Medical Malpractice Data and Inquiries

The current issue of Journal of Empirical Legal Studies includes an interesting data resource and survey by Bernard Black, et al., titled Medical Liability Insurance Premia: 1990–2016 Dataset, with Literature Review and Summary Information. Having just talked briefly about med mal premia and healthcare regulation last week, I was interested to read through the review and description of some of the data and trends. The authors have compiled data from the Medical Liability Monitor, “the only national, longitudinal source of data on med mal insurance rates.”  But they don’t stop there.

We link the MLM data with several related datasets: county rural-urban codes (from 2013); annual county- and state-level data on population (from the Census Bureau); number of total and active, nonfederal physicians, with a breakdown by specialty (from the Area Health Resource File, originally from the American Medical Association); annual state-level data on paid med mal claims against physicians from the National Practitioner Data Bank (NPDB), available through 2015; and data on direct premiums written by med mal insurers from the National Association of Insurance Commissioners (NAIC), available through 2015. We also provide a literature review of papers using the MLM data and summary information on the association between med mal insurance premia and other relevant features of the med mal landscape.

The data appendix, public data, and STATA code book (for cleaning the dataset) are also available from SSRN here. The survey includes a summary of some research into possible explanations for and consequences of medical malpractice premia: effect of med mal risk on healthcare spending, effect of med mal reform on med mal premia, effect of med mal rates on C-section rates and physician supply, effect of med mal payouts on med mal premia.

Noticeably absent from the literature they summarize, which they claim are the principle prior studies using MLM data, is any attention to or focus on market structure issues. Doubly so since there has been a consistent drop in rates over the past 15 years that is generally unexplained in the cited literature. Now, I don’t specialize in health care industry research, but I do know that in the past 15 years there has been an ongoing trend of consolidation among both health insurance companies and medical providing companies (e.g., hospital networks, physician groups, both).  I could easily hypothesize a couple potential dynamics:

  • Increased consolidation among insurance companies may lead to contractual incentives (by way of contract rates and performance measures) that affect the expected cost of med mal insurance.
  • Increased consolidation among hospital networks and physician groups leads to more consistent or standardized practices across larger populations of patients/services, thereby reducing uncertainty or volatility of medical service provision/quality and, thereby, expected cost of med mal insurance.

I suspect there are several potential channels, but it would seem a potentially fruitful area of research–and now there is a more convenient data set with which to play.

How mergers affect innovation…maybe?

Justus Haucap and Joel Stiebale with the Düsseldorf Institute for Competition Economics (DICE) at the University of Düsseldorf have a recent paper analyzing the effects of mergers on innovation in the European pharmaceutical industry. The develop a model that suggests mergers reduce innovation not only in the merged firms, but among industry competitors as well. Their data bear this out, as explained in the abstract:

This papers analyses how horizontal mergers affect innovation activities of the merged entity and its non-merging competitors. We develop an oligopoly model with heterogeneous firms to derive empirically testable implications. Our model predicts that a merger is more likely to be profitable in an innovation intensive industry. For a high degree of firm heterogeneity, a merger reduces innovation of both the merged entity and non-merging competitors in an industry with high R&D intensity. Using data on horizontal mergers among pharmaceutical firms in Europe, we find that our empirical results are consistent with many predictions of the theoretical model. Our main result is that after a merger, patenting and R&D of the merged entity and its non-merging rivals declines substantially. The effects are concentrated in markets with high innovation intensity and a high degree of rm heterogeneity. The results are robust towards alternative specifications, using an instrumental variable strategy, and applying a propensity score matching estimator.

While I haven’t yet read the paper in detail, a cursory examination suggests they have ignored another possibility: mergers in high-intensity R&D industries could be a leading indicator of decreased innovation productivity (i.e., lower returns to investment in R&D). Consider that as research advances, the “low hanging fruit” are collected first before the more difficult (and lower return) investments are pursued. As companies in a high-intensity R&D industry exploit all of the low hanging fruit, particularly internally, one might expect mergers as a way of expanding the available set of lower-cost/higher-return R&D investment opportunities. Since firms are competing in the same science space, a slow-down in one firm is likely to be spuriously correlated with slowdowns throughout the industry.

“Affect” is a word of causation. To suggest that mergers cause a reduction in innovation is a strong statement–especially when paired with a merger policy implication. This may be something that bears more scrutiny since, as the authors note, the entire subject is one on which relatively little light has thus far been shed.

New POLCON Data Set Released

Witold Henisz (Wharton) has just released the latest update to his POLCON (political constraint) index, which includes data up through 2012. The previous (2010) release included data only up through 2007. The POLCON data uses a spatial modeling technique to synthesize a number of variable characterizing the structures and ideological alignments of countries’ political system, including the number and types of veto points and the party control (and fractionalization) of different government bodies.

The data are made available for no fee except the promise of an appropriate citation. The 2013 release is available for download in both STAT and Microsoft Excel formats and a code book is provided.

This index is an excellent resource for scholars interested in cross-country comparisons that take into account political uncertainty, and Henisz has been doing the academic community a great public service in maintaining and updating this resources since its inception.

Conducting Empirical Legal Scholarship

The 12th Annual Conducting Empirical Legal Scholarship Workshop will take place at the USC Gould School of Law May 22-24, 2013. The workshop is for law school faculty, political science faculty, and graduate students interested in learning about empirical research and how to evaluate empirical work. Leading empirical scholars Lee Epstein and Andrew Martin will teach the workshop, which provides the formal training necessary to design, conduct, and assess empirical studies, and to use statistical software (Stata) to analyze and manage data. Participants need no background or knowledge of statistics to enroll in the workshop. For more information, see the conference website.

Most economics grad students get a lot of exposure to econometric and statistical concepts, but not necessarily to an understanding of how to actually apply that knowledge to conduct empirical research. I suspect the organizers may be willing to let even economics grad students participate.