
This paper compares the structural approach to economic policy analysis with the program evaluation approach. It offers a third way to do policy analysis that combines the best features of both approaches. I illustrate the value of this alternative approach by making the implicit economics of LATE explicit, thereby extending the interpretability and range of policy questions that LATE can answer.
Two recent papers, Deaton (2009) and Heckman and Urzua (2009), argue against what they see as an excessive and inappropriate use of experimental and quasi-experimental methods in empirical work in economics in the last decade. They specifically question the increased use of instrumental variables and natural experiments in labor economics and of randomized experiments in development economics. In these comments, I will make the case that this move toward shoring up the internal validity of estimates, and toward clarifying the description of the population these estimates are relevant for, has been important and beneficial in increasing the credibility of empirical work in economics. I also address some other concerns raised by the Deaton and Heckman-Urzua papers.
Instruments, Randomization, and Learning about Development, by Angus Deaton
There is currently much debate about the effectiveness of foreign aid and about what kind of projects can engender economic development. There is skepticism about the ability of econometric analysis to resolve these issues or of development agencies to learn from their own experience. In response, there is increasing use in development economics of randomized controlled trials (RCTs) to accumulate credible knowledge of what works, without overreliance on questionable theory or statistical methods. When RCTs are not possible, the proponents of these methods advocate quasi-randomization through instrumental variable (IV) techniques or natural experiments. I argue that many of these applications are unlikely to recover quantities that are useful for policy or understanding: two key issues are the misunderstanding of exogeneity and the handling of heterogeneity. I illustrate from the literature on aid and growth. Actual randomization faces similar problems as does quasi-randomization, notwithstanding rhetoric to the contrary. I argue that experiments have no special ability to produce more credible knowledge than other methods, and that actual experiments are frequently subject to practical problems that undermine any claims to statistical or epistemic superiority. I illustrate using prominent experiments in development and elsewhere. As with IV methods, RCT-based evaluation of projects, without guidance from an understanding of underlying mechanisms, is unlikely to lead to scientific progress in the understanding of economic development. I welcome recent trends in development experimentation away from the evaluation of projects and toward the evaluation of theoretical mechanisms.
Silly Bandz
With two girls under six in my house, Silly Bandz are all the rage. According to the Wall Street Journal, they’re the latest little kid fad, and earning they’re maker big-time revenue:
BCP Imports LLC, maker of the Silly Bandz bracelets that have become an accessory de rigueur on elementary school playgrounds, is the latest small company looking for a way to extend its appeal with pint-sized customers.
Retailers selling the packs of 24 bracelets for $5 to $7 a pop cannot keep them in stock, and the company has had to hire more than 350 employees since October, according to founder Robert Croak. Silly Bandz have generated more than $100 million in annual sales, he says.
The kids can’t get enough. As parents, we use them as incentives to get ours to do what they otherwise find difficult (practicing piano, putting away toys, etc.). Everybody seems reasonably happy with our very local economy. When this fad dies, our currency will crash. Of course, they’ll be another.
Firms as Irrational Actors
The leadoff article in the Competition Policy International Journal’s (paywall) recent “focus issue” on behavioral economics is a primer on the literature regarding systematically irrational behavior by firms and its implications for competition law and policy. The topic is in itself unusual, as the authors Mark Armstrong and Steffan Huck note, inasmuch as a large part of the research on sub-optimal decision making focuses on consumers. Firms, by contrast, are frequently presumed–especially by courts and policymakers–to be model rational actors. The results surveyed by the authors strongly suggest that this is very frequently not the case.
One example with significant practical consequences is the manner in which cartel formation deviates from theoretical models under the influence of different enforcement regimes. Game theory predicts the ineffectiveness of leniency programs, such as those employed by competition authorities in the US and Europe, under which lower penalties or complete amnesty are given to the first participant to turn in a price-fixing conspiracy. Theories postulate that leniency can even encourage cartel formation because rational participants can use it as an enforcement mechanism to punish cheaters. But experimental results have found lower prices and incidence of cartel formation under leniency. The reasons for these deviations from rational maximization are speculative, but the fact that they occur is highly relevant to the formulation of competition policy.
Similar results are also relevant to the emerging practice by US courts of evaluating and often dismissing antitrust complaints based on their facial “plausibility” without allowing any investigation of the actual facts underlying a claim:
Courts and regulators in some jurisdictions may not consider seriously conduct (such as predatory pricing, for instance) which does not appear to make “business sense” according to their judgment. Leslie reports that “if a plaintiff’s complaint describes a conspiracy that the judge concludes is irrational, then the court rules that the conspiracy must not have happened as a matter of law, regardless of the evidence presented by the plaintiff to support its claim.” In the light of the theories and evidence reported in this article, we suggest that a dogmatic attitude towards the pervasiveness of business rationality may lead to instances where harmful behavior goes unpunished.
I will be watching for how lawyers use the growing body of behavioral economic evidence in this area to oppose dismissal in antitrust cases, and whether courts will pay it any heed.
The filibuster, part 6: Deliberate, not too long, mostly bills
This post is part of a multi-post series on the filibuster in the U.S. Senate. An index to all other posts in the series, as well as a list of main sources that have informed this series, is included in the first post.
When I began this summer project on the filibuster, I didn’t know enough about it to form my own opinion on whether and how to reform it. Now I do. I’m sympathetic to the notion that it is far too easy for the Senate’s minority to obstruct and delay. Scott Lilly makes a good case that the Senate
fails to complete much of the work for which it is responsible and falls so far behind schedule in completing the work it does do as to seriously undermine the capacity of the entire federal government to respond in an effective and efficient way to the problems facing our country.
Yet I’m also sympathetic to the notion that debate more deliberative and open than occurs in the House is of some value. Hence, the filibuster should be reformed, but not too much.
These are two principles at odds. Providing the minority rights to consider and debate, but not to unduly delay and obstruct is a balancing act. But the two principles can be better balanced. The key is to permit reasonable delay so senators can review bills and nominees, but not to allow them to be open ended. The goal should be to strengthen cloture, not weaken filibuster or make it more onerous.
As explained in the Congressional Research Service report on this topic, cloture has far more value than just ending debate. It is also a means of streamlining the process of amendments. As such, it’s used as an organizing force, to get things done, even when obstructionist motivations are not present. It helps senators with contradicting incentives accomplish the nation’s business. In fact, that’s the purpose of the two-day ripening period between submission of the cloture petition and the cloture debate. Senators need to prepare and submit their amendments. Once the two-day window closes, so does the opportunity to submit an amendment.
So, cloture has value that can be enhanced and strengthened while providing time for senators (minority and majority alike) to participate. To do so, and in looking over the various proposals I reviewed last week, I suggest the following as one set of possible reforms:
- If there need be a reduction in the cloture threshold, do it across the board, once and for all, not in steps that lengthen the legislative process, as proposed by Tom Harkin. For example, reduce the cloture threshold from a three-fifths (60 vote) to a four-sevenths (58 vote) majority.
- Adopt Michael Bennet’s idea and impose expiration dates on nomination holds, executive and legislative alike. A hold in this case should be for purposes of review, not for indefinite obstruction. Extend the expiration date in instances where the hold is bipartisan. Perhaps the length of delay could be proportional to degree of bipartisanship in some fashion.
- Institute Bennet’s proposal to reverse the bias of the cloture vote. Force the minority to produce 41-votes to continue debate, not the majority to produce 60 to end it.
- Create a one-filibuster-per-bill rule by eliminating the ability to filibuster a motion to proceed. (Separate filibusters on amendments would be possible.) Forcing the majority to wait out two rounds of cloture for each bill is unnecessarily dilatory (once on the motion to proceed, once on the bill itself).
- Further limit debate. Thirty hours per cloture vote is unnecessary and rarely fully used (never before 2003, though a few times since). Cut the time to some length that seems binding. Lilly suggests 16 hours.
Those are just my favorite ideas to date. I’m sure we’ll hear about others. Notice that they mostly focus on making cloture easier, not making filibuster harder. Moreover, they preserve important but reasonable pauses in the legislative process. Readers who can’t get enough of this stuff and want to know what someone more expert than I thinks about it should read Jonathan Berstein’s favorite reform ideas.
We should all want a more efficient Senate, provided all sides of a debate are heard as much as necessary, but no more, and largely on bills of import (not motions and minor executive nominations). So, my charge to the Senate: deliberate, not too long, mostly bills (h/t Michael Pollan).
Health insurance and Hispanics in Massachusetts
Racial and ethnic disparities in health care and insurance is no where near my expertise. But I have questions, such as, “Why are so many Hispanics uninsured in Massachusetts (or Boston, really)?”
Witness results from the Boston Area Community Health (BACH) survey, as presented by John McKinlay and Carol Link at this year’s AcademyHealth Annual Research Meeting. The BACH is a survey of a representative sample of Boston residents ages 30-79. (In the figure below, region between each adjacent pair of vertical bars is 20 percentage points.)
The bars illustrate how health insurance coverage changed with health reform in Massachusetts. Though Hispanics experienced the greatest gains in coverage (15.3 percentage points), they are still the group least likely to be covered (10.9% uninsured). Why?
The individual mandate is working in Massachusetts
The following is a re-post of my most recent Kaiser Health News column.
In Massachusetts, the individual mandate requiring state residents to buy health insurance is working. Yet, a similar requirement remains among the more controversial elements of the new national health reform law. Opponents of the mandate resent being required to purchase a product they may not want. Proponents claim that the mandate is necessary to prevent an unraveling of the broader set of reforms in the law. But will it work? It is in Massachusetts, and that should give reform advocates some confidence.
First of all, what does it mean for the mandate to “work?” The purpose of the mandate is to counter a potential threat to health insurers’ stability when they are required to accept all comers, even those with preexisting conditions. If individuals have access to insurance coverage whenever they please but are not required to have it all year, some will choose to enroll only when sick and then drop coverage when healthy. If too many people were to do just that, then insurers would be on the hook for more health care expenses than they could cover with collected premiums.
No insurer could survive a sufficiently severe level of such “adverse selection” (policyholders’ health care costing much more than their collective premiums can cover), and the individual mandate is designed to ensure that they won’t face one. Requiring individuals to purchase coverage–and pay premiums–even when they’re healthy guarantees that insurers have sufficient funds to cover the claims of the sick.
The individual mandate is working in Massachusetts because it is preventing a destabilizing level of adverse selection. Although there are individuals gaming the system in the state—by waiting to purchase insurance until they need it–the overall coverage rate is high (about 96% insured) and the associated degree of adverse selection is very low.
In a recent report released by the Massachusetts Division of Insurance, actuaries estimated that part-year insurance purchasing in Massachusetts’ combined individual and small group market increased premiums by 0.5 percent to 1.5 percent. Based on an average individual premium in Massachusetts of about $5,000 per year, that translates into an annual premium increase of $25 to $75, far too low to have a major impact on the market. Insurance companies can pass that level of premium increase on to consumers without many of them dropping coverage.
This is good news for Massachusetts and for the country. The penalty for lack of compliance with the state’s mandate is slightly lower on average than what the fully phased-in penalties will be (in 2016) under the new national health reform law. (I’ve estimated them to be about $674 per person per year under the national law and $537 under Massachusetts law.) Thus, all else being equal, individual mandate compliance ought to be at least as high under the new law as it is in Massachusetts. The results seen in the state imply that little gaming should be expected nationally.
Still, one should not be too complacent. Not every state is like Massachusetts (as some might rejoice). It is possible that individuals in conservative states where the mandate is not popular would be more likely to make short-term insurance purchases. However, the new health reform law has one thing that the Massachusetts reforms lack, the ability for insurance exchanges to impose open enrollment periods, something Massachusetts Gov. Deval Patrick (D) and Massachusetts Senate President Therese Murray (D) both advocate for their state’s version of the exchange.
Year-round access to insurance facilitates gaming the system, so limiting access to certain months should reduce it. It may also reduce levels of coverage overall, so there is a catch. Another perfectly reasonable reform that avoids this trade-off would be to increase the penalties for non-compliance.
No doubt tweaks to the Massachusetts and the national law such as these will be made. But it is reassuring that they’re just that: tweaks. The fundamental structure of both laws and the role of the individual mandate they include appear to be sound. It’s working in Massachusetts. That’s good news for all of us.
Subscribe
