The Effects of “Early Offers” in Medical Malpractice Cases: Evidence from Texas

• Date: 01 December 2009
• DOI: http://doi.org/10.1111/j.1740-1461.2009.01158.x
• Published date: 01 December 2009
• Author: Bernard Black,David A. Hyman,Charles Silver

The Effects of “Early Offers” in Medical

Malpractice Cases: Evidence from Texasjels_1158723..767

Bernard Black, David A. Hyman, and Charles Silver*

Medical malpractice litigation is costly and time consuming. Professor Jeffrey O’Connell,

with various co-authors, has long advocated “early offer” rules that would encourage defen-

dants to offer to settle for economic damages plus attorney fees, and punish plaintiffs who

refuse such offers. Using detailed closed claims data from Texas for 1988–2005, we simulate

the effects of these “early offers.” Under a base set of assumptions, early offers will sharply

reduce payouts in cases with small economic damages (under $100,000, all amounts in 1988

dollars); will moderately reduce payouts in currently tried cases with economic damages

from $100,000–$200,000 and would normally increase payouts (and therefore will not

be made) in tried (settled) cases with economic damages over $200,000 ($100,000). Overall,

we predict that early offers will be made in 72 percent of all cases, and will result in a 16

percent (20 percent) decline in payouts in tried (settled) cases. Almost all this effect comes

from the sharp decline in payouts in cases with small economic damages. Defense costs will

drop by roughly 60 percent (20 percent) in currently tried (settled) cases in which an early

offer is made, and by about 13 percent overall. An early offer program will have very different

effects on different types of plaintiffs, with especially large payout reductions for elderly and

deceased plaintiffs. An early offer program also overlaps substantially in its effects with a

statutory cap on noneconomic damages (which 26 states already have). Defendants in many

of these states have already realized large reductions in payment of noneconomic damages;

the additional reductions from an early offer program are modest and would often affect

plaintiffs whose recoveries were already limited by damage caps. Our mixed results contrast

sharply with dramatic claims by O’Connell and co-authors, who predict 70 percent reduc-

tions in both payouts and defense costs. Their estimates reflect the compound effects of a

series of unreasonable assumptions.

I. Introduction

Critics of medical malpractice (med mal) litigation assert that it is slow, expensive, and

levies an unjustified “tort tax” on defendants. Similar criticisms apply to personal injury

*Address correspondence to Bernard Black; email: bblack@northwestern.edu. Black is Hayden W. Head Regents

Chair for Faculty Excellence, University of Texas Law School, and Professor of Finance, University of Texas, Red

McCombs School of Business; Hyman is Richard and Marie Corman Professor of Law and Professor of Medicine,

University of Illinois; Silver is McDonald Endowed Chair in Civil Procedure, University of Texas Law School.

We thank Jill Horwitz and participants in the 2008 Conference on Empirical Legal Studies, 2008 Midwest Law and

Economics Association Meeting, and workshops at Northwestern and University of Michigan Law Schools for helpful

comments, and Hyun Kim for superb research assistance.

Journal of Empirical Legal Studies

Volume 6, Issue 4, 723–767, December 2009

723

litigation more generally. Because the overwhelming majority of cases settle, rules that

encourage faster settlement are an obvious strategy for improving the performance of the

tort system.

In this article, we analyze one important early settlement proposal, first made by

Professor Jeffrey O’Connell in 1982.1Under the current version of this “early offer” pro-

posal, defendants can offer to settle for full economic damages plus attorney fees, possibly

with a minimum damages offer in some cases. Plaintiffs who refuse the offer face a large

stick in the form of a higher burden of proof; the defendants are “making offers that can’t

be refused.”2Recent studies by O’Connell and co-authors estimate huge payout reductions

from such a program—on the order of a 70 percent drop in both payouts and defense

costs.3

We simulate the impact of an “O’Connell” early offer program on payouts. We use a

data set of all closed medical malpractice claims in Texas from 1988–2005 with payouts over

$25,000 (1988$).4We study both tried and settled cases, and find similar results for both sets

of cases. All amounts in this article are in 1988 dollars; multiply by 1.82 to convert to 2008

dollars.

Under a base set of assumptions, we estimate that an early offer program could

produce a 16 percent overall decline in payouts in currently tried cases, and a 20 percent

decline in settled cases. We explore sensitivity analyses how this estimate will vary depending

on program details and one’s assumptions. This decline in payouts will result primarily from

large payout reductions for plaintiffs with limited or no economic damages ($0–$100,000).

There will also be large variation in how an early offer program affects different plaintiff

groups, with large payout declines for elderly and deceased plaintiffs, limited effects on

employed adults in nondeath cases and on children, and almost no effect on baby cases.

There is also substantial overlap between the cases in which an early offer program will

reduce payouts and the cases in which a cap on noneconomic damages (adopted by 26 states)

already does so.

Defendants will make early offers only if they expect to gain by doing so. This means

that early offers will be made principally in cases with “small” economic damages ($0 to

$100,000). They may also be made in currently tried cases with economic damages from

$100,000–$200,000, depending on likelihood of liability and other case characteristics.

Early offers would generally produce higher payouts in tried (settled) cases if economic

1O’Connell (1982).

2Hence the title of Professor O’Connell’s initial article, which begins “Offers That Can’t Be Refused.” Id.

3Hersch et al. (2007); O’Connell and Born (2008); O’Connell and Robinette (2008).

4This article is one of a series using the Texas closed claims database to explore different aspects of medical

malpractice and personal injury litigation. Other pieces of this project include Black et al. (2005) (trends in overall

payouts); Hyman et al. (2007) (comparing jury verdicts with actual payouts); Zeiler etal. (2007) (physician policy

limits and out-of-pocket payments); Black et al. (2008) (analyzing defense costs in medical malpractice claims);

Hyman et al. (2009a) (estimating effect of various damages caps); and Hyman etal. (2009b) (analyzing effect of

insurer duty to settle).

724 Black et al.

damages are over $200,000 ($100,000), and thus will normally not be made in these cases.

Overall, we predict that early offers will be made in 72 percent of all cases, which represent

about 42 percent of current payout.

The 70 percent payout reductions estimated by O’Connell and co-authors reflect the

compound effects of a series of unreasonable assumptions. These include: (1) assuming no

minimum damages offer in many cases; (2) assuming that two-thirds of paid damages are

noneconomic (we estimate that noneconomic damages represent 40 percent (52 percent)

of payout in tried (settled) cases); (3) assuming that current payouts include full payment

of economic damages (they do not in many tried cases and are unlikely to in settled cases);

(4) treating a fee of 10 percent of economic damages as a market-clearing price for the

services of plaintiffs’ attorneys (it isn’t close); (5) ignoring plaintiffs’ out-of-pocket costs; (6)

ignoring the time value of money; and (7) assuming that liability is certain.

Defense costs will decline moderately. We estimate that defense costs will drop by

roughly 60 percent (20 percent) in currently tried (settled) cases that are resolved with an

early offer, or about 13 percent across all cases. This contrasts with O’Connell’s assumption

that defense costs will drop by 70 percent in both tried and settled cases.

Our approach and data set have important limitations. Our data come from a single

state, albeit a large one. Our simulations require a series of assumptions. For tried cases,

these include assumptions about how to allocate payouts to different components of

damages, equilibrium fee levels, and plaintiffs’ ex ante chances of winning at trial in cases

that they actually win ex post. For settled cases, we also need to estimate the economic and

noneconomic components of payouts. We believe, however, that our assumptions for both

tried and settled cases are far more realistic than O’Connell’s. Our simulation approach

ignores how an early offer program might affect which cases are brought, which are taken

to trial, and how they are tried. Note, too, that any payout reductions from an early offer

program are a wealth transfer, and not a direct social savings. We do not evaluate here the

efficiency implications of a change in payouts.

Section II provides background on the pretrial settlement process, and describes the

Texas medical malpractice data set and our simulation methodology. Section III discusses

our data set, and Section IV describes our simulation methodology for tried cases. Section

V analyzes how an early offer program will affect tried cases. Section VI extends the analysis

to settled cases. Section VII examines how an early offer program will affect different types

of plaintiffs. Section VIII examines how such a program will affect defense costs, total

defendant costs, plaintiff’s net recovery, and system efficiency. Section IX compares our

early offer results to those found by O’Connell and co-authors. Section X discusses some

implications of our findings. Section XI concludes.

HOV reply to this article in the next volume of this journal, and we then respond to

their reply.5We urge those interested in fully understanding the differences in assumptions

and analysis that drive the differences in results to read those two articles, as well as this one

and the original HOV article. In particular, our response contains a more extensive state-

ment of our assumptions, and how they differ from HOV, than is offered below.

5Hersch, O’Connell & Viscusi (forthcoming); Black, Hyman & Silver (forthcoming).

“Early Offers” in Medical Malpractice Cases 725