






































Microsoft Word - 11keopke_baten.doc


 History of Meteorology 2 (2005) 147 
 

 

 

 

 

 

 

 

 

 

Climate and its Impact on the Biological Standard of Living  

in North-East, Centre-West and South Europe 

 during the Last 2000 Years 
 

Nikola Koepke and Joerg Baten 

 
University of Tuebingen and CESifo 

 

Abstract 

 

We argue that historical climatology is crucial for understanding human living standards, 

for which anthropometric indices are an important proxy variable, given the biological 

relationship with quality and quantity of nutrition. For example, did climatic change cause the 

demographic catastrophes of the 14
th

 and 17
th

 centuries, as Galloway (1986) argued (see e.g. also 

Kelly, n.y.)? This study uses recent estimates of human stature over the last two millennia in 

three different European regions and compares them directly with estimates of temperature. We 

employ both a climatic index based on a number of series, and a recent series by Mann and Jones 

(2004). The basic finding is that overall, the impact of temperature is economically, but not 

statistically significant. Starting in the 9
th

 century statistical significance is given, however, when 

population density exceeds a previously unknown level. It seems that population pressure made 

the European populations, especially north of the Alps, more vulnerable to climatic shocks. 

 

Climatic influence on height 

 

One of the most fascinating topics of long-run economic history is the relationship 

between climate and human living standards. Especially in pre-industrial times, we would expect 

an impact of climate on agricultural production (especially protein production), and thus on the 

quality of nutrition, and therefore mean height. Furthermore, due to more humid or colder 

winters, food storage becomes more difficult in Central Europe. Indeed, the impact on human 

history was immense: on the 18
th
 century climate-height effect see Baten (2002). Grove (2002) 

demonstrated how the switch from the medieval warm period (900 to the early 13
th

 century) to 

the Little Ice Age, starting around the late 13
th
 century, has decreased harvests and protein 

production from cattle and sheep.
1
 Not only did temperatures decline, but as colder winters 

tended to be generally correlated with more frequent weather extremes, other climatic problems 

also created a deadly synergy. For example, cattle epidemics spread rapidly in Northern and 

                                            
1
 Grain yields were falling between 1220 and 1320, see Grove (2002), figure 2. 



 Biological Standard of Living 148 
 

 

Western Europe by the 13
th

 and early 14
th

 century, killing a large part of the cattle stock. 

Therefore Grove argued that the agricultural production decline took place before and parallel to 

the Black Death of the mid-14
th
 century. 

Although plague is a highly infectious disease that is only mildly influenced by 

malnutrition, lower nutritional status might have weakened the immune system of the European 

population, contributing strongly to the large population loss of the 14
th

 century. In addition, 

during famines people often leave their households and start to move around in search of other 

subsistence possibilities (Mokyr and O‘Grada, 2002). The most Northern cattle- and fishery-

based economies of Europe suffered most. Iceland lost most of its population, and the European 

population of Greenland completely disappeared. 

The 15
th

 century and the first two thirds of the 16
th

 century were warmer again, but the 

17
th

 century represents the next climatic catastrophe. Pfister (1988) has described how climate 

reduced Swiss nutritional status in the last decades of the 16
th

 and most of the 17
th

 century. While 

most of the population decline of the 17
th
 century is traditionally attributed to the Thirty Years‘ 

War as well as to the hunger and the infectious disease that accompanied it, the rapid climatic 

deterioration could have contributed to the large number of deaths from (at least partially) 

nutrition-related diseases. The synergy between protein malnutrition and death from a large array 

of diseases can also explain why the population stagnated or even declined in countries that did 

not directly participate in the Thirty Years‘ War. Milder episodes of climatic deterioration in the 

late 18
th

 and mid-19
th

 centuries coincided with milder demographic effects on average, even if 

some regions and countries were severely affected (Grove 2002). 

These developments in individual centuries suggest that there was an impact of climate 

on nutritional conditions in some centuries of the 2
nd

 millennium. But what we want to clarify in 

this study are three questions: Was there a relationship of temperature and mean height also for 

the long run from 0 to 1800 AD? Did increasing population density lead to higher vulnerability? 

And if so, at which point in time did this happen? 

 

Climate series 

 

To check this we created a climate index using different series. Recent research offers 

estimates from Alpine and Scandinavian glacier movements, Greenland ice cores, oak tree rings 

and lake sediments to quantify climatic change over these centuries.
2
 All of those series appear to 

be correlated in general. We used the European glacier movements mainly as explanatory 

variables, because they are available for the ancient period, and the evidence might be less 

indirect for the region under study compared with, for example, Greenland oxygen isotope ratios 

(see Lamb 1982; Patzelt, 1994; Heide, 1997; Grove, 2002, p. 316). However, the literature 

emphasizes that glacier movements reflect temperature changes with a certain time lag. 

Therefore we calculated the average of the previous and the current century’s glacier movement. 

Additionally, we corroborated our glacier series with tree-ring series from North Sweden that 

also stretches back to the ancient period (and compared both with a shorter tree-ring series on the 

Alpine area: they moved in accordance, see Huntley et al., 2002, p. 278).
3
 For comparison, we 

                                            
2
 See e.g. Frenzel et al. 1997, or for an overview of possible methods see e.g. Wigley et al., 1981. 

3
 We experimented with local temperature series for the three regions North/Eastern, Central/Western and 

Southern Europe, but the differences between the series were extremely small, so we abandoned this avenue of 

temperature measurement. 



 History of Meteorology 2 (2005) 149 
 

 

used the recent temperature series estimates by Mann and Jones on the Northern Hemisphere 

(2003a, 2003b, 2004), which starts in the early 3
rd

 century AD. 

 

Height series 

 

Mean height has often been used as an indicator for the quality of nutrition.
4
 We 

estimated height trends for the Mediterranean, Northeastern and Central-Western Europe for the 

1
st
 to the 18

th
 century A.D. Because of dating limitations for a regular archaeological site, the 

unit of analysis is restricted to the century. In a related study we devoted considerable space to 

describe our strategies, here we will only summarize them.
5
 We could rely on a sample of 9477 

height estimations from 314 sites. In some cases previous investigators aggregated heights of two 

to 360 individuals; thus we have 2974 separate height numbers. We used both weighted 

regressions (weighted with square roots) and regressions with individuals only to estimate height 

trends by gender and by the three European regions. The regression approach allowed us to 

control for migration
6
 and social status

7
, as far as we (and earlier scholars) were able to 

determine this using grave goods and similar information. We arrived at trends as given in Figure 

1 and Figure 2. The overall picture shows stagnating heights indicating no real progress in 

European nutritional status until around 1800 but there is considerable variation between 

                                            
4
 This is the common procedure in anthropometric research: see e.g. Komlos, 1989, Komlos and Baten, 1998, 

Steckel, 1995. But because of the study period we got our height estimations not from written sources, but from 

physical anthropological analysis of bones from excavated cemeteries, see Koepke/Baten 2005. 
5
 Koepke and Baten, 2005. 

6
 Concerning migration:  A number of anthropologists are still convinced that genetic height potentials play a 

large role, whereas other anthropologists have doubts whether genetic height potentials explain any variation in 

average height of a population – in contrast to individual height which is clearly influenced by genetic factors 

(Bogin, 1988, Mascie-Taylor and Bogin, 1995). Anthropometric historians found that environmental circumstances 

during growth have the most important impact on variation in mean height. Two points are important in this respect. 

Firstly, most migrants experienced a different environment during their first years of life, compared with the 

autochthonous population. For example, if they were born in a Northern or Eastern European agricultural 

environment and than migrated to the Mediterranean in their later life, we would expect them to be significantly 

taller. Secondly, if immigration is large enough, agricultural production techniques might be transferred to the target 

region, if they turn out to be sufficiently efficient in the new environment. We know that the most important 

migration streams moved from the Mediterranean region into Central and Western Europe in the first to third 

century, and there were important Germanic (and other) migrations from Northern Europe to Eastern, Central and 

Southern Europe and later to the British Isles from the fourth to sixth centuries. 

Migrants from the Mediterranean to Central Europe (especially Roman soldiers and officers, as well as 

administrative staff) turned out to be 4 cm shorter than the rest of the population. But skeletons that could be 

identified as “Germanic migrants” were not significantly different from Eastern Europeans. Also not statistically 

significant, but economically meaningful was their coefficient in the “Mediterranean” regression:  Germanic 

migrants, who died in the Mediterranean region, were 1.63 cm taller. 
7
 Social status is an important variable, as many studies on the 18

th
 to 20

th
 centuries found height differences of 

typically 2-4 cm among adults of lower versus middle and upper class (see e.g. Baten, 2000). In our data set, we 

relied mostly on the classification schemes of the original studies. If skeletons were not of higher social rank, the 

excavation reports often did not find this fact worth mentioning. We therefore assigned dummy variables only to the 

cases of middle and upper class social origin (leaving a “lower or unknown” group for the constant). This also 

means that we should not over-interpret the coefficient of this social status variable. However, this variable is not 

only important by itself, but is also necessary to control for the social composition and potential social selectivity 

when we analyze height trends. Although the bulk of our measurements stems from burial sites that represented all 

social strata, we wanted to exclude the possibility of social selectivity causing height trends as far as possible. 



 Biological Standard of Living 150 
 

 

centuries.
8
 How could we make sure that this was a reliable estimate of height development? 

Naturally, this kind of estimation (for non-modern periods) has many limitations – although our 

sample is much larger than earlier studies, the number of cases is still small in comparison to 

data sets on more recent periods. But it is reassuring that counterchecking height trends for 

separate European regions and for genders moved in similar directions, except where we 

expected them to diverge. For example, we expected a worsening development for Northern and 

Eastern Europe during the Little Ice Age (14-17
th

 centuries) because of the more extreme impact 

of the temperature change there, and the Northeastern Europeans actually lost their favorable 

position during these centuries.
9
 In contrast, the conditions were more favorable in more 

continental Central-Western Europe during this period. 

Female mean height is naturally always lower than male height. But female growth is 

also determined by discrimination of females. Female heights were even more depressed 

relatively to males during the Middle Ages than in the other epochs, whereas gender dimorphism 

decreased in the Renaissance period.
10

 This fact also fits our expectations. 

Apart from those expected deviations, height trends moved relatively similarly in the long 

run. Hence, we conclude that the estimates of development were reasonably reliable. But we 

applied an additional strategy to ascertain reliability: we checked burial sites that were in use for 

more than a century. If those shared the same trend with the large region, we could be sure that it 

was not a random regional composition effect that caused our trends. Among those cases with 

large sample numbers the majority pointed in this direction. 

The highly synchronistic development of human heights by itself might suggest an 

influence of temperature conditions, because these probably have been more similar across 

European regions than economic conditions. 

 

                                            
8
 During Roman times we have more or less stagnating heights; this is interesting as (having archaeological 

studies in mind) one would expect an increase in the 2
nd

 century, and a more pronounced decline in the 4
th

 century. 

Remarkably is the increase in the 5
th

 and 6
th

 century despite the migration period temperature pessimum. The 11
th

 

and 12
th

 century were favourable for mean height - this is also the medieval warm period, see Crowley and Lowery, 

2000. The decrease in the 13
th

 century may be explained by bad climate (beginning of the Little Ice Age). The 

decrease in the 17
th

 century could be a consequence of the thirty-year’s war and parallel climatic deterioration. 
9
 In general people in the North-Eastern region are the tallest, Mediterraneans have the lowest mean height: 

However, not due to genetics, but due to environmental factors: The Northeast has low population density and 

animal protein rich diet due to cattle husbandry, whereas in the South it is the opposite (high population pressure and 

Mediterranean diet based on less meat consumption, in this concentrated on pork): see e.g. King, 1999. In between 

are the people of Central-Western Europe, which is the region that has been under Roman occupation in the 

beginning of our study period. 
10

 See e.g. Ulrich-Bochsler, 1996. 



 History of Meteorology 2 (2005) 151 
 

 

167

168

169

170

171

172

173

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Birth Century

H
e

ig
h

t 
in

 c
m

 

 
Fig. 1. Height development, 1

st
 to 18

th
 centuries A.D. (in cm, male and female). Source: see Table 1. The level of 

heights was adjusted to male heights of an average European (using the regional coefficients and weighting 

them with sample weights). 

 

167

168

169

170

171

172

173

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Birth Century

H
e

ig
h

t 
in

 c
m

centre_west mediterr north_east

 
Fig. 2. Height development by major regions (in cm), 1

st
 to 18

th
 centuries A.D. Source: see Table 1. 

 

Comparison of temperature and height development 

 

There are some similarities and many differences between the height and the temperature 

series (Figure 3). The well-documented climatic optimum of the 11
th

/12
th

 centuries and lower 

values before and after are visible in the height series. The low values of the 7
th

 and 8
th

 centuries, 



 Biological Standard of Living 152 
 

 

and the crisis of the 17
th

 century could have been caused by adverse climatic conditions.
11

 

Important deviations relate to the 1
st
 to 6

th
 and to the 13

th
 centuries. 

In our opinion the most likely interpretation is that - despite of the phase of Migration 

Period Pessimism - after the breakdown of the Roman Empire - average height increased because 

of better nutritional status and improved environmental conditions, due to various phenomena:  

(1) population density and urbanization decreased after invasions and plague epidemics. The 

consumers moved back to the proximity of nutrient production. Infectious disease might have 

appeared less frequently (although the (supposed) 
12

 second occurrence of the plague in the sixth 

century contradicts it). (2) Germanic invaders brought their agricultural methods that emphasized 

protein production. Even if this specialization was inefficient in the Mediterranean, the settlers 

might have kept them for a transitory period. In Central and Western Europe, the methods were 

efficient as long as population density was low. 

These two developments might explain why the temperature-height relationship is not 

visible for the first eight centuries.
13

 A second possible interpretation is that after the 9
th

 century, 

population density was so high that the European population became more vulnerable to climate. 

A comparison of the development of mean height and the temperature index created by 

Mann and Jones (2004) (Figure 4) indicates more parallel movements. But also in this case, we 

cannot see a connection over all the centuries of our study period. Interestingly, this similar 

development starts with the 9
th

 century A.D., whereas it is very obvious that from the 1
st
 until the 

8
th

 century A.D. both series do not move together at all. 

 

167

168

169

170

171

172

173

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Birth Century

H
e
ig

h
t 

in
 c

m

9,10

9,15

9,20

9,25

9,30

9,35

9,40

9,45

T
e

m
p

e
ra

tu
re

 I
n

d
e

x

Height Temp.index

 
Fig, 3. Height (in cm) and temperature development (based on glacier movements and tree rings), 1st to 18th 

centuries A.D. 

 

                                            
11

 Around 1700 was (probably) the coldest phase of the Little Ice Age: see Bradley/Jones, 1993. 
12

 The so-called Antonine plague is regarded as the first one. 
13

 The low height value of the 13
th

 century is particularly interesting and deserves further study. Was it because 

of the rapid urbanization of this period (more infectious disease, less milk for rural-to-urban migrants)? Or because 

of more social or gender inequality? Does a measurement error bias the height variable? 



 History of Meteorology 2 (2005) 153 
 

 

 

167

168

169

170

171

172

173

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Birth Century

H
e

ig
h

t 
in

 c
m

-0,45

-0,4

-0,35

-0,3

-0,25

-0,2

-0,15

-0,1

-0,05

0

0,05

T
e

m
p

e
ra

tu
re

 I
n

d
e

x

mean height (in cm) temperature index (Mann/Jones)

 
Fig. 4. Height (in cm) and temperature development (index by Mann-Jones), 1st to 18th centuries A.D. 

 

Results: impact of temperature on height 

 

Using both our composite index and the Mann/Jones temperature index we come to the 

conclusion that warmer climate has a positive, but insignificant impact on mean height for the 

whole period (Table 1, col. 2-5): warmer temperature is good for harvests and protein production 

in the relevant range, and this is favorable for height. The difference between two standard 

deviations of our climatic series is 0.12, the difference between minimum and maximum is 0.20. 

The coefficient of the more appropriately specified model is 2.97. The difference between 

“good” and “bad climate” was therefore about 0.4 cm, the difference between the extremes about 

0.6 cm. Both values are at the margin of being economically significant.
14

 Without controls for 

period, this variable is economically unimportant. But the tall stature of North-Eastern Europeans 

in the warm 11
th

/12
th

 century, and its dramatic decline afterwards lends support for the 

importance of this variable. Using Mann and Jones’ climate index (see Table 2, col. 2 and 3) the 

difference between minimum and maximum is 0.22. The coefficient of 7.01 indicates that two 

standard deviations (0.10) imply 1.4 cm height difference between typical favorable and 

unfavorable periods. 

If we conduct regressions starting in the 9
th

 century A.D., the relationship between 

temperature and height becomes statistically significant (see Table 1, col. 6 and 7). Again, this 

result is robust using both temperature indices (see also Table 2, col. 4 and 5). This is interesting, 

because in the 9
th

 century, population density exceeded a previously unknown level. It seems as 

                                            
14

 Temperature is statistically insignificant as explanatory variable for mean height; without controlling for 

period it is even economically unimportant. 



 Biological Standard of Living 154 
 

 

if population pressure made the European populations, especially north of the Alps, more 

vulnerable to temperature shocks. 

 

 
Table 1. Two regressions : determinants of mean height in Europe, using our composite index for climatic 

conditions. P-Values in columns 3, 5, 7 in italics. Weighted Least Squares Regression: number of cases 

adjusted for aggregated observations using square roots. Constant refers to a hypothetical height value for 

the Early Middle Ages, and Central/Western Europe. Columns 2 - 5: 1
st
 - 18

th
 century A.D. Columns 6-7: 

9
th

 - 18
th

 century A.D.
. 
Source: see Koepke/Baten 2005. 

 

1 2 3 4 5 6 7 

CENTURIES 1st-18th  1st-18th  9th-18th  

 Coefficients p-values Coefficients p-values Coefficients p-values 

Constant 144.4 0.00          164.37 0.00           -14.03               0.00 

       

Climate warm 2.97 0.52 0.82 0.84           18.94 0.02 

Gender inequality - 0.31 0.50 -0.29 0.46 1.60 0.02 

Urban share 0.16 0.23 0.2 0.14 0.25 0.08 

Population density - 0.06 0.37 -0.08 0.20 -0.03 0.55 

Roman Bath/ 

Technology 
  -2.05 0.01   

Social inequality   -0.17 0.58   

Mediterranean - 1.66 0.05 -1.67 0.04 -3.56 0.01 

North-Eastern 

Europe 
1.17 0.03 0.89 0.07 0.92 0.17 

Antiquity - 1.68 0.01     

Late Medieval 

Period 
- 0.48 0.52   -3.00 0.02 

Modern (15
th

 to 

18
th

 c.) 
- 0.76 0.59   -0.79 0.49 

Adj. Rsq 0.33  0.38  0.60  

N 36  36  17  

 

 
 



 History of Meteorology 2 (2005) 155 
 

 
Table 2. Two regressions: Determinants of mean height in Europe, using the Mann-Jones index for climatic 

conditions. P-Values in columns 3, 5 in italics. Weighted Least Squares Regression: number of cases 

adjusted for aggregated observations using square roots. Constant refers to a hypothetical height value for 

the Early Middle Ages, and Central/Western Europe. We used the data on the Northern hemisphere 

provided by Mann/Jones to get best possible correspondence; unfortunately no data especially on Europe is 

given. Columns 2 - 3: 3
rd

  - 18
th

 century A.D. Columns 4-5: 9
th

 - 18
th

 century A.D.
 
Source: see 

Koepke/Baten 2005 and Mann/Jones 2004. 

 

 

1 2 3 4 5 

CENTURIES 3rd-18th  9th-18th  

 Coefficients     p-values    Coefficients         p-values 

Constant 175.78 0.00 172.05 0.00 

     

Climate warm 

(Mann/Jones 2004) 

7.01 0.23 15.51 0.03 

Gender inequality -0.69 0.23 0.60 0.91 

Urban share 0.22 0.12 0.34 0.03 

Population density -0.08 0.18 -0.09 0.11 

Mediterranean -2.05 0.02 -3.33 0.01 

North-Eastern Europe 1.10 0.05 1.33 0.07 

Antiquity -1.79 0.01   

Late Medieval Period -0.16 0.82 0.73 0.37 

Modern (15
th

 to 18
th

 c.) -0.48 0.75 1.40 0.34 

Adj. Rsq 0.31  0.54  

               N 32  17  

 

 

One could argue that cold climate might not be harmful but rather beneficial in the 

Mediterranean, because precipitation could become more frequent there, when temperature gets 

colder. We tested (using our index) whether our results would change if we exclude the 

observations on the Mediterranean, and found the coefficient for temperature unchanged: the 

coefficient was 2.74, compared with 2.97 when the Mediterranean was included (see Table 3). 

 



 Biological Standard of Living 156 
 

 
Table 3. Regression: Determinants of mean height only in western and northeastern Europe, using our composite 

index for climatic conditions. P-Values in column 3, in italics. Weighted Least Squares Regression: number 

of cases adjusted for aggregated observations using square roots. Constant refers to a hypothetical height 

value for the Early Middle Ages, and Central/Western Europe. Source: see Table 1. 

 

1 2 3 

CENTURIES 1
st
-18

th
  

 Coefficient p-values 

Constant 146.51 0.01 

   

Climate warm 2.74 0.60 

Gender inequality - 0.32 0.53 

Urban share 0.17 0.29 

Population density - 0.07 0.37 

North-Eastern Europe 1.15 0.06 

Antiquity - 1.76 0.02 

Late Medieval Period - 0.36 0.67 

Modern (15
th

 to 18
th

 c.) - 0.56 0.72 

Adj. Rsq 0.21  

N 30  

 

 

Except for the bundle of explanations given above for the missing relation of temperature 

and mean height in the long run, it is possible, of course, that there is measurement error, 

especially for the early period, for which the temperature estimates are known to be particularly 

imprecise.
15

 We also have to keep in mind the probability that other climatic factors, e.g. 

precipitation, are also important determinants, which we could not control for in detail, because 

no data series are available fitting with our long-run study period for now. 

Which other factors influenced the development of mean height? Apart from temperature, 

most of the other variables are statistically insignificant, but bear the expected sign in the 

regression analysis (see Table 1):  Population density comes closest to statistical significance; in 

unweighted regressions the p-value is even as low as 0.15. This suggests that lower population 

density is advantageous for the biological component of the standard of living that is reflected in 

stature in pre-industrial times, after controlling for large-region effects and inequality. The 

analysis of economic significance for population density yields a height effect of about 1.0 cm 

for the typical “high” and “low” population density of the time, and 2.2 cm between the most 

extreme observed values. In other specifications, the economic significance of population density 

would even be one third greater. Malthusian theory of land as limiting factor for human 

development seems to be confirmed (for period until 1800). Gender inequality and social 

inequality both had negative signs in the regressions for the whole period (col. 2-5). Given that 

these results are similar to those of many other studies on the 18
th

 to 20
th
 centuries, we tend to 

attribute fairly large credibility to them. In terms of economic significance, social inequality 

meant 0.63 cm between high and low, and 0.74 between extremes, whereas the effect of gender 

inequality was about half of that. The slightly astonishing result for gender inequality during the 

                                            
15

 But it is also possible that the relation is truly weak. The interpretation could be awkward, because the 

temperature series does not correspond to real conditions exactly enough. For example, by now the discussion is still 

in progress, when the Little Ice Age started: compare e.g. de Menocal et al., 2000; Jones et al. 2001. 



 History of Meteorology 2 (2005) 157 
 

 

9
th

 to 18
th
 centuries (Table 1, col. 6: positive) is not robust. When using the Mann-Jones climate 

index, gender inequality is insignificant, and the coefficient less than half the size. The “Roman 

bath/technology” dummy
16

 in the regression without time dummies actually has an unfavorable 

impact on mean height. In sum, population density is definitely of economic significance, but not 

of statistical significance. Social inequality and gender inequality are at the margin of being 

economically significant. 

 

Conclusion 

 

Our study is based on the first anthropometric estimates on the biological standard of 

living in central Europe of the two millennia A.D. In the long run no general increase in mean 

height took place. The height development is quite synchronistic in the three “large regions” of 

Europe. Population density is an economically significant determinant of mean height: it has a 

negative influence on height due to Malthusian factors. Urbanization has a positive impact, 

if we control for population density. Roman health system/technology, social and gender 

inequality, as well as temperature are of marginal significance. 

The results regarding temperature are robust using different temperature indices. 

Remarkably, we found a statistically significant relationship between temperature and height 

using data from the 9
th

 century AD onwards; probably extremely increased population density 

made the Europeans more vulnerable to climatic changes. 

 

 

                                            
16

 In order to test how Public Health developed over the last two millennia - especially, whether the picture of 

an advanced Roman bathing system and water supply technology on the one hand, and a decline of hygienic 

conditions after the end of the Roman empire on the other hand, is correct - we coded a “Roman Bath” dummy 

variable (1 for the Mediterranean for the centuries 1 to 4, and for Central/Western Europe for the centuries 2 to 4). 

As this specification might also captures other aspects of Roman technology and the imperial economic system we 

named this variable “Roman bath/technology”. In the regression without time dummies this dummy becomes 

statistically significant. However, contrary to popular belief Roman bath technology and especially other technology 

(e.g. in agricultural terms) could not compensate negative effects of the Roman economic system: the coefficient of 

this variable is negative. 



 Biological Standard of Living 158 
 

 

References 

 

Baten, J., 2000, Economic Development and the Distribution of Nutritional Resources in 

Bavaria, 1797-1839, Journal of Income Distribution 9, 89-106. 

Baten, J., 2002, Climate, Grain Production and Nutritional Status in 18th Century Southern 

Germany, Journal of European Economic History, 30.1, 9-47. 

Bogin, B., 1988, Patterns of Human growth. Cambridge University Press, Cambridge. 

Bradley, S., and P. Jones, 1993, ‘Little Ice Age’ summer temperature variations: their nature and 

relevance to recent global warming trends, The Holocene, 3,4, 367-376. 

Crowley, T., and T. Lowery, 2000, How warm was the Medieval Warm Period? Ambio, 29, 51-

54. 

deMenocal, P., et al., 2000, Coherent High- and Low-Latitude Climate Variability during the 

Holocene Warm Period, Science, 288, 5474, 2198-2202. 

Frenzel, B., et al., 1997, Glacier fluctuations during the Holocene. Stuttgart. 

Galloway, P., 1986, Long-Term Fluctuations in Climate and Population in the Pre-Industrial Era, 

Population and Development Review 12.1, 1-24. 

Grove, J., 2002, Climatic Change in Northern Europe over the Last Two Thousand Years and its 

Possible Influence on Human Activity. In: Wefer, G. et al. (eds.), Climate Development 

and history of the North Atlantic Realm, Hanse Conference on Climate History, Springer, 

Berlin/Heidelberg, 313-326. 

Heide, A., 1997, Das Wetter und Klima römischer Antike im Westen des Reiches. Unpublished 

Ph.D. thesis, Mainz. 

Huntley, B., et al., 2002, Holocene Climatic History of Northern Europe as Recorded by 

Vegetation Changes: Possible Fluctuations upon Human Activity, In: Wefer, G. et al. 

(eds.), Climate Development and history of the North Atlantic Real, Hanse Conference on 

Climate History, Springer, Berlin/Heidelberg, 223-232. 

Jones, P., et al., 2001, The Evolution of Climate Over the Last Millennium, Science, 292, 5517, 

662-667. 

Kelly, M., (no year given), Climate and Pre-Industrial Growth, Working paper on: 

http://www.ucd.ie/economic/staff/mkelly/papers/climate.pdf. 

King, A. 1999, Meat Diet in the Roman World: A Regional Inter-Site Comparison of the 

mammal bones, Journal of Roman Archaeology 12, 1999, 168-202. 

Koepke, N., and J. Baten, 2005, The Biological Standard of Living in Europe During the Last 

Two Millennia,  European Review of Economic History, 9.1, 61-95. 

Komlos, J., 1989, Nutrition and Economic Development in the Eighteenth Century Habsburg 

Monarchy: An anthropometric History, Princeton University Press, Princeton, NJ  

Komlos, J., and J. Baten, 1998, Height and the Standard of Living. Journal of Economic History, 

57.3, 866-870. 

Lamb, H. 1982, Climate, History and the Modern World, Methuen & Co,. London. 

Mann, M., and P. Jones, 2003a, Global Surface Temperatures over the Past Two Millennia, 

Geophysical  Research Letter,s 30.15, 1820. 

Mann, M., and P. Jones, 2003b, 2,000 Year Hemispheric Multi-proxy Temperature 

Reconstructions, IGBP PAGES/World Data Center for Climatology. Data Contribution 

Series 2003-015. NOAA/NGDC Paleoclimatology Program, Boulder CO, USA. 



 History of Meteorology 2 (2005) 159 
 

 

Mann, M., and P. Jones, 2004, 2,000 Year Hemispheric Multi-proxy Temperature 

Reconstructions, IGBP PAGES/World Data Center for Climatology, Data Contribution 

Series, NOAA/NGDC Paleoclimatology Program, Boulder CO, USA 

Mascie-Taylor, C., and B. Bogin, 1995, Human Variability and Plasticit,y Cambridge Studies in 

Biological Anthropology, Cambridge University Press, Cambridge. 

Mokyr, J., and C. O‘Grada, 2002, What Do People Die of During Famines? The Great Irish 

Famine in comparative perspective, European Review of Economic History 6.3, 339-363. 

Patzelt, G. 1994, Die klimatischen Verhältnisse im südlichen Mitteleuropa zur Römerzeit, In: 

Bender, H., ad Wolff, H. (eds.), Ländliche Besiedlung und Landwirtschaft in den Rhein-

Donau-Provinzen des Römischen Reichs, Espelkamp. 

Pfister, C. 1988
3
, Klimageschichte der Schweiz 1525 – 186,. Das Klima der Schweiz von 1525 – 

1860 und seine Bedeutung in der Geschichte von Bevölkerung und Landwirtschaft, Haupt 

Verlag, Bern/Stuttgart. 

Steckel, R. 1995, Stature and the Standard of Living, Journal of Economic Literature, 33,4, 

1903-1940. 

Ulrich-Bochsler, S., 1996, Anthropologische Befunde zu Frau und Kind in Mittelalter und 

Neuzeit unter besonderer Berücksichtigung der Stellung der Früh- und Neugeborenen: 

soziobiologische und soziokulturelle Aspekte im Lichte von Archäologie, Geschichte, 

Volkskunde und Medizingeschichte, Basel, Univ., Diss. Mikrofiche. 

Wigley, T., et al., 1981, Climate and History, Studies in Past Climates and their impact of Man, 

Cambridge University Press, Cambridge. 
 

 


