Next | ToC | Prev


People Died at Three Mile Island

Gordon MacLeod sat across from the governor of Pennsylvania. It was October 9, 1979. MacLeod had been state secretary of health since twelve days prior to the accident at Three Mile Island.

A tall, trim Bostonian, MacLeod was a lifelong Republican who had served in Richard Nixon's Department of Health, Education, and Welfare. As both a medical doctor and an engineer he had moved from a research fellowship at Harvard Medical School to a chairmanship at the University of Pittsburgh's Graduate School of Public Health.

In 1979 Governor Richard Thornburgh, a neighbor of MacLeod's, had urged him to take charge of the state's Department of Health, which was in disarray. MacLeod had resisted, but finally agreed, with the understanding he would serve just two years, then return to academia.

Now, eight months later, as controversy still raged over how much radiation had been released at Three Mile Island, the governor's office called the secretary of health for a conference. The meeting began with some small talk, MacLeod told us a year later. And then Thornburgh got to the point. "`Gordon,' the governor said, `I'm going to have to ask for your resignation.'"

"I just sat there," MacLeod told us, "stunned. After going to all that trouble to get me to come on board, he was now telling me to leave after just eight months because things were `just not working out.'"[1]

Thornburgh's public explanation for MacLeod's firing was a "difference in institutional style." But the state media had other ideas. As the UPI reported it, MacLeod had been "state government's harshest critic of the way the Thornburgh administration responded to the Three Mile Island accident. And that may have been why he was fired."[2]

Indeed, MacLeod's problems with Thornburgh had begun on March 29, the day after news of radioactive releases from TMI began to spread. MacLeod had, in his words, "recommended and, on the next day, urged the governor in the strongest possible terms to call for the departure of pregnant women and young children from an area within five miles of the Three Mile Island plant." MacLeod told us later that if he had a chance to do it over, he would also have urged the departure of children in puberty, who are also extraordinarily radiation-sensitive.

But the state's nuclear engineers and radiation health physicists disagreed with MacLeod, and they told the governor there was no need for an evacuation. Initially Thornburgh advised area residents to stay indoors, but said nothing about evacuating.[3]

Meanwhile Dr. Ernest Sternglass had gone to Harrisburg the day after the accident. After testing on his own and finding high radiation levels, he urged that the state evacuate pregnant women and small children. He was worried in particular that I-131 doses could prove devastating to the small children and infants in utero who were particularly vulnerable to miscarriages, stillbirths, malformations, childhood leukemias, and other radiation-linked problems. Thornburgh publicly charged Sternglass with being an alarmist and stood firm in his refusal to call for an evacuation.

That night the state's Department of Environmental Resources announced that because the holding tanks at TMI were overloaded with radioactive liquids, Met Ed had been flushing them for hours into the Susquehanna River. No one had bothered to notify communities downstream that were continuing to draw their drinking water from the river.[4]

Finally Thornburgh asked NRC chairman Joseph Hendrie, a nuclear engineer, what he would do if he had a pregnant wife in the area. Hendrie replied that he would get her out "because we don't know what is going to happen."[5]

Thornburgh then decided to do what MacLeod had quietly urged and what he had attacked Ernest Sternglass for publicly suggesting. At noon on March 30--two days after the start of the accident--he announced that he was "advising those who may be particularly susceptible to the effects of radiation, that is, pregnant Women and pre-school-age children, to leave the area within a 5-mile radius of the Three Mile Island facility until further notice."[6]

1. Gordon MacLeod, interview, October 1980; see also, Anna Mayo, "The Nuclear State in Ascendancy," Village Voice, October 22-28, 1980.

2. York Daily Record, November 5, 1979, p. 4-A.

3. MacLeod interview, and Gordon MacLeod, "Three Mile Island and the Politics of Public Health," prepared for presentation for the New York City and Old Westbury chapters of Physicians for Social Responsibility, Columbia University, New York, November 22, 1980 (hereafter cited as "Politics"); see also, Gordon MacLeod, "Some Public Health Lessons from Three Mile Island: A Case Study in Chaos," Ambio 10, No. 1 (1981).

4. Washington Post, Crisis.

5. Kemeny Report, p. 118.

6. Washington Post, Crisis.

Public Health in Crisis

Meanwhile Gordon MacLeod was desperately trying to choreograph an official health response. Despite the recommendations of the Shapp Commission at Shippingport six years earlier MacLeod found the state woefully unprepared for a nuclear incident. "There was not even a book on radiation medicine in the department," he said. "Nor was there a single physician specially trained in radiation medicine anywhere in the Pennsylvania state government."

Apparently the NRC was equally unprepared. As the accident progressed, MacLeod asked the commission to send in a doctor trained in the field of radiation health. "I was told," MacLeod said, "that the NRC had no physicians on its staff, much less a physician trained in the field of radiation medicine."[7]

The commonwealth also tried to get from the federal government a supply of potassium iodide, a liquid that may be taken to block the ingestion of radioactive iodine from the atmosphere, and thus prevent harm to the thyroid.[8] Finally, five days after the accident--far too late for it to do much good--eleven thousand "little brown vials" arrived. According to MacLeod, six thousand of them were unlabeled. Many of the droppers yielded only half the correct dose. Other droppers did not fit the vials at all. And many of the vials contained "hairlike filamentous material and other particulate matter."[9]

Despite his belief that the official health response to the accident had been grossly inadequate, MacLeod remained a supporter of atomic energy. "I personally believe," he told a conference at Columbia University eighteen months after the accident, "that nuclear power can be made relatively safe if we don't ignore the public health lessons of the past."[10]

But MacLeod worried that those "health lessons" were being ignored. Several months after the accident he clashed with Thomas Gerusky, chief of radiation monitoring at the state's Department of Environmental Resources, who had opposed the early evacuation of pregnant women and small children from the area. MacLeod found Gerusky's testimony to the Kemeny Commission inquiry into the accident to be misleading. In the fall he wrote a letter to Commission chairman John Kemeny outlining his objections to what Gerusky had said.[11]

That, apparently, was the last straw for the Thornburgh administration. Word soon spread that MacLeod's behavior during the accident had been "erratic." By mid-October he had been removed from his position.[12] The state media interpreted MacLeod's dismissal as a concession to the nuclear industry. Some atomic backers had deeply resented the call for the TMI evacuation, saying it had been unnecessary, unwarranted, and had served only to frighten the population.

But the firing of Gordon MacLeod hardly ended the controversy over the health impact of the accident and how it had been handled. In November, Ernest Sternglass charged that figures from the nearby Harrisburg and Holy Spirit hospitals indicated that infant deaths there had doubled from six during February through April of 1979 to twelve in May through July. Only one infant had died at the Harrisburg Hospital in May through July of 1978; seven had died there in those same three months following the accident. The statistics seemed tragically reminiscent of the era of nuclear bomb testing. The NRC, the state, and the utility had all claimed--as had the AEC after so many atomic explosions--that radiation releases had been too small to have more than a very marginal health impact, if any at all. Sternglass asserted the authorities had failed to account for the extreme sensitivities of fetuses in utero in claiming a very marginal health impact from the accident's releases.[13]

He also pointed out a crucial shortcoming in the method of calculating estimated doses from a nuclear accident. An average population dose had been set by estimating how much radiation had been released and making calculations around the two million people in the fifty-mile radius around TMI. But, he said, the winds during the most crucial hours of the accident--when most of the radiation was released--generally headed to the west, northwest, and north. Thus the real doses were impacting not the vast surrounding population, but the people specifically in the path of the plume. And as Chairman Joseph Hendrie had confirmed on March 30, in the midst of the accident, the doses to individual areas where the plume touched the ground were "husky" and in the range of 120 millirems per hour and more, quantities easily large enough to cause severe damage to fetuses in the womb.[14]

Sternglass now charged that the doses had in fact impacted people in the path of the plume, and with visible effect. Syracuse, Rochester, and Albany had all received windblown doses from the plant, he said, and had suffered rising infant deaths. A preliminary study by the Canadian journal Harrowsmith indicated a possibly similar pattern emerging among infants born at eastern Ontario and western Quebec hospitals, due to radiation from nearby Nine Mile Point.[15]

In December of 1979, Sternglass carried his conclusions much further. In a paper delivered to the Fifth World Congress of Engineers and Architects at Tel Aviv, he said that data from the U.S. Bureau of Vital Statistics showed that there were "242 [infant] deaths above the normally expected number in Pennsylvania and a total of 430 in the entire northeastern area of the United States," a rise of clear statistical significance. The linkage with TMI was clear because "large amounts of radioactive Iodine-131 were released from the plant" and the peak of infant mortality came within a matter of months thereafter. The greatest rises took place near the plant, with effects decreasing as a function of distance away from Harrisburg.

He backed up his case by analyzing the amount of radiation to which pregnant women downwind might have been subjected. Accepting minimum official estimates, Sternglass calculated that the doses of radioactive I-131 alone could have been on the order of one hundred millirems to individual pregnant women in the path of the plume. Such doses, he said, were clearly capable of causing rises in infant mortality.[16]

Using federal statistics, Sternglass then demonstrated that Pennsylvania's infant death rate in July was the highest of any state east of the Mississippi that month (except for Washington, D.C.), although Pennsylvania usually has one of the lowest rates in the nation. He went on to say that a similar rise was evident in infant-mortality rates in northern New England--where wind had carried fallout from the plant--as opposed to southern New England, where it had not.[17]

The hypothesis was confirmed by the fact that infant-death rates began to fall again after the accident. This, he said, was predictable because embryos in utero who were too small to have developed a thyroid, or who were conceived after the accident, would not have been affected by their mothers' ingestion of radioactive iodine.

But I-131 was not the only radioactive element released from TMI--nor were infants the only humans likely to be harmed. Strontium 90, cesium 137, noble gases, and other disease-causing isotopes may also have escaped. Overall, said Sternglass, increases in cancers, leukemia, and a wide range of other diseases were "likely to occur." The Three Mile Island accident, he predicted, "will turn out to have produced the largest death toll ever resulting from an industrial accident, with total deaths from all causes likely to reach many thousands over the next 10 to 20 years."[18]

7. MacLeod,"Politics."

8. A discussion of the efficacy of potassium iodide in preventing harm during a nuclear accident appears in Jan Beyea, Some Long-Term Consequences of Hypothetical Major Releases of Radioactivity to the Atmosphere from Three Mile Island (Princeton, N.J.: Center for Energy and Environmental Studies, 1979).

9. MacLeod interview.

10. MacLeod, "Politics."

11. MacLeod interview.

12. Pawlick, "Silent Toll. "

13. Ernest Sternglass, "Infant Mortality Changes Following the Three Mile Island Accident," presented at the 5th World Congress of Engineers and Architects, Tel Aviv, Israel, 1980 (hereafter cited as "Infant Mortality Changes"); and Pawlick, "Silent Toll."

14. Pittsburgh Post-Gazette, April 16, 1979, p. 1.

15. Pawlick, "Silent Toll," and Sternglass, "Infant Mortality Changes."

16. Ibid.; and Sternglass interview.

17. Sternglass, "Infant Mortality Changes"; see also, Sternglass, "TMI Update," Nation, April 25, 1981.

18. Sternglass, "Infant Mortality Changes."

Pennsylvania Denies Infant Deaths

The charge that TMI had actually killed area infants provoked a storm of outrage from the government of Pennsylvania. The state responded--as it had at Shippingport six years earlier--that the official statistics Ernest Sternglass had used were, after all, inaccurate. Dr. George Tokuhata, director of the state's Department of Epidemiological Research, said a "printing error" on the part of the U.S. Bureau of Vital Statistics had skewed the state's infant-mortality figures. There were thus eighty-eight fewer infant deaths in Pennsylvania in the summer of 1979 than originally recorded.[19]

Sternglass, however, held his ground. Discrepancies between state and federal data are not uncommon. But this particular case seemed "suspicious." The discrepancy in infant deaths between the two sources for the period of April 1 through June 30, 1979, had been two; from October 1 through December 31 it had also been two. For eighty-eight to surface between July 1 and September 30, precisely in the controversial summer months after the TMI accident, seemed unlikely.[20]

But even after subtracting those eighty-eight infant deaths, Sternglass said, Pennsylvania's infant-mortality rate still exceeded the average U.S. rate--contrary to normal patterns--"and also exceeded it for every month following the accident up to December." He thus maintained that his overall conclusions about infant deaths in the Northeast remained unchanged.[21]

Tokuhata nonetheless told The New York Times, which accepted his remarks uncritically, that Sternglass's analysis had been based on "the wrong number." Tokuhata further charged that Sternglass's reports were "full of problems" and had been built around "methodologies inconsistent with standard epidemiological procedures." He had, said Tokuhata, "selected areas for analysis that fit his hypothesis ignoring those close to the reactor where the infant mortality rate was very low."[22]

Sternglass replied--though not in The New York Times--that he had not initially studied those close-in areas in his Tel Aviv report "because Tokuhata himself had refused to make [those figures] available to me at the time I was doing my study."[23]

In an interview with us Tokuhata denied having ever refused to give Sternglass the data. "He must have asked some other department," Tokuhata said. "We never refused."[24]

But in fact the infant-mortality statistics around TMI only became public in the winter of 1980, when Dr. MacLeod--who had since returned to the University of Pittsburgh--began receiving calls from his former colleagues. Anonymous members of the department told MacLeod that the state was suppressing statistics that indicated a rise in infant-mortality rates near TMI. Alarmed by what MacLeod termed a "restrictive policy" on health data, he released the numbers in a pulpit address at Pittsburgh's First Unitarian Church. That, in turn, forced the Department of Health to make the figures officially public.[25]

And the numbers apparently confirmed the public's worst fears. In the six-month period following the accident, in a ten-mile radius around TMI, thirty-one infants had died. In 1978 the number was only fourteen; in 1977 it had been twenty.

Tokuhata told the Times the apparently sharp rise in infant deaths in 1979 was not significant because that was an absolute number, not a rate of deaths against live births. Times reporter Jane Brody paraphrased him thusly: "When the 31 infant deaths were considered in relation to the number of live births, no statistically meaningful difference was found."[26]

But in fact, said Sternglass, these "preliminary" figures showed an infant-death rate within that ten-mile radius of 7.2 per 1000 live births in 1978; in 1979, after the accident, the number had risen to 15.7 per 1000--a more than doubling.

The numbers for infant-death rates within a five-mile radius of TMI--though small--were even more damning. In 1978 the rate had been 2.3 infant deaths per 1000 live births; in 1979, after the accident, it was 16.2--a jump by a factor of seven.[27]

But the state had an explanation. At a press conference in April of 1980 Dr. H. Arnold Muller, who had taken MacLeod's place as secretary of health, announced that the TMI-area infant-death rates showed "no statistically significant difference in the mortality rate than for the state as a whole."[28] To support its case the state introduced a racial factor. Black people, it said, are known to suffer a higher rate of infant mortality than whites. Thus the presence of large numbers of blacks in Harrisburg--some of whom had been counted into the figures for the ten-mile radius around TMI--had made the local infant-mortality rates seem unduly large. As Tokuhata was paraphrased by Jane Brody in the Times: ". . . when analyzed without taking into account Harrisburg, where the large black population ordinarily has a much higher infant mortality rate than the rest of the region, the rate for the population living within 10 miles of the plant was the same as that for the state as a whole" (our emphasis).[29]

The analysis was deceptive.

The charge that TMI had killed nearby infants had nothing to do with a comparison with the state average. It had been based on comparing death rates in the same area from the spring and summer of 1977 and 1978--before the accident--against those of the spring and summer of 1979--after the accident.

To subtract the figures for black people from the 1979 statistics without doing the same for 1978 and 1977 would have made sense only if they had all moved into Harrisburg the day of the accident.

And in fact, prior to the accident, the infant-mortality rate in the TMI area had been generally lower than the state average. If it now equaled the state average, that would really mean a significant rise in the normal local infant-mortality rate.

But the state's own statistics showed the local rates were actually above the state average anyway. From April 1 to September 30, 1979, the state infant-mortality rate was put at 13.3 per 1000. For that period in the ten-mile radius around TMI it was 15.7; for the five-mile radius (which excluded Harrisburg's blacks) it was 16.2. Thus in the same press release in which the state had claimed preliminary proof that the TMI accident had harmed no babies, its own figures indicated precisely the opposite.[30]

Meanwhile, another controversy had erupted over an unexpected outbreak of thyroid-deficiency problems among infants born to the southeast of the plant. Again Gordon MacLeod was responsible for making the information public. Having been informed by colleagues still within the Department of Health that the numbers had surfaced, MacLeod privately asked the state--four times--to release the data. The state refused all four times. So MacLeod alerted a UPI reporter that "there's a story over there," and it was soon in the newspapers.[31]

The problem focused around thirteen cases of infant hypothyroidism in an area where normally three such cases would be expected. MacLeod was particularly sensitive to the state's withholding data on hypothyroidism because it is a disorder that can be easily treated at birth with the administration of iodine supplements. Untreated, it can cause brain damage and a wide variety of other serious defects. "The most important thing was that there be some opportunity to prevent the disease," MacLeod told us. Warning the public "might help pick up any case that might otherwise slip through the cracks."[32]

But the question raised a sensitive issue. Thyroid problems were well known to have surfaced among Marshall Islanders downwind from atomic tests. To imply an outbreak downwind of TMI was potentially to indict the reactor.

Industry response was thus immediate and sharp. "There cannot be any connection" between TMI and the disease, said Dr. Victor Bond of the Brookhaven National Laboratory. "I can say that unequivocally. For thyroid effects, the doses would have to have been thousands of times higher than they were." "A link just cannot be there," added William Dornsife, a nuclear engineer with the state Bureau of Radiation Safety.[33]

Tokuhata later told us that a case-by-case investigation in Lancaster County, where seven of the hypothyroid cases had surfaced, showed "no evidence" to link TMI to the disease. One case, he said, had surfaced "before the accident." Another was "inherited" and a third baby was "born three months after the accident and could not have been affected." Two others "had the thyroid in the wrong locations, which is a developmental disorder and highly unlikely to have anything to do with TMI." The final two cases were "unexplained." But having eliminated five of the seven victims, Tokuhata said the remaining two fitted the "normal pattern" of two hypothyroids per fifty-five hundred live births.[34]

By culling the cases Tokuhata brought the hypothyroid statistics down to what the state termed a "normal" rate. But Gordon MacLeod charged that even while accepting Tokuhata's subtractions, the extraordinary concentration of cases in a short period of time after the accident still reflected a five- to tenfold increase in the expected number.[35] The New York Times, however, accepted Tokuhata's analysis without question. In Jane Brody's April 15, 1980, report--entitled "Three Mile Island: No Health Impact Found"--the paper definitively exonerated the crippled reactor. In the instance of the hypothyroids a "potentially biased selection of cases led to the conclusion that radiation had damaged fetal thyroid glands."

The article dismissed the infant mortalities, saying incomplete figures and incorrect survey methods were the real culprits. Relying on statements from Tokuhata, other Pennsylvania officials, and the Atlanta-based Center for Disease Control--and without interviewing Gordon MacLeod--Brody concluded that "health officials say that thus far data do not support" claims of any extraordinary health problems among infants downwind.[36]

Unfortunately, in taking utility, state, and federal officials' word at face value, Brody had failed to notice a middle ground developing. Dr. Thomas Foley of the Pittsburgh Children's Hospital told The Washington Post that no cause-and-effect relationship had been definitely proven linking TMI to the hypothyroids. But "the fact that it did follow the accident raises an issue," he said. The timing was "peculiar and curious."[37]

The Times also overlooked the possibility that radiation from TMI might have been only partially at fault, and that emissions from normal operations at the nearby Peach Bottom reactor might also have contributed to the problem. A survey of cases in nearby Maryland might have helped clear up the issue, but none was done.[38]

Three days after Jane Brody's story appeared, the staff editorial "Nuclear Fabulists" appeared in the Times. It charged that "those scare stories about radiation damage from the accident at Three Mile Island look increasingly far fetched." Just as "reports of bizarre deformities among farm animals and wildlife" had been squelched by the Department of Agriculture, so concerns about infant mortality and hypothyroidism had been "effectively laid to rest by state and Federal health investigators." Findings that "little radiation escaped, posing virtually no threat to public health" had been "supported by all major investigations of the accident," said the Times editorial board.

In fact the real problem was "experts" who had "inflamed public fears by dealing recklessly with statistics." Among them was Dr. MacLeod, who "irresponsibly publicized some of the raw data suggesting the existence of health problems," and Dr. Sternglass, who was "accused by neutral [!] health authorities of mishandling data to demonstrate health damage. . . . Even in nuclear fables," concluded America's journal of record, "there are people who cry wolf."[39]

19. George Tokuhata, interview, February 1981.

20. Pawlick, "Silent Toll"; and MacLeod, "Politics."

21. Pawlick, "Silent Toll."

22. Brody, "3 Mile Island."

23. Pawlick, "Silent Toll."

24. Tokuhata interview. In this February 1981 interview Tokuhata denied having refused to send Sternglass the data in question. Later in the interview, however, we asked him for infant-mortality data for 1975 and 1976. He said he could not send it to us. We also asked him about Gordon MacLeod's remark to us (in our October 1980 interview) that Tokuhata had called MacLeod at 7:00.A.M. one morning in the winter of 1979-80 to discuss the hypothyroid cases that had surfaced in southeastern Pennsylvania. Tokuhata denied having made the call.

25. MacLeod interview.

26. Brody, "3 Mile Island."

27. Sternglass, "Infant Mortality Changes"; and Pawlick, "Silent Toll." See also, Sternglass, Secret Fallout, pp. 241-275.

28. Pennsylvania Department of Health, Health Department Releases Preliminary Study on Infant Deaths in TMI Area (Harrisburg: Department of Health, April 2, 1980) (hereafter cited as State Preliminary Study).

29. Brody, "3 Mile Island."

30. State Preliminary Study. This press statement also included a long discussion of infant-death rates in the January 1-September 30, 1979, period and the October 1, 1978-March 31, 1979, period, neither of which was particularly relevant to the question of what the accident did or did not do to local infant-death rates. Both sets of figures included seasonal changes and neither offered the year-to-year comparisons that were the heart of the matter. See Pawlick, "Silent Toll," for a lengthy discussion of which periods were relevant and which were not.

31. MacLeod interview.

32. Ibid., and MacLeod, "Politics."

33. Philadelphia Inquirer, February 22, 1980.

34. Tokuhata, interview, April 1980.

35. Gordon MacLeod, interview, June 1981. MacLeod explained to us that the state had averaged the two "accepted" hypothyroid cases over a full year's period, when in fact they had both occurred in a matter of months after the accident. One might also question Tokuhata's methods of culling the cases. Overall hypothyroid incidence rates are calculated on a total number of reported cases. For the state to cull the count in this single instance was to give it special treatment not given the method of calculating the "normal" rate.

36. Brody, "3 Mile Island." Among other things Brody's article incorrectly stated there had been twenty infant deaths in the ten-mile area around TMI in 1978; in fact state statistics showed fourteen for that year, and twenty for 1977.

37. Victor Cohn, "A-Plant Involved in Probe of Thyroid Ills," Washington Post, February 21, 1980.

38. The suggestion that Maryland infant-mortality rates be investigated came from Dr. Alice Stewart.

39. New York Times, "Nuclear Fabulists."

Infants Died at TMI

As debate over hypothyroids and infant deaths intensified, so did the anger and fears of the people around Three Mile Island. In March of 1980, at the first anniversary of the accident, some eleven thousand people gathered at the state house in Harrisburg to demand a permanent shutdown at TMI. The controversy largely focused on Met Ed's "other" reactor, Three Mile Island Unit One.

At the time of the 1979 accident TMI-1 had been just about to come back on line after a refueling shutdown. Ironically TMI-1 had one of the best operating records of any reactor in the U.S. The accident at TMI-2 had kept it shut. Now Metropolitan Edison, operating at the brink of bankruptcy, facing a bill of at least one billion dollars for cleaning up TMI-2, wanted desperately to get TMI-1 reopened and back in its rate base.

Increasingly fierce local opposition was in the way. In March of 1981 local citizens burned fifty thousand dollars in Met Ed electric bills on the state-house steps. The next day, on the accident's second anniversary, fifteen thousand people gathered to demand a permanent shutdown. The latter rally marked a major turning point in public attitudes toward nuclear power because it had been sponsored by eleven international trade unions--representing some seven million workers. For years organized labor had been portrayed as a staunch, unified supporter of atomic energy. But now increasing concern over health, safety, and economic issues had helped change that.

At the same time, however, the newly inaugurated Reagan administration was pushing ahead to restart atomic licensing. Through the last months of the Carter administration industry supporters had been lobbying hard to call the post-TMI moratorium to an end. The industry had learned much about safety, they said and it was time to build more plants. Led by NRC chairman Joseph Hendrie a strong drive to shorten the licensing process, and limit public participation in it, gained momentum.

But critics charged that this was precisely the time the reactors were proving more dangerous than ever. One particularly harsh fight was developing over the Diablo Canyon Nuclear Station, a double-reactor project on the coast of California that sat just three miles from a major earthquake fault. The plant had been tied up in legal battles since it was completed in 1976. Hundreds of arrests had taken place at the site, and in 1981 a national-scale confrontation erupted there.

And in the midst of increasing polarization evidence continued to surface that nuclear power was far more dangerous than anyone had believed, and that the "wolf" Gordon MacLeod and Ernest Sternglass were pointing to was in fact very real.

As early as October of 1979--six months after the TMI accident, and the same month in which Gordon MacLeod was fired--the Arkansas Department of Health issued a study indicating a sharp rise in stillbirths in Pope County, where Arkansas Power and Light's Nuclear One is located. The infant-mortality rate had dropped slightly. But the stillbirth rate had soared so significantly that the combined total had climbed from 20.3 per 1000 live births in 1974, when Nuclear One opened, to 25.4 in 1975, 27.6 in 1976, and 26.8 in 1977. The combined rate in control counties farther from the site had, by contrast, dropped sharply.

Arkansas Power and Light quickly denied any likelihood that Nuclear One "would have any effect on the health of newborns. We have worked closely with the hospital there," said AP&L vice-president, Charles Kelly, "and every indication we've had in monitoring the health effects is that there is none." The study, added Director Robert Young of the Arkansas Health Department, was "inconclusive" and offered no evidence that Nuclear One was to blame for the escalating stillbirth rate.

But Drs. George Carlo and Carol Hogue, the epidemiologists at the University of Arkansas Medical Sciences campus who prepared the study, warned that "a pattern of risk" seemed to be developing. "The situation should be monitored closely," they said, because "we may be detecting a weak signal."[40]

The signal from TMI seemed considerably stronger. In February of 1981 Pennsylvania released its 1979 vital statistics--seven months later than normally expected. Missing from the data for the first time were general disease figures from the town of Aliquippa, near Shippingport. Also omitted were county-by-county tallies of congenital malformations, and information on how many infants had been born under a weight of fifteen hundred grams. "This key information," charged Ernest Sternglass, "is needed to study the possible effects of radioactive iodine from the release of TMI or the other large reactors in the state of Pennsylvania.

"The pattern is clear," he said in an angry article in The Nation. "Two years after the TMI accident, the nuclear industry and the state of Pennsylvania continue to mislead the public about its adverse effects on human health."[41]

Just prior to the 1981 union-sponsored rally at the state capital, the Department of Health released what it termed a "final" report on infant deaths near the plant. On March 20, Dr. George Tokuhata told the media that the rate of infants dying in their first year of life had definitely not gone up after the accident. In fact, he said, in a ten-mile radius around TMI, the infant death rate per 1000 live births was an identical 19.3 in the quarter before the accident and in the quarter after it. There was "no difference" in the infant-mortality rate in January through March 1979 as opposed to April through June. If the accident had killed any area infants, "a significant increase in infant deaths during the last six months of 1979 would have occurred." Rather the death rate dropped to 12.7 in July through September of 1979 and 13.4 in October through December.

Thus, Tokuhata concluded, "there is no evidence to date that radiation from the nuclear power plant influenced the rise or fall of statistics."[42] An aide in the Department of Epidemiological Research told us that Sternglass and others had erred by relying on "provisional data," and that this latest, "definitive" report would settle all that.[43]

In general the state and nuclear industry focused their defense of TMI on attacking Ernest Sternglass. But as in the past Sternglass's primary role had been to call attention to the issue. Ultimately it was the state's own numbers that would indicate whether or not TMI had harmed local infants.

In this case, as he had done in years previous, Dr. Tokuhata had compared time frames that were essentially irrelevant to each other. In his public analysis he had emphasized that the 1979 infant-death rates near TMI in the winter months of January through March--before the accident--had been the same as in the spring months of April through June--after the accident.

But infant-death rates usually drop in the spring. The fact that they were as high in the spring of 1979 as they had been that winter was extremely significant.

Indeed, a strong case could be made that the December 1978 opening of TMI-2--not just the accident--could have been related to a significant increase in infant mortality. The 1979 winter death rates were far higher than those of 1977 and 1978.

In the winter of 1977 state figures showed that 14.7 infants had died per 1000 live births in the ten-mile TMI radius. The figure was equal to the state average for that period of that year. In 1978 the figure was 14.0--less than the state average for that period of that year. But in 1979, after the opening of TMI-2, the number had soared to 19.3--far above the state average for that period of that year, and far above the rate for that period of the previous two years in the same area. Something had caused a jump in infant deaths in TMI area the winter of 1979.

As for the spring of 1979--the three months after the accident--the contrasts were even more striking. In April through June of 1977, infants had died in the ten-mile TMI area at the rate of 11.7 per 1000 live births, less than the state average. In 1978 the figure was 9.8, again less than the state average. But in 1979 the number jumped to 19.3, far above the state average, and nearly doubling the rate of the previous two years.

	  Infant Death Rate per 1000 Live Births

	 Ten-Mile Radius Around Three Mile Island

				     Excluding     State
   Winter     Total   Harrisburg     Harrisburg    Average
    1977      14.7       24.8          10.9         14.7
    1978      14.0       30.7           7.5         14.3
    1979      19.3       33.8          12.6         13.3

    1977      11.7        8.1          13.1         14.4
    1978       9.8       11.5           9.1         14.0
    1979      19.3       29.7          14.7         14.0

    1977       9.2      10.9            8.5         12.9
    1978       4.9       3.3            5.5         11.8
    1979      12.7       9.9           13.9         12.1

    1977      14.7      16.9           14.0         13.7
    1978      15.1      25.9           11.5         13.6
    1979      13.4      31.7            5.9         14.4

Source: Pennsylvania Department of Health, "TMI Area Death Rates No Higher than State Average, Health Department Report Says," Harrisburg, March 20, 1981, Table 5.

The figures for the July through September summer months were also striking. Though fewer infants died in the summer of 1979 than in the spring of 1979, the rate was clearly higher than it had been in the summers of 1977 and 1978. The numbers were 12.7 in 1979 as opposed to 9.2 in 1977 and 4.9 in 1978. The figures then leveled off for the fall months of October through December, with infant-death rates actually lower in 1979 than they had been for the same period in 1977 and 1978.

The numbers were small in absolute terms, but of clear statistical significance in the time periods that were most crucial. Overall the pattern seemed to fit the worst-case scenario for a radioactive accident. Comparative infant-death rates in a ten-mile radius around TMI had risen when the second reactor opened there and had skyrocketed in precisely those critical first three months after the accident. The figures for Harrisburg itself--where some argued that the worst of the plant's radioactive emissions had set down--seemed even more frightening.

	  Neonatal Death Rate per 1000 Live Births

	  Ten-Mile Radius Around Three Mile Island

				    Excluding     State
  Winter     Total    Harrisburg    Harrisburg    Average
   1977      12.4        20.7          9.3         10.7
   1978       8.6        19.2          4.5          9.9
   1979      17.2        33.8          9.4          9.3

   1977       8.5        -0-          11.6         11.1
   1978       7.6         7.6          7.6         11.1
   1979      19.3        29.7         14.7         10.5

   1977       6.1         7.3          5.7         10.1
   1978       1.0        -0-           1.4          9.3
   1979       7.8         6.6          8.3          9.3

   1977      10.5        12.7          9.8         10.1
   1978      10.8        17.2          8.6         10.5
   1979       9.3        21.1          4.4         10.1

Source: Pennsylvania Department of Health, "TMI Area Death Rates No Higher than State Average, Health Department Report Says," Harrisburg, March 20, 1981, Table 4.

In 1977, in the April through June spring quarter, infants in Harrisburg died at the rate of 8.1 per 1000 live births, well below the contemporary state average. In 1978 the figure was 11.5, again well below the state average. But in 1979 it reached a horrifying 29.7, more than doubling the state average and nearly tripling the Harrisburg figures for the two previous years. Though the absolute numbers were small, the changes were of clear statistical significance.[44] Nor did the figures account for any of the pregnant women who fled the area during the accident to have their babies elsewhere, and who might well have been affected by the emissions of those terrifying first two days.

But the fiercest toll of all seemed to be taken on infants born in Harrisburg right after the accident, who then died within twenty-eight days. Infant-mortality rates--on which most of the public debate on TMI health impacts centered--are based on infants' dying in their first year of life. Neonatal death rates, a subsection of infant-mortality statistics, focus on infants who die in their first month of life. And in those tragic three months after the TMI accident every Harrisburg baby listed as an infant-mortality statistic had in fact died in the first twenty-eight days of life. Thus, in April through June of 1979, state neonatal statistics indicated that infants one month old or less died in Harrisburg at a rate of 29.7 per 1000 live births. In 1978 the number had been 7.6. In 1977 it had been zero. Though the absolute numbers were again small, the changes were of even greater statistical significance than those in the infant-mortality rate.[45]

40. Arkansas Gazette, October 31, 1979, p. 8-A.

41. Sternglass, "TMI Update"; see also, "The First Casualty at T.M.I." and "The Lethal Path of T.M.I. Fallout," Nation, February 28 and March 7, 1981.

42. Pennsylvania Department of Health, TMI Area Infant Death Rates No Higher Than Statewide Averages, Health Department Report Says (Harrisburg: Department of Health, March 20, 1981) (hereafter cited as State Final Report).

43. Ed Degan, Pennsylvania Department of Health, interview, March 1981.

44. State Final Report, Table 5.

45. Ibid., Table 4.

The Mental Fallout

In those frightening first days of the accident at Three Mile Island fifth and sixth graders in nearby Middletown had gotten together to write their last wills and testaments.[46] The terror was at the gut and instinctual levels, and it dated all the way back to Hiroshima. "The first application of nuclear energy was the atomic bombs which destroyed two major Japanese cities," explained the Kemeny Commission in 1979. "The fear of radiation has been with us ever since, and is made worse by the fact that, unlike floods or tornadoes, we can neither hear nor see nor smell radiation."

Thus, predicted the commissioners, the "major health effect of the accident" appeared to be "mental distress" felt by "certain groups" living near the reactor. The problem, they said, was "short lived."[47] But nine months later state researchers confirmed a 113 percent jump in the number of TMI neighbors using sleeping pills and an 88 percent rise in those using tranquilizers. The use of alcohol was up by 14 percent and cigarette smoking had increased by nearly a third.

As documented by one thousand telephone interviews, a wide range of "psychosomatic illnesses" had surfaced, including chronic headaches, diarrhea, loss of appetite, sweating, rashes, and hypertension.[48] "The symptoms people are suffering are similar to those suffered by people who work at dangerous jobs," we were told by Dr. Robert Holt, a New York University psychologist who studied the TMI area. "In those situations you expect an increase of tension, shortened tempers, mood swings and more physical symptoms like hyperventilations, ulcers, and asthma."[49]

In addition to finding stress they also discovered that the population had become somewhat politicized. Fierce debate raged over such issues as the venting of krypton gas from TMI-2, the dumping of more radioactive water into the Susquehanna, the reopening of TMI-1, who should pay for the cleanup of the site, and whether or not Metropolitan Edison should be allowed to go bankrupt.

Meanwhile news continued to surface of abnormal radiation levels in test wells around the plant site, and in area groundwater.[50] Such reports had an effect. "I am scared to death," Mary Enterline told The New York Times. "I have a two-year-old son and every night when I pull his shade down at bedtime, and look out the window and see the cooling towers, I nearly cry."[51]

"I live in fear every day," Donna Umholtz told the state Public Utilities Commission. "I am ready to evacuate on a second's notice."

"I won't allow my children to be exposed to low-level radiation," added Joanne Topolsky, who--like many others--was trying to sell her home and move out of the area. "We had so many dreams, and they are shattered now because of TMI."[52]

As the 1980s began, rallies, marches, and a utility-rate revolt continued to rock what had long been a quiet, staunchly conservative area. Dislike and distrust of Met Ed, the state, and the NRC continued to grow. Some public forums had degenerated into bitter shouting matches. "We will never forgive or forget what you have done here," twenty-six-year-old Michelle Stewart yelled at an NRC panel in the spring of 1980. "You have created tension between husbands and wives. You have turned us into cynical people. . . ."[53]

"My husband is a construction worker who helped build both those reactors, and now he's damned sorry for it," one local housewife told us in 1980. "No one in the world can possibly understand what we have lived through here."

We asked what, of all problems, bothered her the most. She thought a moment. "I'm tired," she then replied, "of having my children's health used in an experiment."[54]

Meanwhile some TMI neighbors questioned the fact that so much attention had been given to the mental fallout from the accident, and so little to its physical health effects. According to The New York Times, at least fourteen psychological surveys were taken of area residents in the wake of the accident, based in part on grants of $375,000 from the National Institute of Mental Health and another $52,000 from the utility industry. The state of Pennsylvania, which had mustered a bare hundred work hours to study area animals and which had established no systematic ongoing survey for possible physical damage from TMI radiation, did conduct a one-thousand-person telephone poll on the mental impact of the accident.[55] "It makes you wonder," Jane Lee told us, "how they can get so much money to study the psychological effects of this accident when they can't seem to pull it together to look at the physical effects on animal and human health."[56]

46. Mayor Robert Reid of Middletown, quoted in Robert Del Tredici, The People of Three Mile Island (San Francisco: Sierra Club Books, 1981) (hereafter cited as People).

47. Kemeny Report, p 35.

48. Pennsylvania Department of Health, Health Department Releases TMI Stress Study (Harrisburg: Department of Health, April 17, 1980) (hereafter cited as State Stress Study); and Tokuhata, interview, April 1980.

49. Robert Holt, interview, April 1980.

50. New York Times, February 21, 1981.

51. New York Times, March 27, 1980.

52. Jim Hill, "Residents Still Fearful of TMI, Survey Reports," York Daily Record, March 13, 1980, p. 1.

53. Steve Lawrence, "`We'll Never Forgive or Forget,'" Daily News (New York), March 27, 1980.

54. Anonymous, interview, March 1980.

55. State Stress Study; Ben Franklin, "Researchers Find Anxiety in the Air Near 3 Mile Island," New York Times, March 27, 1980; and Ben Franklin, "Long Distress Found Over Atom Incident," New York Times, April 18, 1980.

56. Jane Lee, interview, March 1980.

The Taste of Tragedy

For many in the TMI area the outcome of the reactor accident now seemed as obvious as it had become at Bikini Island, St. George, Utah, and other communities downwind from years of nuclear bomb testing; among the GIs who had helped clean up Hiroshima and Nagasaki; among the 300,000 who had served as guinea pigs at the tests in Nevada and the Pacific; among millions of citizens exposed to too many medical X rays; among workers in the uranium mines and mills such as Church Rock and Shiprock, and at nuclear facilities such as Hanford, Portsmouth, Paducah, Piketon, U.S. Radium Dial, American Atomics, and Rocky Flats; among citizens living downwind of Windscale, Kyshtym, American Atomics, Rocky Flats, and downriver from Church Rock, Durango, and other mill-tailing sites; among thousands of Americans living near those tailings piles, some of whom built homes with them, others of whom suffer from them in their water supply and air; among millions of Americans near low- and high-level waste dumps with reason to fear for their own and their children's long-term health; among farmers near the Shippingport, Arkansas Nuclear One, West Valley, Vermont Yankee, Rocky Flats, and Fitzpatrick and Nine Mile Point facilities with reason to believe that their animals are coming to the same ugly end as the sheep caught in the "Dirty Harry" bomb fallout of 1953; and among citizens near the Dresden, Humboldt, Indian Point, Shippingport, Millstone, Arkansas One, and seventy-odd other American reactors with reason to fear that their babies are being killed by radiation before they live even a month.

Now, in the wake of TMI, the patterns were repeating themselves in central Pennsylvania. In the fall of 1979 one York couple sued Metropolitan Edison over the August stillbirth of their daughter. A Hershey engineer named Steven Scholly saw his daughter born with Down's syndrome the summer following the accident. The state, he said, had assured him the reactor emissions could not have caused it. But, he told us, "We know radiation causes genetic defects."[57] "It's unbelievable," noted Diane McCleary of Valley Green, less than five miles from the site. "I've talked to so many people in just this area who have lost their babies, miscarried or carried them almost full term then lost them. I've lived in different places and never heard of anything like this." "It's like I was saying last night," added her neighbor Deborah Frey in an interview with Harrowsmith, "reading the death columns in the paper, I've never seen so many babies that live a day or two and then they're dead."[58] "I've been seeing a lot of strange things," Dr. Joseph Leaser told us a year after the accident. "It's nothing you can pin down, exactly. But there are symptoms surfacing here that just can't be explained by nerves alone."

A general practitioner in Middletown, Leaser is a father of four children, a part-time horse breeder and a longtime resident of the area. "We've had a real run on unusual rashes, allergic reactions, dermatitis, skin lesions, itch, and people complaining of a funny taste in their mouths," Leaser told us in 1980.

He also wondered about an uncommon aberration he had noticed among his patients. "We have found abnormal counts of eosinophils--that's a type of white blood cell--in what I would say is a significant number of patients," he told us. "It isn't a scientifically controlled study. But I'd say that when I review blood smears, it seems to me I see more." A high count of eosinophils, he added, was a "well-known symptom of excessive radiation exposure."[59]

The mysterious, scientifically unexplained "funny taste" Leaser said his patients mentioned was cited by numerous residents of both sides of the Susquehanna--as it had been by Utah residents after the 1953 "Dirty Harry" shot. "We had very bad tastes in our mouths, like an iron or metal taste," said Fran Cain, the dog breeder living within eyesight of the plant. "It came right in the house to us. We had it three or four times."[60]

"It tasted like, you know, like when you're a kid and you put money in your mouth?" said Jane Lee. "And we all had it."[61]

"It was like having a penny in your mouth," said Bill Whittock, a retired engineer in his seventies living a quarter mile from the plant. "I'd be curious to know what degree of radioactivity I was exposed to."[62]

There were some other symptoms as well. "I got kind of scorched the first day," said Vance Fisher, a local farmer. "I didn't know what was going on and I had outdoor work to do, so I was out most of the day. Got a little burn out of it. Well, it wasn't a sunburn, anyhow. My face got red."[63]

In the midst of the accident Celeste Crownover of nearby Londonderry began suffering an unexplained twitch in her leg, which--as of our June 1980 interview--had not entirely gone away. During the worst of the radioactive emissions tears began to gush from her eyes, she suffered a bad metallic taste in her mouth, a burning sensation covered her arms and legs, and a "fiery blister" broke out on her shoulder. In the summer her hair fell out "by the hands full." "I am fifty-one years old," she told us, "and nothing like that has ever happened to me before or since."[64]

"My daughter got real sick," Becky Mease of Middletown told an NRC panel. "She had diarrhea for three days straight and headaches and she became anemic. I didn't know what to do. My little girl is still getting colds and sinus problems," Mease added. "Now if that's not because of that power plant, you tell me what it is."[65]

"I haven't felt myself since about three months after the accident," we were told by Louise Hardison, whose goats--at 1.2 miles from the reactor--produced milk with high radiation counts. "I've been tired all the time. Maybe it's all in my head. But maybe it isn't."[66]

Hardison's complaints of tiredness came nearly two years after the accident, and were echoed by others in the TMI area, as well as by people living near Vermont Yankee and Rocky Flats.[67]

Dr. Joseph Leaser told us in early 1981 that "since about six months after the meltdown, I've noticed that the problem with the white blood cells has disappeared and hasn't come back."

He added that during the fall 1980 venting of krypton gas from the TMI-2 core, "a number of patients who didn't know it was happening came in independently complaining about the funny taste in their mouths. I hadn't heard of that since the accident, and I haven't heard any of it again since the venting."

Leaser said he thought it all indicated something very frightening and dangerous. "These are a rock-ribbed, churchgoing, Bible-thumping lot of people. They are not the kind who will go off running at the mouth. Some of these symptoms you can explain away from psychological stress, there's no doubt about it.

"But some you just can't."[68]

57. Steven Scholly, interview, March 1980.

58. Pawlick, "Silent Toll,"

59. Joseph Leaser, interview, March 1980.

60. Del Tredici, People.

61. Ibid.

62. Bill Whittock, interview, March 1981; and Del Tredici, People.

63. Del Tredici, People.

64. Paxton Herald, March 26, 1980; and Celeste Crownover, interview, June 1980.

65. Lawrence, "We'll Never Forgive."

66. Louise Hardison, interview, February 1981.

67. Green, "Vermont Yankee"; and Mixon, interview, May 1981.

68. Leaser, interview, February 1981.

Next | ToC | Prev

back to Killing Our Own | radiation | rat haus | Index | Search