LOMBORG-ERRORS                       Home
                        List of chapters
Chapter 15:

Air pollution



P. 164 right: (COMMENT)

"The British environmental scientist Peter Brimblecombe has produced a model . . ". Comment: Brimblecombe did not produce a model, but used a model already described by F. B. Smith in 1976. This model has been applied to cities of varying sizes at the present time, and it turns out that it gives values of air pollution which lie within about 20 % of the mean values measured at stations in those cities. Therefore, it is not acceptable that Lomborg "adjusts" the values produced by this model by mulitplying them with a factor of 4 to 5 (see remarks to note 1163).

P. 164 right: ERROR

" . . such that the levels of the 1980s-1990s are below the levels of the late sixteenth century . . " Error: This is not true. Lomborg has manipulated Brimblecombe´s data by multiplying them with a factor of 4 to 5. Actually, Brimblecombe´s data are that the smoke concentration in the late sixteenth century was about 10 µg/m³, whereas recent levels (around 1990) seem to be slightly higher than this.Therefore, Lomborg´s conclusion that smoke pollution has been much worse in previous centuries than in the 20th century, is not warranted.

P. 164 bottom right: ERROR

About sulphur dioxide pollution: " . . such that again the levels of the 1980s-1990s are below the levels of the late sixteenth century." Error: This is not true. Lomborg has manipulated Brimblecombe´s data by multiplying them with a factor of 4 to 5. Actually, brimblecombe´s data are that SO2 concentrations in the late sixteenth century were 10 - 20 µg/m³, which is certainly lower than the values found in the 1980s, which were at about 50 µg/m³. Therefore, Lomborg´s conclusion, that air pollution is an old problem, is not warranted.

P. 165, fig. 86 and note 1163: ERROR

"Data for 1585-1935 are estimated from coal imports and have been adjusted to the average of the measured data." In the note: "This necessitates an upwards adjustment of Brimblecombe´s data by approximately a factor of 4." Error: Actually, the directly measured SO2 concentrations during the period 1934 to 1964 were fairly constant around an average of abut 330 µg/m³, probably with a somewhat higher average around 400 µg/m³ in the first years. There also exist data based on another method, in which the concentration was not measured directly, but inferred from the amount deposited. A conversion factor has been applied to transform these latter data into concentration data, but the validity of this conversion is uncertain. Data are presented in fig. 1 in Laxen & Thompson (1987): Environmental pollution 43: 103-114 (see also example 3 here in Lomborg-errors). From their figure it is evident that the deposition method gives higher values than direct measurement. What Lomborg has done, is to use only the measurement that gives high values around 1935, and only the measurement that gives low values around 1960. In this way, the curve artificially is made steeper. Next, this steeper slope has consequences for the calibration with the coal import data, and this distorted calibration then leads to that Lomborg mulitplies Brimblecombe´s curve with a factor of 4. Without this erroneous multiplication, there would not have been higher values from 1700 to 1900 than in the middle of the 1900s, and the whole point in Lomborg´s argumentation would collapse. Also, it is not true that "it must be assumed that central London was much more polluted". The SO2 data are from the most polluted part of central London. Altogether, the figure is a case of deliberate manipulation.

P. 169 right: BIAS

"The reason for the dramatic fall in particle levels is partly . . " Flaw: When mentioning causes of the fall in the level of particle pollution, Lomborg fails to mention the role played by regulations. He mentions technical advances which have brought about reductions, but avoids the mention of all evidence for the effect of public regulation. In the USA, Chicago and Cincinnnati began a nationwide trend of municipal air quality control by enacting laws to regulate smoke and soot already in the year 1881. Thus, long before the Clean Air Act of 1970, improvements in air quality were "hastened by state and local ordinances that regulated the incineration of garbage and the burning of coal or high-sulfur fuel in residential, commercial, and industrial furnaces" (cited via D.A. Kysar from Paul R. Portney, pp. 77-99 in Portney & Stavens (eds. 2000): Public Policies for Environmental Protection). So, the downward trend before 1970 in figure 88 could very well be partly or mainly due to public regulation.

P. 170 left and note 1212: BIAS

"Many studies have - perhaps surprisingly - not been able to document any noteworthy effect." Flaw: The cited statement is based on Powell 1997. However, on his p. 19, Powell concludes: "Overall, the results of this analysis suggest that mandated pollution control investments have often had a significant effect in reducing maximum air pollutant concentrations." So, Lomborg´s citation of Powell is so biased that the conclusions are turned to nearly the opposite.This is deliberately misleading.


After having stated that many studies have not been able to document any noteworthy effect of regulating air pollution, Lomborg makes no mention of that information which does document "noteworthy effects". Error: We know that Lomborg has read and studied the EPA report to the US Congress on benefits and costs of the Clean Air Act. This is evident from the end of his note 1186. This report deals with the American Clean Air Act of 1970 and calculates what the air pollution in US would have been in 1990 if the act had not been implemented. Concerning the difference between the 1990 scenarios with and withour regulation of air pollution, the summary of the report says: ". . the differences between the scenarios were so great that, under the socalled "no-control" case, an additional 205,000 Americans would have died prematurely [of these: 184,000 due to particulate matter] and millions more would have suffered illnesses ranging from mild respiratory symptoms to heart disease, chronic bronchitis, asthma attacks, and other severe respiratory problems. . . Other benefits which could be quantified and expressed in dollar terms included visibility improvements, improvements in yields of some agricultural crops, improved worker attendance and productivity, and reduced household soiling damage." This is the conclusion of the latest report (available at www.epa.gov/oar/sect812/), but the 1997 report which Lomborg saw, has similar conclusions. Thus, we certainly are dealing with "noteworthy effects", and this is true for the situation in the USA in general. When Lomborg knows this very well, but does not mention it, this omission is so serious that it is fair to call it an error made in order to mislead deliberately.

P. 170 left: FLAW

"It seems likely that in the absence of the Clean Air Act of 1956 substantial improvements in air quality would have occurred anyway." - After the smog catastrophe in 1952 with an estimated 4,000 premature deaths, great public concern led to the Clean Air Act of 1956, amended in 1964 and 1968, and subsequently, the situation improved. Lomborg cites a paper which documents that the improvement was slow, had started before 1956, and continued long after 1956. We get an impression that the improvement was little better than if people in London had voluntarily reduced combustion of raw coal because it was common sense to do so, rather than because they had been forced to it by law. Flaw: The point is that the change in heating systems was indeed, to a great extent, voluntary. The authors of the paper cited by Lomborg were not British residents and probably did not understand that in British culture, you rely very much on voluntary agreement. The Clean Air Act was not a compulsory regulation (Wall (1976): International journal of environmental studies 9:239-245). Each district council in London was free to decide if they would engage in smoke control, and as late as 1966, some councils had not yet done so. So, it is natural that improvement was only gradual. You cannot use this example in a debate on compulsory regulations, because the Clean Air Act was not a compulsory regulation in the usual sense of the word. Concerning SO2, Laxen and Thompson (cited by Lomborg) have shown that regulations of the S content of oil in central London from 1972 actually fits nicely with the downward trends in SO2 concentration after that. Lomborg must have read this. When he fails to mention it, his text is at least partially biased.

P. 170 left and note 1214: BIAS

"In a study of three US cities, it was found . . " Flaw: This refers to Powell 1997. As described above for note 1212, Lomborg´s review of what Powell says is biased.

P. 170 bottom right: ERROR

"But only within the last 20 years have we become clearly aware . . " Error: This is not true. Right from the first introduction of tetra ethyl lead additive to petrol in 1923, there has been awareness of the dangers of this compound. The recommendations of an expert committee in 1926 led to a limit to the lead content of petrol. This limit was raised in 1958, which aroused much concern about lead. Patterson (1965: Arch. env. health 11: 344-360), described that the general pollution with lead was not far from the level that is poisonous to man.

P. 171 left: FLAW

"The US started phasing out lead in gasoline in 1973 . . " Flaw: Lomborg ought to have mentioned that the motivation for this phasing out was mainly the worry about possible effects on humans. The producers of car engines and lead additives opposed the phasing out of lead very vigorously. That is, the reduction was not a part of general technical advances, but were the result of public concern. Admittedly, the effect of lead on catalytic converters also played a certain role, but the converters too were introduced because of concerns for public health. Considering that Lomborg repeatedly tries to convince us that the improvements we see would have come anyway, he should have mentioned that at least in the case of lead, these improvements would not have come without public intervention. Instead, he falsely postulates that the concern about the health effects of small amounts of lead appeared only around 1980, i.e. after the phasing out of lead had started.

P. 173 right: ERROR

"Ozone is not believed to have any actual life-threatening effect." Error: This statement is very much wrong. Ozone is an unstable and very reactive compound, which, therefore, is acutely toxic to plants and animals. In humans, exposure for one hour to more than 1 ppm ozone ( = 2000 µg/m³) causes lung damage. Lomborg has read the PORG report from 1997. The executive summary contains sentences like "Concentrations of ozone exceed thresholds for effects on vegetation and human health throughout the UK", and "A recent study has demonstrated that ozone levels in London are associated with changes in daily mortality rates." As Lomborg writes directly contrary to what he has read, this error seems to be deliberate. In his note 1253, Lomborg refers to that there is little evidence for long-term chronic effects on most of the population. This - that most of the population, i.e. more than 50 %, are probably not affected, is not the same as that there is not any effect. It may be added that several recent studies have demonstrated that children exposed to ambient levels of ozone may suffer chronic damage to their lung function (see J. Sunyer (2004): Eur. Respir. J. 23: 185-186).

P. 173 right: FLAW

"The UK expert panel on Air Quality Standards "found no evidence . . ." Flaw: This statement is outdated and wrong. The expert panel report from 1994 is no longer available at the internet, so it has not been possible to check if it has been quoted correctly by Lomborg. But Lomborg has read the PORG report from 1997, which refers to a 1996 study demonstrating a relation between ozone and public health (H.R.Anderson et al. (1996): Air pollution and daily mortality in London: 1987-92. Brit. Med. J. 312: 665-669). This paper, in turn, refers to evidence that was available already in 1991 and 1993 about the relation between ozone and lung diseases in London. One of these is an official health advisory group. It seems strange that the expert panel on air quality should not know about this. Altogether, Lomborg´s choice to use just this quote seems very biased and deliberately misleading.

P. 173 right: (COMMENT)

"Ozone pollution is generally measured in peak concentrations . . " Comment: The British authorities use another measure, viz. the number of days that the concentration exceeds a concentration of 100 µg/m³, because this measure is believed to relate to the health risks. Annual averages are also reported.

P. 173 right, note 1257 : ERROR IN REFERENCE

Error: Note 1257 is obviously irrelevant to the contents of the text at this place, and it is seen to be an erroneous duplication of note 1260. Probably, the contents of the note should have been those that are now in note 1258.

P. 173 right: ERROR

"Ozone levels have not been consistently monitored at the national level in the UK." Error: Actually, monitoring in rural aeas started in the UK in 1986/87, and monitoring in urban locations started in 1990.

P. 173 bottom right and note 1258: FLAW

"In the 1997 UK ozone revies, it was concluded . ." Flaw: It seems that note 1258 should have been note 1257. The contents of note 1258 should probably have been the following quote from the same paragraph of the same reference: "The magnitude of peak concentrations has declined . . . This is a very important improvement in air quality . . ". However, in spite of this text which Lomborg most likely had intended to quote, the standard aimed at by the British authorities is the number of days per year that exceed a certain threshold. The trend in this measure may be seen at www.defra.gov.uk/environment/statistics/airqual/kf/aqkf28.htm.Here it is evident that there is no improvement at rural sites over the period 1987 to 2003. The trend for urban sites is steadily worsening during the whole period. So, Lomborg avoids using the same measure that the authorities use, and instead uses another measure which happens to give the false impression of steady improvement.

P. 173 bottom right, figure 93 and note 1260: FLAW

". . as figure 93 indicates, London peak data show a fall of 56 percent from 1976 till today." Flaw: It has not been possible to trace the source for Lomborg´s data concerning London. It seems suspect that the curve has been made by the combination of two data-sets. The available data are that the ozone concentrations at the most polluted urban sites in UK are rising, and so the impression given by Lomborg is misleading. It should be added that the highest ozone concentrations are found in the countryside, not in the cities, because NO2 from traffic in the cities reacts with ozone and removes it from the air. Lomborg is aware of this, as is evident from the Danish version of his book.

P. 173 bottom right: FLAW

"For agriculture, it is estimated . ." Flaw: This text is greatly biased to the optimistic side, and is partially contradicted by Lomborg´s own note 1259. Actually, Lomborg´s source, the PORG report from 1997, states that across Europe, rural background ozone concentrations have doubled since the latter part of the last century, with most of this increase occurring since 1940. As to the situation in UK, the executive summary says in its introductory paragraph: "There is some evidence of a decline in the magnitude of peak concentrations during the last decade [this is the only part of the evidence that Lomborg uses!] although no general downward trends in either average concentrations or exceedance of thresholds for effects on vegetation in response to reduced NOx emission have yet been detected." In conclusion, Lomborg´s text is deliberately biased.

P. 176 left, note 1280: (COMMENT)

In note 1280: "Note, that a number of later studies have discussed and questioned the approach in Shafik . . " Comment: The kind of relationship demonstrated by Shafik for air pollution is called an environmental Kutnets curve (EKC). The main impression from Lomborg´s text is that EKCs apply generally. The closest that he comes to admit that this is not generally so, is the cited sentence.

P. 176 left, note 1280: FLAW

In note 1280: ". . but for local pollution problems like SO2 and particles, the general inverse-u shape has been reaffirmed (Grossman and Krueger 1995; Torras and Boyce 1998 . . ). Flaw: Out of the papers cited, the two referred to from 1995 and 1998 contain original analyses. But they do not unequivocally confirm that Kuznets curves apply. Grossman and Krueger (1995) have data for SO2, and for this parameter, they obtain an N-shaped curve, i.e. a curve which rises once more at incomes above $4000. The concentration of suspended particles appears to decline even at low-income levels. Torras and Boyce (1998) likewise found that the concentration of SO2 once more resumes a rising trend at high income levels, i.e. an N-shaped curve is implicated here too.

P. 176 left and figures 96-97: ERROR OF BIAS

"The World Bank set about investigating . . The result can be seen in Figure 96." Error: Shafik´s investigation encompassed ten indicators of the quality of the environment, of which some, however, were about approximately the same parameter. The ten were: Safe drinking water; lack of sanitation in cities; yearly deforestation; total deforestation up to now; amounts of waste per capita; carbon dioxide emission per capita; sulfur dioxide in city air; suspended particles in city air, oxygen content of rivers; coliform bacteria in rivers. Out of these ten indicators, only two showed the kind of bell-shaped curve that is presented in figs. 96 and 97 and which is called a Kuznets curve. It is true that two other cases with different patterns are presented in fig. 109 on p. 202, and fig. 113 on p. 206, but in the present chapter we get the impression that the results are general. Concerning deforestation, Shafik found a Kuznets type curve, but it was not significant. But in chapter 10 on forests (p. 117), Lomborg nevertheless postulates that economic growth reduces deforestation. Throughout Lomborg´s book it is a pervading theme that the environment is best improved by having maximum economic growth. Here, we see, however, that this is true only for 2 out of 10 parameters. Lomborg tries to give the impression that this - the existence of Kuznets curves - is a general phenomenon, e.g. by using "the result" in the above quote, rather than "for a minority of the parameters, the result was . . ". As this issue concerns the absolutely central theme of Lomborg´s book, this bias is grossly misleading. Those interested in general reviews on environmental Kutnets curves may consult Theodore Panayotou (2000): Economic growth and the environment, Center for International Development Working Paper no. 56, at www.cid.harvard.edu/cidwp/056.htm, and Simone Borghesi (1999): The environmental Kutnets curve: A survey of the literature, at www.feem.it/Feem/Pub/Publications/WPapers/WP1999-085.htm.

P. 177 right: (COMMENT)

"There is also good reason to believe that the developing world, following our pattern, in the long run likewise will bring down its air pollution." Comment: Maybe, but this is not borne out by the studies on environmental Kuznets curves. These are mostly cross-country studies, comparing poor and rich countries for the same time period. There exist only few studies where the trend in single countries has been studied over time, one for a few OECD countries, and one for countries in south east Asia. These studies do not reaffirm that the situation over time follows what should be expected from cross-country studies. J. R. Vincent (1997): Environment and development economics 2: 417-431, claims that the cross-country version of the Kuznets curve is just a statistical artifact and should be abandoned. Specifically, he mentions that literacy, political rights, and civil liberties are highly correlated with environmental wuality even in low-income countries. Others have concluded that "more could be learnt from examining the experiences of individual countries at varying levels of development as they develop over time" (cited form Borghesi, see above)..

P. 176 bottom right: (COMMENT)

" . . the connection between income and air pollution looks much like the pollution trend for London from 1585 that we saw in Figure 86." Comment: As note above, Figure 86 is manipulated and does not give a true picture.