Saturday, June 08, 2013

Nagin: Death Penalty Deterrence: Defended & Advanced

Nagin: Death Penalty Deterrence: Defended & Advanced 
Dudley Sharp, independent researcher, death penalty expert, former opponent, 832-439-2113, CV at bottom
minor updates, 12/2022

This is a rebuttal to two National Research Council (NRC) reviews, 1978 and 2012 (1), which allegedly show that the methodological integrity of some econometric studies, finding for death penalty deterrence, was so poor as to render them useless.

Such assertion is false, and you should speak to the Researchers, of those studies finding for deterrence (2), as well as finish this paper.

Have the studies, finding for death penalty deterrence, been nullified? No.

Keep in mind, as you read this, as well as Nagin's reports, that gross homicide rates cannot be how deterrence is measured.

Nagin must know this. So why didn't he stick to deterrence, per se, but instead, abandoned it for homicide rates?

The 10 quotes, below, are from an article, whereby academic detractors of deterrence, discuss the NRC Nagin reviews (3).

"Sharp" is my response to those comments.

"Nagin" -- The NRC reviews which include both studies and Dr. Nagin's fellow researchers for those two NRC reviews.

"Researchers" -- Those researchers, finding for deterrence, whose studies were included in those two NRC reviews.

1) "For decades, scholars have attempted to answer a seemingly simple question: Does the death penalty deter murder? Two National Research Council (NRC) reports, conducted more than three decades apart, have reached basically the same conclusion—we still do not know." (3)

Sharp: Untrue. The death penalty deters some, just as all criminal sanction deter some (4). Scholars, as all of us, know how people respond to negative and positive incentives. This has been known for thousands of years and is the foundation of economics.

What Nagin challenges is the methodology of some few of those Researchers.

Nagin's reviews claim that the studies finding for deterrence were flawed and, because of that, should not be relied upon.

Those are two very different things - evaluating research of a few studies and if there is a deterrent. Deterrence cannot be made to go away. Some studies are better, some worse, than others.

Neither Nagin review tells us that the death penalty does not deter. They can't. All sanctions deter some.

It seems clear that the 1978 Nagin review was, fully, rebutted (5).

Nagin referenced that 1978 review, repeatedly within the 2012 review. Yet, he failed to note the important rebuttals of the 1978 Nagin review. That is unfortunate and misleading, undermining Nagin's credibility.

2) “Research to date on the effect of capital punishment on homicide is not informative about whether capital punishment decreases, increases, or has no effect on homicide rates,” Nagin said, quoting the 2012 report.

Sharp: Again, untrue. Research, history and case studies show that the death penalty deters, just as all negative prospects/incentives do, even if there is no noticeable effect on homicide rates (5), as the deterrent effect would be measured by the net number of homicides being 1) higher, absent the death penalty/executions or 2) lower, with the death penalty/executions, whether or not the gross murder rates rose, lowered or stayed the same.

Negative incentives, just as positive ones, matter and effect behavior. What we are speaking of, here, is a review of a few econometric studies, wherein Nagin finds the Researchers lacking.

There is no indication that those Researchers agree with Nagin's conclusions. Nagin did not offer those Researchers the ability to review and comment on Nagin's findings, prior to publication, as should have occurred.

Nagin made sure that such review did not occur, because Nagin would have been rebutted, just as he was in 1978.

There are rebuttals to Nagin's criticism from the 1978 Researchers (5).

It appears that those Researchers are more credible than Nagin, as reviewed.

To accept Nagin's criticism is to accept that these Researchers:

1) don't know their science; 2) have no methodological credibility; 3) work within a discipline that has no credibility; and 4) don't know how to plan or conduct a study which includes proper constants, variables and controls, in order to measure their intended subject.

The problem, for Nagin, is that none of that is true. I encourage you to speak to those Researchers, listed below (2). Nagin, intentionally, didn't. 

These Researchers are respected in their field. Did their research and methodology, suddenly, fall apart, only in their death penalty research?  Of course not.

Is it surprising that this area of research, which has resulted in so many studies finding for death penalty deterrence (24 since 1996), is under assault? No.

Academia is, nearly, 100%, opposed to the death penalty. Some of the authors, finding for deterrence, have, publicly, stated their opposition to executions:

Mocan: "The (deterrence) results are robust, they don't really go away." "I oppose the death penalty. But my results show that the death penalty (deters) — what am I going to do, hide them?" "Science does really draw a conclusion. It did. There is no question about it." "The conclusion is there is a deterrent effect." Naci Mocan. Economics Chair, Louisiana State U, former Chair, Economics, U of Colorado (Denver).

Academics have biases like everyone else. Problems enter when that bias affects their research.

Is it a wonder why Nagin did not allow the Researchers to vet his paper? No.

Nagin's Conflicts of Interest

Dr. Nagin's academic chair is financed by a well-known anti-death penalty trust, Heinz, a major financier of anti-death penalty efforts (6). Two of the three funding agencies for the 2012 Nagin review are, also, very well-known supporters of anti- death penalty efforts, to the tune of millions of dollars, per year (6).

So, three out of four of Nagin's financial supporters, for this study, are anti-death penalty.

Conflicts of interest are, rarely, this obvious or severe. Should that make all question the NRC published reports? Yes. It's the elephant in the room, as a few other points are, as well.

Does NRC have no controls for such conflicts?  NRC states: "Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the views of the organizations or agencies that provided support for the project."

In other words, they take a pass on any credibility issues, a good idea with the Nagin reviews. Conflicts of interest don't seem to be an NRC concern.

Brutalization

I think it goes to Nagin's bias that he mentioned, within the 2012 paper, Dr. Joanna Shepherd's study finding for brutalization -- that executions increase murders (7).

Nagin, intentionally, omitted crucial points or he could not comprehend them. Ask him.

It appears Nagin just wanted to take a dig at the deterrence studies, by showing that even Shepherd, who found deterrence in a number of her other, previous, studies, now, found for brutalization.

What Nagin left out.

Even with those instances of brutalization, in Shepherd's study, the deterrent effect had a greater net effect, the net deterrent effect overwhelmed the brutalization effect -- A net number of innocent lives - 1672 - were saved (7), because death penalty deterrence overwhelmed brutalization.

Shepherd's study found that the deterrent effects overwhelmed the brutalization effects. Nagin either, intentionally, left that out or he couldn't understand Shepherd's results.

Either way, it destroys Nagin's credibility.

Just as problematic for Nagin, is that Nagin omitted that Shepherd found that if all state jurisdictions executed at least one murderer every two years, that deterrence would rise, substantially, that many more innocent lives would be spared and . . . brutalization disappears (7).

Shepherd's brutalization study provides more evidence for deterrence. Something we would expect a neutral academic to point out. Nagin did the opposite or he didn't understand the results. Either way, Nagin's credibility is zero.

Nagin, intentionally, left out those important findings, because they supported deterrence and because Shepherd's methodology was similar to the other Researchers and/or Nagin just didn't understand the study.

Shepherd's methodology is, virtually, identical, to some of those Researchers finding for deterrence, yet, Nagin did not criticize Shepherd's methodology, as Nagin did the other Researchers or, again, Nagin had no clue.

Why? Because Nagin liked Shepherd's results for brutalization, but not the other studies which found for deterrence, just as Shepherd's brutalization study did, as well.

Nagin either didn't understand Shepherd's "brutalization" study and/or he didn't want to expose its similar methodology to the Researchers' studies finding for deterrence and the fact that Shepherd's "brutalization" study found greater support for deterrence.

It is either one or the other, either of which should have excluded Nagin from being in this or any other research group and which destroys Nagin's credibility.

Either way, Nagin looks very bad, indeed.

3) "The (Nagin) committee . . . eventually reached the potentially controversial conclusion that research on the death penalty’s deterrent effect is so flawed that it cannot be used to determine if the death penalty does indeed affect homicide rates."

Sharp:  What is controversial and clear is Nagin's obvious bias and his severe conflict of interest.

The Nagin committee's conclusions are not controversial but are empty in the context of the Nagin bias, errors and conflicts.

It seems obvious that no one should seriously consider any of Nagin's conclusions based upon those problems, alone.

For anyone wishing a credible review of the Researchers methodology, we would need a neutral referee, who would provide solid peer review, inclusive of the Researchers and absent such obvious bias and conflicts of interest that Nagin brought to this latest review.

Speak to those Researchers and read Ehrlich's comments to the 1978 Nagin review, wherein Ehrlich appears to have, fully, rebutted Nagin's complaints (5). 

Ehrlich writes: "While the methodological advances in recent research on deterrence have, to a considerable extent, come from work by economists, and while studies following the economic approach area major focus of the (Nagin) Panel's work, not a single practitioner of the economic approach to crime is to be found among the panel's interdisciplinary roster of members. In contrast, the panel does include scholars who have pursued approaches in criminology that are seriously challenged by the economic approach and whose past work exhibits considerable skepticism, if not philosophical hostility, toward the deterrence hypothesis."

" . . . the imbalanced composition of the (Nagin) Panel may be partly responsible for the shortcomings of its work and conclusions, which are elaborated in the following sections." (5a, page 3)

From here, Ehrlich eviscerates both Nagin's understanding of the science and Nagin's conclusions.

NOTE: There is a common misunderstanding, parroted by Nagin, about the death penalty's effect on "homicide rates". see "HOMICIDE RATES & EXECUTIONS", below.

4) Nagin "singled out the key question from the report: Is capital punishment more or less effective as a deterrent than alternative punishments, such as a life sentence without the possibility of parole?" “None of the studies we reviewed—none—accounted for the noncapital portion of the sanction regime,” Nagin said.

Sharp:  I think Nagin is in error. I believe Ehrlich, in the earlier studies (1978), did have some controls for other sanctions. Ask Ehrlich. At this stage, asking Nagin would be unwise, unless you, also ask the specific Researcher that Nagin is referencing. 

The more recent studies may not have needed to review other sanctions, specifically, in order to isolate executions, as within the panel studies other sanctions are constants, within those jurisdictions and are, therefore, controlled.

Ask the Researchers, as Nagin did not.

Again, including Ehrlich's feedback, as well as other Researchers, would have been welcome, but Nagin excluded that, on purpose.

I suggest that adding other sanctions into the recent studies may be problematic and unnecessary. I think Nagin should consult with the Researchers on this, as Nagin should have, prior to publication.

It is a good point by Nagin, that we would all like to see comparisons of deterrence with life sentences vs executions.

Obviously, it would have been more advantageous, for us, if Nagin had reviewed that with the Researchers, prior to publication, so that we could have a discussion of those protocols within the report.

Again, Nagin excluded that critical opportunity. Let's not pretend that Nagin really wanted such a discussion, which Nagin, intentionally, excluded. I suspect Nagin knew the Researchers would have, totally, rebutted all of Nagin's positions, altering the outcome. Therefore, Nagin did not consult with them.

Maybe, next, time NRC may suggest to Nagin that some standard, peer review protocols should be followed, such as contacting the Researchers of those studies being reviewed, so that knowledge can be shared, and misunderstandings corrected.

One would have thought that Nagin learned that lesson from the earlier report. There is no question that Nagin didn't want those Researchers input, otherwise he would have contacted them.

Better yet, exclude Nagin from such future reviews and have a more balanced review, with a neutral referee, inclusive of some of the Researchers/authors of the studies being reviewed.

The key question is not is the death penalty an enhanced deterrent over life without parole. The evidence, without these studies, is that it is (4).

The question within the NRC report was the methodological reliability of the deterrence studies. Again, two very different things.

Is it possible that Nagin misunderstands the science?

Based upon Ehrlich's rebuttals (5), and Nagin's misuse of Shepherds's "brutalization" study, yes, it appears so.  

Either that and/or Nagin is so hostile to the death penalty and any research that may be seen to support it, as Ehrlich pondered, as Nagin's comments often seem to support that theory, just as Nagin's conflicts of interest suggest.

Does Nagin misunderstand these more recent studies, today? It seems likely, based upon Nagin's problems with the Shepherd "brutalization" study and his misuse of murder rates.

I think we need to wait to hear from the Researchers (2).

5) "In probing the studies, the NRC committee also found that they did not provide any plausible evidence on how potential murderers perceive, and respond to, capital punishment. Many studies fail to address how perceptions are formed, how they correspond with reality, or how they vary across states or over time. They simply infer that potential murderers respond to the objective risk of execution, Nagin said."

Sharp: This comment is curious. Criminals perceive from experience and learning and from outside sources, such as media, colleagues and other acquaintances, just as we all do.

Criminals understand incentives, negative or positive, and know that sanctions await them, should they be caught. There are police, witnesses, cameras, evidence, jails, prisons and executions -- criminals are very aware of the world they live in, again, just as we all are. The responses to those, by both criminals and potential criminals, may be either objective or irrational, thoughtful or subconscious, the same as responses by non-criminals, as with all, within their worlds.

When criminals "case" potential crime targets and decide not to commit the crimes, based upon any of those, it is based, solely, within deterrence.

Nagin appears to be the only person unaware of this.

6) "In 1975, economist Isaac Ehrlich published a now “infamous” study in the American Economic Review that concluded that eight lives are spared for every execution that takes place." "Though this finding has been widely rebuked since then, the work was up to the journal’s standards for its time, and thus the field, Manski said. So, he quizzed, what standards of proof should we now apply?"

Sharp: This is an interesting comment, because Nagin very much disagreed with Manski, as Nagin found many problems with Ehrlich's work, as per Nagin's earlier NRC published rebuke of Ehrlich.

As Manski, correctly states, Ehrlich's standards were solid for the time, therefore Nagin's criticism was inaccurate and Ehrlich agreed with Manski, as Ehrlich rebutted Nagin's criticism (5).

Again, Nagin's credibility suffers.

Nagin a very strange choice to head up a report on the same topic, this many years, later. There is also the major issue of the conflicts of interest.

Ehrlich: " . . . the (Nagin) Panel's reservations toward the reported findings of apparent deterrent effects stem not from any mistakes uncovered or from any fundamental methodological disagreements but instead are founded upon various conjectures - a level of criticism quite different in kind. Indeed, the impression derived from the entire document of the (Nagin) Panel is that the authors were not so much interested in rational and objective evaluation of the empirical evidence on deterrence as they were intent on showing that evidence to be defective. While the specific interpretations of statistical findings may quite rightly become the object of scholarly dispute, none of the work of the panel and its commissioned papers attempts to provide a systematic and comprehensive alternative explanation for the amalgam of cross-sectional and time series evidence consistent with the deterrence hypothesis. It seems inappropriate that evidence consistent with a set of detailed behavioral propositions emanating from a theory that also have proven useful in explaining a variety of other expressions of human behavior is hardly given equal weight to a set of speculations and some ad hoc behavioral propositions which do not derive from logical principles of general applicability." (5a, p 15)

7) "Then in the 1990s, a new generation of researchers took up the question. But this new generation of researchers committed many of the same mistakes as the previous one." “It’s like people in the 1990s didn't even read the 1978 report,” Manski said.

Sharp: This is in conflict with the Researchers finding for deterrence, who  claim they addressed the concerns from the prior criticisms. Were Nagin and Manski unaware? Ask the Researchers (2), don't just depend upon Nagin or Manski. The academic question is did Manski speak to those Researchers about his concerns? Probably not. It is all but guaranteed that all of the Researchers knew that Ehrlich had eviscerated Nagin's 1978 report, which Nagin, improperly, hid from readers.

So, we seem to have a series of conflicting claims which have not been resolved, as they could have been had the Researchers been invited for commentary but, instead, were, intentionally excluded.

8) "Policymakers and analysts almost always opt to use models that draw stark black-and-white conclusions, Manski said. But such certainty typically rests on flawed models and may lead to erroneous conclusions. Instead, he suggested reporting a range of estimates for the impact of a particular policy change, derived from a corresponding range of plausible models. This would generate more honesty in policy analysis, he said."

Sharp: Honestly, these recent series of independent studies, by different Researchers, found a very wide range of deterrent effects, from 1-28 potential murderers deterred, and did so with a wide range of research which was the opposite of a stark black and white conclusions, with most researchers noting that their conclusions should not be looked at as, let's say, "stark black and white conclusions." While the Researchers, firmly, stand by their research, methodology and conclusions, they note the variables within their work and findings and conclude that their work is suggestive and/or supportive of deterrence, but that is about as stark a conclusion as can be made or has been made with a soft science.

To be clear, the uncertainty is not the presence of deterrence, but the measurable impact of deterrence, by the social sciences.

Cloninger and Marchesini “Using differing methodologies and data sets at least five groups of scholars each working independently (and often without knowledge of the others) have arrived at the same conclusion, there is significant and robust evidence that executions deter some homicides. While there may be merit in some of (Donahue & Wolfers’) specific criticisms, none addresses the totality of the collection of studies. The probability that chance alone explains the coincidence of these virtually simultaneous conclusions is negligible.” (8a)

It should be noted that the Researchers destroyed  Donahue and Wolfer's criticism (8a), but Nagin, repeatedly depended upon their work within the 2012 review, without mentioning those rebuttals, another blow against Nagin's neutrality and credibility.

Criminals, as all of us, respond to incentives. Executions are the most negative of incentives that criminals will face.

No one should be surprised that Researchers confirm such, as we already know it to be true.

9) "Though it concludes that studies to date are not rigorous enough to prove one way or the other the effects of capital punishment on homicide rates, “the report does point out that judgments about whether capital punishment deters or not are still relevant to policy deliberations,” Nagin stressed. “It’s just that people should just not appeal to this evidence in support of their opinions about whether capital punishment has a deterrent effect or not.”

Sharp: Even without the recent 24 studies finding for deterrence, we know the death penalty deters and deters more than a life sentence, just as all sanctions deter some (4). What is being questioned is the reliability of the methods used to reach that conclusion.

Based upon the many major problems, Nagin's review should be rejected.

NOTE that there have been two additional reviews, critical of researchers finding for deterrence, and both of those have been rebutted (8).

10) “The great value of this report is that it clears the air about what we know—and don't know—about the death penalty,” Manski said.

Sharp: No, the Nagin review is about methodological credibility, which is not, remotely, clear, as conflicting claims remain.

This is about the studies of the Researchers, only, not about the death penalty, in general, or deterrence, in particular, but the credibility of those studies.

Nagin's methodology, if any, seems to be the major problem, as with his bias, credibility, errors and conflicts of interest.

Manski's claim that Nagin clears the air is far from reality.

Nagin, once again, has muddied the waters.

With the contributions of the Researchers and a neutral referee and the absence of Nagin, this could have been a more thorough, reliable and useful review.

See ADDITIONAL NAGIN PROBLEMS, below

======

"HOMICIDE RATES & EXECUTIONS"

Accepting that 1-28 homicides are deterred, per execution, as per the Researchers, would mean 33- 924 homicides prevented per year, based upon the deterrent effect, or an average of 478 per year, or only 2.6% of the approximate average of 18,000 murders per year (1973-2012).

While that is a huge savings in innocent life it may have very little impact on homicide rates, possibly, causing a murder rate of 10 to drop 0.2, based only on deterrence, but with other factors, possibly, raising it above 11 or below 8, while still accounting for the lives saved by that 0.2 murder rate reduction, because of deterrence - the obvious reason that you can't measure deterrence by gross murder rates, as Nagin seems unaware.

Because of other variables, each year, homicides and homicide rates may go up, down or remain the same, even while an average of 478 innocent lives are being saved, every year, via deterrence.

Let's say that Iceland and Singapore have the lowest crime rates and murder rates in the world. As every other country and city in the world have higher crime and murder rates, does that mean that no murderers nor other criminals are deterred in all other countries and cities, only because their rates are higher than Iceland and Singapore? 

Of course not.

That is why gross homicide rates cannot be how deterrence is measured.

Nagin must know this. So why didn't he stick to deterrence, per se, but instead, abandoning it for homicide rates?

Although executions in the modern era didn't start until 1977, some studies found that just the existence of the death penalty deterred murders. Therefore, I began the period at 1973, which was the first year that new death penalty statutes came into law, post Furman v Georgia.

======

ADDITIONAL NAGIN PROBLEMS

Nagin stated they were only going to discuss deterrence, yet Nagin, somehow, brought several anti-death penalty deceptions into the 2012 Nagin review.

1) "Liebman, Fagan and West (2000) found that two-thirds of capital sentences are reversed on appeal."

From Nagin at
http://www.nap.edu/openbook.php?record_id=13363&page=89 

I suspect Nagin didn't fact check/vet, because the study is an anti-death penalty staple and isn't true.

Some fact checking/vetting, which I did, with Nagin skipping that important process.

James Liebman's Broken Study: Review of A Broken System

2) Likely, Nagin never fact check the cost studies, either. Again, a hit on Nagin's credibility. Nagin cost references:



Some fact checking/vetting, which I did, with Nagin skipping that important process.

Death Penalty Costs: Saving Money

======
Footnotes:

1) a) National Research Council. (2012). Deterrence and the Death Penalty. Committee on Deterrence and the Death Penalty, Daniel S. Nagin and John V. Pepper, Eds. Committee on Law and Justice, Division of Behavioral and Social Sciences and Education. Washington, DC: The National Academies Press.

b) DETERRENCE AND INCAPACITATION - ESTIMATING THE EFFECT OF CRIMINAL SANCTIONS ON CRIME RATES (1977), Editor(s): A BLUMSTEIN ; J COHEN; D NAGIN, National Academy of Sciences, https://www.ncjrs.gov/App/Publications/abstract.aspx?ID=44669

2) These are but a few of the researchers, from whom others may be contacted, as well.

Dale Cloninger, Professor of Finance and Economics, Interim Dean of the School of Business and Public Administration, U of Houston (Clear Lake), cloninger@cl.uh.edu

Hashem Dezhbakhsh, former Chair, Economics, Emory U., econhd@emory.edu

Issac Ehrlich, Chair, Economics, U of Buffalo (SUNS), mgtehrl@buffalo.edu


H. Naci Mocan, Chair, Economics, Louisiana State U, and  former Chair, Economics, U of Colorado (Denver), mocan@lsu.edu

3) The 10 quoted sections all come from "The Death Penalty: Does It Deter Killings?", Institute for Policy Research, Northwest University, NEWS2013

4) OF COURSE THE DEATH PENALTY DETERS: A review of the debate

5) Rebuttals to Nagin's 1978 review

a) Fear of Deterrence -- A Critical Evaluation of the Report of the Panel on Research on Deterrent and Incapacitative Effects, Journal of Legal Studies, Vol. 6 (2), June 1977.

b) Capital Punishment and Deterrence: Some Further Thoughts and Additional Evidence. Ehrlich, Isaac, Journal of Political Economy, Vol. 85, No. 4, pp. 741-88, August 1977. Available at SSRN: http://ssrn.com/abstract=961491

c) Sensitivity Analyses of the Deterrence Hypothesis: Let's Keep the Econ in Econometrics. Ehrlich, Isaac and Liu, Zhiqiang, Journal of Law & Economics, Vol. 42, No. 1, pp. 455-87, April 1999.  Available at SSRN:, http://ssrn.com/abstract=961447

6) The conflicts of interest could hardly be more obvious and severe.

Daniel Nagin, Chair, the Teresa and H. John Heinz III University Professor of Public Policy and Statistics, Carnegie Mellon University.

Nagin's income/Chair is financed by a liberal foundation, The Heinz Family Foundation, which provides millions of dollars for anti-death penalty efforts

Two of the three funding groups for the later NRC study, Tides Foundation and the Proteus Action League, are well known anti death penalty financiers, to the tune of millions of dollars per year.

Tides gets major funding from Heinz. Check on Proteus.

Proteus Action League

7) Brutalization & The Death Penalty: More Support for the Deterrent Effect

8) a) DEATH PENALTY DETERRENCE: Rebuttal to Donahue and Wolfers: