* This story was expanded since its original publication to further detail Stuckler's response to his critics.
Does "shock therapy" economic reform really kill? The question has been debated by experts since a British medical journal, “The Lancet,” published a study last year claiming there were links between mass privatization and rising death rates in the former Soviet republics.
Economists and sociologists have been closely examining the work of a team led by David Stuckler -- a sociologist and public-health expert at Oxford University. The latest studies say Stuckler's original Lancet article was wrong because it was based on faulty assumptions.
Economist John Earle says he was surprised when he first read the work of Stuckler's team. As a professor at Central European University in Prague and Budapest, Earle has spent years since the collapse of communism studying the economic transition to capitalism.
Looking at the same data used in Stuckler's original “Lancet” study, Earle and his colleagues found no evidence to support the widely-reported claim that shock therapy reforms caused mortality rates to rise during the 1990s.
"The original study received a lot of attention from press around the world -- in Asia, North America, and Europe," Earle told RFE/RL. "We were very surprised by the findings and the fact that the [media] reporting on it was generally so uncritical -- that is, accepting the findings at face value. So we thought that it deserved another look."
"That was our motivation," Earle continues. "It seemed to be becoming part of received wisdom on the transition that this had occurred. Once we found that it should not be part of the received wisdom, we thought that it was important to set the record straight."
Hot Debate
The latest edition of “The Lancet” includes a summary of Earle's findings along with his criticism of Stuckler's research model. Earle's follow-up study also has been published in its entirety by the Michigan-based Upjohn Institute for Employment Research, where Earle also is employed.
"In summary, what we found was, through a number of checks, that the original results were highly sensitive to the nature of the statistical analysis that was employed," Earle says. "One could not say that there is, in fact, any relationship between privatization and mortality in those countries."
Earle says one fault was that Stuckler's team did not account for any time lag between the start of privatization programs and changes in mortality rates.
"The original analysis had claimed that privatization would lead to layoffs of workers, and [when] these laid off workers would become unemployed, their health would deteriorate and they would die," he says. "One would not expect when a policy has begun to be implemented that immediately people would start dying. Instead, it would take at least a year or two."
That same point was raised in a letter, published in the latest edition of “The Lancet,” from a research team led by Christopher Gerry -- a senior lecturer at University College London's School of Slavonic and East European Studies.
Noting that the Russian privatization program was announced in December 1992 and completed in June 1994, Gerry says the reforms cannot plausibly be claimed to have affected Russia's mortality rates at all during 1992 and, at most, only weakly during 1993.
Earle says once data used by Stuckler is checked for time lags, "the correlation between privatization and mortality is greatly weakened and eventually disappears."
Response To Critics
Stuckler responded to the latest arguments in the debate by charging that Earle and Gerry manipulated his original data and committed a series of factual errors -- including dates for key reforms.
He also accused Earle and Gerry's teams of "data torture" -- that is, "blatant statistical manipulations to fit their hypothesis" that he says owe more to efforts to discredit his research than trying to find scientific truths.
"They weren't replicating our study," Stuckler told RFE/RL. "They started off by replicating our study but then they made a series of statistical manipulations that fundamentally changed our study so that they were no longer testing whether mass privatization is linked to mortality."
Stuckler also says that his team has been surprised by the "vitriolic nature" of other critiques they have faced during the past year by "advocates of privatization." He said many of the critics have conflicts of interest because they took significant funding from institutions for advising governments to push for radical, rapid reforms.
Stuckler says his work sought to explain why United Nations data shows that one of the world's worst peacetime mortality crises occurred during the transition to capitalism in former Soviet republics.
"The United Nations estimated that in the early 1990s, there were 3 million avoidable deaths that occurred over and above historical trends," he explained. "This has been a puzzle to the field of public health. No one expected so many working-aged men, in particular, to die in connection with the transition to capitalism.”
“What our study shows is that these deaths were not simply inevitable, but that they were connected with a specific strategy -- in particular, the 'big bang' rapid approach to building capitalism out of communism," Stuckler says.
Still, Stuckler says journalists were wrong to report that his research shows mass privatization was a direct cause for rising mortality rates.
"As with any statistical study, it is nearly impossible to make the leap from association to causation," Stuckler says. "But what a good scientist will do is test as many implications of the theory or the hypothesis as possible. As a social scientist, I think of it as building a case for causality. We were very careful in the paper to say that mass privatization was 'strongly linked' and 'associated' with these rises."
‘Disingenuous’ Reply
Earle quotes Stuckler's original “Lancet” article and rejects his explanation of misrepresentation as "disingenuous."
"This is a direct quote: 'Privatization was a crucial determinant of differences in mortality rates across the countries.' That is a claim about causality," Earle says. "'Determinant' means one variable determines another."
"The authors [of the original Lancet study] also draw policy conclusions from their analysis," Earle continues. "They say that future governments should be cautious about privatizing firms because, according to them, mortality rose after privatization. If this is merely a correlation and not causation, then there is no reason to draw that policy conclusion."
At University College London, Gerry charges that Stuckler's team failed to understand the "dynamic nature of the processes that underpin the mortality trends" in the former Soviet republics. For example, Gerry says, there were no controls to account for deadly diseases stemming from exposure to pollution in the earlier years of communism.
Earle says the original research should have examined longer mortality trends in the former Soviet republics instead of focusing only on the 1990s. If Stuckler's team had done that, Earle says, they would have discovered that mortality rates began rising in the Soviet Union as early as 1985 and 1986 -- seriously weakening the claim that post-communist economic shock therapy kills.
Does "shock therapy" economic reform really kill? The question has been debated by experts since a British medical journal, “The Lancet,” published a study last year claiming there were links between mass privatization and rising death rates in the former Soviet republics.
Economists and sociologists have been closely examining the work of a team led by David Stuckler -- a sociologist and public-health expert at Oxford University. The latest studies say Stuckler's original Lancet article was wrong because it was based on faulty assumptions.
Economist John Earle says he was surprised when he first read the work of Stuckler's team. As a professor at Central European University in Prague and Budapest, Earle has spent years since the collapse of communism studying the economic transition to capitalism.
Looking at the same data used in Stuckler's original “Lancet” study, Earle and his colleagues found no evidence to support the widely-reported claim that shock therapy reforms caused mortality rates to rise during the 1990s.
"The original study received a lot of attention from press around the world -- in Asia, North America, and Europe," Earle told RFE/RL. "We were very surprised by the findings and the fact that the [media] reporting on it was generally so uncritical -- that is, accepting the findings at face value. So we thought that it deserved another look."
"That was our motivation," Earle continues. "It seemed to be becoming part of received wisdom on the transition that this had occurred. Once we found that it should not be part of the received wisdom, we thought that it was important to set the record straight."
Hot Debate
The latest edition of “The Lancet” includes a summary of Earle's findings along with his criticism of Stuckler's research model. Earle's follow-up study also has been published in its entirety by the Michigan-based Upjohn Institute for Employment Research, where Earle also is employed.
"In summary, what we found was, through a number of checks, that the original results were highly sensitive to the nature of the statistical analysis that was employed," Earle says. "One could not say that there is, in fact, any relationship between privatization and mortality in those countries."
Earle says one fault was that Stuckler's team did not account for any time lag between the start of privatization programs and changes in mortality rates.
"The original analysis had claimed that privatization would lead to layoffs of workers, and [when] these laid off workers would become unemployed, their health would deteriorate and they would die," he says. "One would not expect when a policy has begun to be implemented that immediately people would start dying. Instead, it would take at least a year or two."
That same point was raised in a letter, published in the latest edition of “The Lancet,” from a research team led by Christopher Gerry -- a senior lecturer at University College London's School of Slavonic and East European Studies.
Noting that the Russian privatization program was announced in December 1992 and completed in June 1994, Gerry says the reforms cannot plausibly be claimed to have affected Russia's mortality rates at all during 1992 and, at most, only weakly during 1993.
Earle says once data used by Stuckler is checked for time lags, "the correlation between privatization and mortality is greatly weakened and eventually disappears."
Response To Critics
Stuckler responded to the latest arguments in the debate by charging that Earle and Gerry manipulated his original data and committed a series of factual errors -- including dates for key reforms.
He also accused Earle and Gerry's teams of "data torture" -- that is, "blatant statistical manipulations to fit their hypothesis" that he says owe more to efforts to discredit his research than trying to find scientific truths.
"They weren't replicating our study," Stuckler told RFE/RL. "They started off by replicating our study but then they made a series of statistical manipulations that fundamentally changed our study so that they were no longer testing whether mass privatization is linked to mortality."
Stuckler also says that his team has been surprised by the "vitriolic nature" of other critiques they have faced during the past year by "advocates of privatization." He said many of the critics have conflicts of interest because they took significant funding from institutions for advising governments to push for radical, rapid reforms.
Stuckler says his work sought to explain why United Nations data shows that one of the world's worst peacetime mortality crises occurred during the transition to capitalism in former Soviet republics.
"The United Nations estimated that in the early 1990s, there were 3 million avoidable deaths that occurred over and above historical trends," he explained. "This has been a puzzle to the field of public health. No one expected so many working-aged men, in particular, to die in connection with the transition to capitalism.”
“What our study shows is that these deaths were not simply inevitable, but that they were connected with a specific strategy -- in particular, the 'big bang' rapid approach to building capitalism out of communism," Stuckler says.
Still, Stuckler says journalists were wrong to report that his research shows mass privatization was a direct cause for rising mortality rates.
"As with any statistical study, it is nearly impossible to make the leap from association to causation," Stuckler says. "But what a good scientist will do is test as many implications of the theory or the hypothesis as possible. As a social scientist, I think of it as building a case for causality. We were very careful in the paper to say that mass privatization was 'strongly linked' and 'associated' with these rises."
‘Disingenuous’ Reply
Earle quotes Stuckler's original “Lancet” article and rejects his explanation of misrepresentation as "disingenuous."
"This is a direct quote: 'Privatization was a crucial determinant of differences in mortality rates across the countries.' That is a claim about causality," Earle says. "'Determinant' means one variable determines another."
"The authors [of the original Lancet study] also draw policy conclusions from their analysis," Earle continues. "They say that future governments should be cautious about privatizing firms because, according to them, mortality rose after privatization. If this is merely a correlation and not causation, then there is no reason to draw that policy conclusion."
At University College London, Gerry charges that Stuckler's team failed to understand the "dynamic nature of the processes that underpin the mortality trends" in the former Soviet republics. For example, Gerry says, there were no controls to account for deadly diseases stemming from exposure to pollution in the earlier years of communism.
Earle says the original research should have examined longer mortality trends in the former Soviet republics instead of focusing only on the 1990s. If Stuckler's team had done that, Earle says, they would have discovered that mortality rates began rising in the Soviet Union as early as 1985 and 1986 -- seriously weakening the claim that post-communist economic shock therapy kills.
