Emassfile - 515standard of Living Essays On Economics History and Religion in Honor of John E Murray Patrick Gray Full Chapter PDF
Emassfile - 515standard of Living Essays On Economics History and Religion in Honor of John E Murray Patrick Gray Full Chapter PDF
Emassfile - 515standard of Living Essays On Economics History and Religion in Honor of John E Murray Patrick Gray Full Chapter PDF
https://ebookmass.com/product/essays-in-economic-history-
purchasing-power-parity-standard-of-living-and-monetary-
standards-lawrence-h-officer/
https://ebookmass.com/product/the-fountain-light-studies-in-
romanticism-and-religion-essays-in-honor-of-john-l-mahoney-
robert-j-barth/
https://ebookmass.com/product/passionate-mind-essays-in-honor-of-
john-m-rist-barry-david-editor/
https://ebookmass.com/product/classical-economics-keynes-and-
money-essays-in-honour-of-carlo-panico-john-eatwell/
Convivencia and Medieval Spain: Essays in Honor of
Thomas F. Glick Mark T Abate
https://ebookmass.com/product/convivencia-and-medieval-spain-
essays-in-honor-of-thomas-f-glick-mark-t-abate/
https://ebookmass.com/product/renewing-philosophy-of-religion-
exploratory-essays-paul-draper/
https://ebookmass.com/product/the-nature-of-the-economy-
aristotelian-essays-on-the-philosophy-and-epistemology-of-
economics-ricardo-f-crespo/
https://ebookmass.com/product/medical-microbiology-9th-edition-
patrick-r-murray/
https://ebookmass.com/product/medical-microbiology-9th-edition-
patrick-r-murray-2/
Studies in Economic History
Patrick Gray
Joshua Hall
Ruth Wallis Herndon
Javier Silvestre Editors
Standard
of Living
Essays on Economics, History, and
Religion in Honor of John E. Murray
Studies in Economic History
Series Editor
Tetsuji Okazaki
Faculty of Economics
The University of Tokyo
Tokyo, Japan
Aims and Scope
This series from Springer provides a platform for works in economic history that
truly integrate economics and history. Books on a wide range of related topics are
welcomed and encouraged, including those in macro-economic history, financial
history, labor history, industrial history, agricultural history, the history of institutions
and organizations, spatial economic history, law and economic history, political
economic history, historical demography, and environmental history.
Economic history studies have greatly developed over the past several decades
through application of economics and econometrics. Particularly in recent years, a
variety of new economic theories and sophisticated econometric techniques—
including game theory, spatial economics, and generalized method of moment
(GMM)—have been introduced for the great benefit of economic historians and the
research community.
At the same time, a good economic history study should contribute more than
just an application of economics and econometrics to past data. It raises novel
research questions, proposes a new view of history, and/or provides rich
documentation. This series is intended to integrate data analysis, close examination
of archival works, and application of theoretical frameworks to offer new insights
and even provide opportunities to rethink theories.
The purview of this new Springer series is truly global, encompassing all nations
and areas of the world as well as all eras from ancient times to the present. The
editorial board, who are internationally renowned leaders among economic
historians, carefully evaluate and judge each manuscript, referring to reports from
expert reviewers. The series publishes contributions by university professors and
others well established in the academic community, as well as work deemed to be
of equivalent merit.
All books and chapters in the Studies in Economic History book series are
indexed in Scopus.
Editorial Board Members:
Loren Brandt (University of Toronto, Canada)
Myung Soo Cha (Yeungnam University, Korea)
Nicholas Crafts (University of Warwick, UK)
Claude Diebolt (University of Strasbourg, France)
Barry Eichengreen (University of California at Berkeley, USA)
Stanley Engerman (University of Rochester, USA)
Price V. Fishback (University of Arizona, USA)
Avner Greif (Stanford University, USA)
Tirthanker Roy (London School of Economics and Political Science, UK)
Osamu Saito (Hitotsubashi University, Japan)
Jochen Streb (University of Mannheim, Germany)
Nikolaus Wolf (Humboldt University, Germany)
(in alphabetical order)
Patrick Gray • Joshua Hall
Ruth Wallis Herndon • Javier Silvestre
Editors
Standard of Living
Essays on Economics, History, and Religion
in Honor of John E. Murray
Editors
Patrick Gray Joshua Hall
Religious Studies College of Business and Economics
Rhodes College West Virginia University
Memphis, TN, USA Morgantown, WV, USA
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Switzerland AG 2022
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
John Edward Murray was the Joseph R. Hyde III Professor of Political Economy
and Professor of Economics at Rhodes College in Memphis, Tennessee, when he
passed away suddenly on March 27, 2018, at the age of 58.
He was born on April 9, 1959, in Cincinnati, and became the first member of his
family to attend college. He worked at a variety of jobs to pay his tuition, including
phlebotomist, house painter, roofer, and ice cream vendor, graduating in 1981 from
Oberlin College with a degree in economics. He later added an MS in mathematics
from the University of Cincinnati, and the MA and PhD in economics from The
Ohio State University, where he wrote his dissertation under Rick Steckel.
John taught high school math before pursuing his graduate work in econom-
ics. After finishing at Ohio State, he accepted a position at the University of Toledo,
where he remained for 18 years before accepting the Hyde Professorship at Rhodes
College in 2011.
He had a lifelong penchant for learning, spending a summer studying the German
language in Schwabish Hall in 1984, and summers as an NEH scholar in Munich in
1995 and at Duke in 2013.
Murray authored two books and co-edited a third. His first book, Origins of
American Health Insurance: A History of Industrial Sickness Funds (Yale University
Press, 2007) was named one of ten “Noteworthy Books in Industrial Relations and
Labor Economics” in 2008 by the Industrial Relations Section, Princeton University.
His second book was co-edited with Ruth Wallis Herndon and titled Children Bound
to Labor: The Pauper Apprentice System in Early America (Cornell University
Press, 2009). Economic History Review said it was “a model for both comparative
and national studies” of childhood and labor in historical context. His third book,
The Charleston Orphan House: Children’s Lives in the First Public Orphanage in
America (University of Chicago Press, 2013), received the George C. Rogers, Jr.
Prize, awarded by the South Carolina Historical Society for the best book on South
Carolina history.
He published book chapters, monographs, encyclopedia and handbook contribu-
tions, and numerous articles in refereed journals including the Journal of Economic
History, Explorations in Economic History, Economic History Review, Agricultural
v
vi Preface
History, and many others. His clear, crisp writing style and ability to explain com-
plicated economic concepts made him a frequent choice to write for the popular
press as well.
John’s scholarly interests were varied, which is reflected in the essays in this
volume. His most recent work centered on coal mine safety, post bellum African-
American labor supply, and families in nineteenth-century Charleston. He published
extensively in the areas of the history of healthcare and health insurance, religion,
and family-related issues from education to orphanages, fertility, and marriage, not
to mention his work in anthropometrics, labor markets, and literacy. His intellectual
work was often informed by his religious convictions, and he spent time studying
Catholic theology at Sacred Heart Major Seminary in Detroit.
John had a deep commitment to his family. His first book was dedicated to his
wife Lynn, and his second and third books to his children Rose and Sarah. He would
share with delight information about his family with colleagues, and his office was
filled with artwork by his children and family photos.
This anthology honors John E. Murray, whose scholarly interests and collegial
network ranged well beyond the economics departments in which he worked
throughout his professional life. His sudden death in March 2018 ended many ongo-
ing conversations in economics, history, and religion. John considered himself a
historian as well as an economist, and he held himself to the scholarly standards of
both disciplines. He interpreted economic data and put it to work in the service of
history. He read history and put it to work in the service of economics. His work was
also informed by his lifelong study of religion, and he maintained lively and colle-
gial friendships with scholars of religion. The essays in this volume reflect John’s
scholarly interests and were written with his interests in mind.
John Murray was a person who conversed with others. The following chapters
continue conversations that John started, encouraged, or inspired. He read second-
ary literature voraciously and would quickly contact the author of an article or book
that caught his interest. His gift for starting conversations brought many people into
his network and led to wonderful collaborations. The four editors of this volume
met him at different moments of his professional life and in very different
circumstances.
1996: John started the conversation that brought Ruth Herndon into his scholarly
community. In 1996, when Herndon was at the Philadelphia Center for Early
American Studies at the University of Pennsylvania (now the McNeil Center), she
published a brief “Research Note” in the Journal of Social History about the signa-
ture literacy of poor people warned out of New England towns in the latter eigh-
teenth century. Literacy and poor people being two of John’s interests, he naturally
read the essay and promptly wrote Herndon at the Philadelphia Center, unaware that
since the article’s publication she had taken up a faculty position in the Department
of History at the University of Toledo, where John was himself teaching in the
Department of Economics. When Herndon received John’s letter, forwarded from
the Philadelphia Center, she picked up her office phone, and called her new UT col-
league. After John got over the shock of this serendipity, he initiated a series of
brown bag lunch conversations that gradually grew into co-authored conference
Preface vii
papers, then a co-authored journal article, then a major research grant proposal sup-
porting their co-edited anthology Children Bound to Labor. Although Herndon sub-
sequently moved to Bowling Green State University and Murray moved to Rhodes
College, they continued their conversation on childhood, parenting, education, and
labor in historical context. Shortly before he died, they had proposed a conference
session together.
2003: Josh Hall first met John when he was teaching at Capital University in
Columbus Ohio. Economic history was what first got Josh interested in economics
and he had heard that there was an Ohio economic history meeting that he might
attend. Having been born in Toledo, he figured that was enough of a connection to
reach out to John Murray by email. And so a correspondence began that touched on
baseball, the Wright Brothers, graduate school in economics, and economic history.
In 2004, John provided advice when Josh applied to doctoral programs in econom-
ics. In 2007, Josh was a finalist for a job at Rhodes College he didn’t get. However,
a year later they were searching for an endowed chair and he encouraged John to
apply. The rest, as they say, is history. Josh greatly misses John’s occasional email
exchanges and is not surprised that so many were touched so deeply by John and
his work.
2004: Javier Silvestre met John at the 2004 Cliometrics World Congress, in
Venice, where the latter chaired the session in which the former presented a paper.
Both shared a broad interest in workplace safety in different countries. Some time
after the Congress, John proposed that Silvestre coauthor a paper on safety in
European coal mining, using an almost unexploited source. However, it was not
until several years later that the real work began. The resulting paper ended up with
a strong focus on technology, to that point an almost entirely unexplored field for
both authors. Once the paper was accepted for publication, in 2014, such an amount
of information on technological change in nineteenth-century European coal min-
ing had been gathered that John proposed that he and Silvestre embark on a project
together. The premise was that, as far as technological change is concerned, perhaps
different strands of the literature, economic history in particular, had been more
focused on the eighteenth and twentieth centuries. Technology in nineteenth-century
coalmining needed to be reassessed. John’s enthusiasm was contagious. Over the
years, regular emails were exchanged on the subject of improvements in mechanical
fans, safety lamps, or explosives. He travelled to Spain a few times. In Zaragoza,
intense work sessions on the “coal project,” as John called it, were combined with
long evening walks and talks. It was difficult not to share some of his many inter-
ests: from freedom of speech to sports, via blues music, as well as dogs, of course,
to mention but a few. He was also a visiting scholar at the University of Barcelona.
His Origins of American Health Insurance book came at a time when the study of
the genesis of the Spanish welfare state was gathering strength among young eco-
nomic historians.
2011: Patrick Gray met John through mutual acquaintances in the Department of
Economics when he moved from Toledo to Memphis in 2011 to become the Joseph
R. Hyde III Professor of Political Economy at Rhodes College. Lunch conversations
regularly turned to such topics as baseball—especially John’s beloved Cincinnati
viii Preface
Reds—and raising children. John was very well read, and he wore his learning
lightly. This made him an outstanding scholar. John was not a member of the
Austrian School, but he agreed with the remark attributed to Friedrich Hayek that
“if you only understand economics, then you don’t understand economics,” and he
exemplified the spirit it expressed. His wide-ranging publications attest to a bound-
less intellectual curiosity and a punctilious attention to detail. John’s endowed chair
came with a generous book budget, and he was not afraid to use it. Theology was a
special interest. His home and office bookshelves groaned under the additional
weight of volumes related to biblical studies, church history, and philosophy.
Copious notes in the margins and underlined passages show that, far from being just
for show, he had actually read them. How to read and teach Augustine and Luther in
the interdisciplinary humanities sequence offered at Rhodes were frequent topics of
conversation. His approach to these texts bespoke an admirable humility that comes
with knowing the limits of one’s knowledge and expertise. Along with his gentle
spirit and hearty laugh, this is what his colleagues will miss.
ix
x Contents
xi
xii Contributors
Abstract Between 1899 and 1929, deaths from waterborne diseases declined dra-
matically in American cities. The major cause of such declines was spending on
sanitation systems (water, sewers, and refuse collection). Cities spent enormous
amounts to build and maintain water and sewer systems, and to collect and dispose
of refuse. We first estimate the size of the payoff to cities of such expenditures,
where the payoff is measured in averted deaths. Using a panel of annual mortality
and municipal expenditure data from 152 cities, we estimate that a 1% increase in
sanitation expenditures was associated with a 3% decline in the mortality rate. In the
second section of the paper, we ask whether the mortality reducing effects of sanita-
tion expenditures differed by the type of water resources available to the city (ocean,
lake, river). The answer is unambiguously yes, with cities located on lakes facing
the most difficult sanitary situation.
1.1 Introduction
As the nineteenth century drew to a close, the demand for public sanitation works in
American cities began to accelerate. No city can exist without a supply of freshwa-
ter, but the widespread acceptance of the germ theory made it clear that the water
should be clean – and not just clean to the senses of taste, smell, and sight, but clean
L. P. Cain
Northwestern University, Evanston, IL, USA
Loyola University Chicago, Chicago, IL, USA
e-mail: [email protected]
E. J. Rotella (*)
University of Michigan, Ann Arbor, MI, USA
e-mail: [email protected]
David Cutler and Grant Miller (2005) studied the effectiveness of urban water
supplies in the early twentieth century by using what they argue is exogenous varia-
tion in both the timing and location of the new technologies to identify the effects of
water improvements. They conclude that the causal influence of water purification
(specifically filtration and chlorination) on mortality was large. They find that clean
water was responsible for nearly half the total mortality reduction in those cities, as
well as for three-quarters of the reduction in infant mortality and nearly two-thirds
of the reduction in child mortality. In a later paper (2006), they argue that this
improvement was not limited to the largest cities. Similarly, Joseph Ferrie and
Werner Troesken (2008) estimate that 35–56% of the decrease in Chicago’s crude
death rate up to 1925 can be attributed to water purification and the eradication of
waterborne diseases. Marcella Alsan and Claudia Goldin (2019) examine the devel-
opment of clean water and effective sewerage systems in Boston between 1880 and
1920 and estimate that those works were responsible for much of the first sustained
decrease in child (under 5) mortality.
In this study, we expand the list of expenditures to include sewage works and
refuse collection as well as waterworks. If such expenditures were effective in
reducing the death rate from waterborne diseases, did they pay off by reducing the
total death rate as well? A decline in the total death rate could have resulted because
deaths from waterborne causes were a large share of total deaths, and the factors
that were responsible for the decline in waterborne deaths determined the decline in
total deaths. Secondly, improvements in water, sewers, and refuse could have led to
reductions in deaths from causes other than waterborne diseases because such dis-
eases were spread by the same vectors, or because declines in morbidity from the
causes responsible for waterborne diseases reduce the likelihood of deaths from
other causes. To paraphrase demographers, people accumulated fewer insults when
these waterborne diseases were averted, and, therefore, they were less likely to suc-
cumb to other diseases. For example, Preston and Van de Walle (1978) argue that, in
the case of intestinal diseases, public health changes led to this effect (see also
Szreter 1988; Wohl 1984; Woods 2000).
This paper examines relationships between US municipal expenditures and death
rates from 1899 to 1929. Urban historians (Glaab and Brown 1967; Mohl 1985)
consider this to be an era of reform, the first awakening of the environmental move-
ment leading to a dramatic expansion in budgetary expenditures on such works. By
1907, virtually every American city had installed sewers, and most big cities were
using filtration and chlorination to assure the safety of their water supplies (Galishoff
1980; Tarr et al. 1980).
From the very late 1890s to the very early 1930s (with a few missing years), the
federal government published compilations of both financial and mortality statistics
for cities.1 This paper stops in 1929 before the onset of the Great Depression and the
1
Data on both finances and mortality are contained in Bureau of Labor Statistics Bulletin, #24, 30,
36, and 42, for the years 1899–1902, and Census Bulletin #20 for 1902–1903. The Bureau of the
Census published Mortality Statistics of Cities annually between 1900 and 1936 and Financial
Statistics of Cities more or less annually between 1905 and 1931.
4 L. P. Cain and E. J. Rotella
availability of federal funds for municipal improvements. Its main focus is to link
statistically both the total death rate and the death rate from diarrhea, dysentery, and
typhoid (diseases spread by impure water and filth) to the expenditures on sanitation
(water supply, wastewater, and refuse works).
A second question addressed in this paper derives from Cain’s (1977) argument
that there are four distinct urban sanitary histories. The differentiating feature is the
type of water resource on which a city is located. Cities located on salt water cannot
draw their water supply from the abundant water close at hand and often have to rely
on sources hundreds of miles removed from the city. On the other hand, these cities
can dispose of their wastes in the adjacent salt water. Cities located on freshwater
lakes have historically used the lake for both water supplies and waste disposal.
Such cities are forced to geographically separate the water intake and sewer outfall
as far as possible to avoid befouling their drinking water with their wastes. This
interdependency creates what are arguably the most difficult sanitation problems
faced by any type of city. Cities located on major rivers simply have drawn their
water upstream from the city and disposed of their wastes downstream, taking care
that the potential sewage backwash cannot reach the water intake. Cities located on
smaller, minor rivers often have had to look elsewhere for an adequate water sup-
ply; they utilize distant lakes and rivers or rely on well water. Such small river cities
still dispose their wastes in the river, but they may have to build sewers to a down-
stream point where the river can receive a large volume of wastewater.
Each of the cities in our sample has been identified as belonging to one of these
groups. We will examine whether the effects of the water, sewer, and refuse vari-
ables differed by city type. Since the different city types faced different costs and
constraints in attempting to reduce mortality by investing in water, sewage, and
refuse works, we expect that the payoff to such investments varied between cities.
Therefore, cities facing different costs and constraints had incentives to invest in
sanitation strategies involving different mixes of these variables.
1.2 Data
Annual data on mortality and municipal expenditures were collected for 152 cities
for the period 1899–1929. The sample was defined to include all cities with popula-
tions over 25,000 in the 1910 Census. A few smaller freshwater lake cities were
added in order to increase the number of observations in that group. While some
cities had to be dropped from the sample because of insufficient data, the pooled
sample used in the reported regressions includes data from 87 cities in 1902 and 125
cities in 1929.
1 Urbanization, Sanitation, and Mortality in the Progressive Era, 1899–1929 5
The mortality data used in this study were collected from the Mortality Statistics of
Cities which annually published death-by-cause statistics. Data were collected on
deaths from typhoid fever, dysentery, and diarrhea as well as all causes taken
together. While historical evidence on death-by-cause is notoriously problematic
because of changing definitions of diseases and changes in diagnoses, the diseases
studied in this paper were well identified in this period.
Typhoid, dysentery, and diarrheal diseases were spread by impure water and
food, and by contact with feces and other filth. In this paper, we follow the conven-
tion of referring to this group of diseases as “waterborne,” even though water is not
the exclusive means of transmission. We expect, as did contemporaries, that these
diseases were controlled by programs to deliver clean water, remove and treat sew-
age, and collect and dispose of refuse.
Deaths from all causes and deaths from waterborne diseases were used together
with population data to calculate the total death rates (TDR) and waterborne disease
death rates (WDR) used as dependent variables in the regressions reported in
Sect. 1.4.
This study makes use of data on annual operating costs and capital acquisition costs
of waterworks, sewage works, and refuse collection and disposal systems. These
data were published in various bulletins up to 1903 and in Financial Statistics of
Cities beginning in 1905. There are few direct figures available for 1904. Not every
series was reported every year, and no Financial Statistics were published in 1913,
1914, or 1920. For 1921 and 1922, information on sewers and refuse were reported
together under the heading “Sanitation.” Interpolation based on expenditures in the
same city in adjacent years was used to apportion the 1921 and 1922 reported fig-
ures between refuse and sewers.
Financial data were used to construct two kinds of variables employed in the
regression analysis: capital variables and current operating cost variables.
Expenditures on capital were aggregated over all years up to the year of observation
and then divided by the population in the year of observation thereby producing an
estimate of the per capita value of the works. The per capita value of sewage facili-
ties (SEWKALL) and refuse collection and disposal facilities (REFKALL) were
constructed in this manner. The accumulated value of capital in waterworks
(WATKALL) includes the value of the waterworks at the beginning of the period.
This value was reported in the Census bulletins, and, for most cities, this is the value
in 1899. Galishoff (1980, 52) includes a graph based on US Public Health Service
data indicating that most cities had selected the source they used and constructed
municipal works before the turn of the twentieth century. Treatment, principally
6 L. P. Cain and E. J. Rotella
filtration, and disinfection, principally chlorination, were adopted after the turn of
the century. A small minority of cities in the sample did not have municipal water-
works and were not included in the main regressions reported in Sect. 1.4.
Information on annual operating expenditures for water, sewers, and refuse were
used to create the variables WATERAV3, SEWERAV3, and REFUSEAV3 which
are the average operating expenditures per capita for the year under observation and
the previous 2 years.
1.2.3 Control Variables
Six variables were collected for control purposes. These include each city’s land
area (LANDAREA) and assessed valuation (ASSDPC) for each year. Land area
provides a measure of geographical size and change within the study period, while
the assessed valuation measures the city’s ability to pay. Since the Progressive era
was a period of annexation and consolidation, the inclusion of these variables con-
trols for this type of city size growth.2
Two series were collected from the historical weather records to control for cli-
matological differences in time and space. The total rainfall in inches measures the
wetness of a particular year, while the length of the growing season in days mea-
sures for how much of the year climactic conditions (temperature, altitude, and
rainfall) permitted normal plant growth.3
Finally, two dummy variables were employed. The first, WAR, includes the
period of the First World War and its aftermath, which included a vigorous inflation
and a virulent outbreak of influenza. The other, LATE20, controls for 3 years in the
late 1920s when many cities overestimated their populations, the figure used to
make per capita calculations.
Annual statistics from 1899 through 1929 were pooled to create a panel data set.
The data used in the regressions cover the period 1902–1929 with data on 1899–1901
used to create variables based on averages and aggregates of past expenditures. The
effects of capital and operating costs on death rates were estimated using a one-way
fixed effects regression model. This technique runs an ordinary least squares
regression on the entire panel, estimating a separate intercept term for each city.
2
Data on Allegheny, Pennsylvania, were collected and added to those of Pittsburgh to incorporate
that annexation explicitly in the sample.
3
Unfortunately, the weather bureau did not collect information for all the cities in the sample, so in
some cases what has been included comes from a city which is a climatological clone of a sam-
ple city.
1 Urbanization, Sanitation, and Mortality in the Progressive Era, 1899–1929 7
Chi-squared tests confirm the superiority of this specification over the simple OLS
model without fixed effects.
The simple OLS model was estimated with the full set of control variables dis-
cussed above plus variables for population density (population/land area), popula-
tion growth (the population this year/the average population in the three previous
years), and the year. The weather variables proved to be very powerful with mortal-
ity substantially higher in the wetter and warmer cities. All the control variables had
the expected signs. While they were important for explaining the urban mortality
experience, all but YEAR, LANDAREA, ASSDPC, and the two dummy variables
(WAR and LATE20) were dropped from the regressions reported in the next section
because their impacts are included in the fixed effects.4
1.4 Results
The first three decades of the twentieth century were years of considerable improve-
ment in medical practice, food delivery and preparation, and urban sanitation. Also,
living standards were rising as personal income was rising throughout the period.
The widespread acceptance of the germ theory of disease led to the adoption of
procedures designed to reduce the spread of many common nineteenth-century dis-
eases. As Mokyr (1983) emphasizes, the evolutionary diffusion of public health
techniques over the twentieth century explains much of the decline in mortality
rates and the emergence of a new demographic regime. As life expectancy increased,
other diseases came to be more common causes of death. The “Second Industrial
Revolution” based on electricity and automobiles introduced potentially more
deadly technologies, while accelerated urbanization increased the potential for vio-
lence.5 The overall pattern of change in urban mortality can be seen in Table 1.1.
In 1902, the 15.124 per 10,000 population deaths attributable to waterborne dis-
eases were 8.9% of all deaths. By 1929, the 1.857 per 10,000 deaths from water-
borne diseases were only 1.4% of the total. The total death rate dropped by 20%
over the period (16.713/1000 to 13.409/1000), but death rates from waterborne dis-
ease dropped by almost 90%. We can get some idea of the importance of this rapid
decline in waterborne diseases by engaging in a simple counterfactual exercise. If,
beginning in 1902, there had been no decline in the death rate from waterborne
diseases, and if the death rate from all other causes had declined at its actual rate,
then the total death rate in 1929 would have been 14.836 instead of 13.323. That is,
instead of falling nearly 20% from 1902 to 1929, the death rate would have fallen
by only 12.2%. From this, we can conclude that 43.2% of the actual decline in the
4
A one-way random effects model was also estimated allowing for city-specific heteroscedasticity
correction using a generalized least squares technique. The results were almost identical to the
OLS specification, and, therefore, only the OLS fixed effects results are reported.
5
The proportion of total deaths from accidents, suicides, and other acts of violence does not appear
to have increased over the study period.
8 L. P. Cain and E. J. Rotella
total death rate between 1902 and 1929 can be attributed to the decline in deaths
from waterborne diseases.
The variables included in the regressions are described in Table 1.2; the results
with the total death rate as the dependent variable are reported in Table 1.3. The
variable WATKALL captures improvements in water quality that resulted from fil-
tration, the construction of filter beds and plants. In each case reported in Table 1.3,
1 Urbanization, Sanitation, and Mortality in the Progressive Era, 1899–1929 9
this coefficient is positive, quite the opposite of what would be expected if there
were a spillover to deaths from other causes from the factors one anticipates would
reduce waterborne diseases. For all the cities in the sample, and in three of the four
city types, the positive coefficient is statistically significant suggesting that an addi-
tional dollar spent on waterworks was associated with higher overall death rates,
after controlling for sewer and refuse expenditures. The variable WATERAV3 cap-
tures improvements in water quality resulting from disinfection, expenditures for
chlorination. In only one case, freshwater lake cities, is this variable negative.
Similar difficulties are present in the sewer and refuse variables.
We conclude from Table 1.3 that a consideration of expenditures on urban sanita-
tion in the years 1899–1929 produces little understanding of what determined the
total death rate. Even though the decline in waterborne diseases was a very large
part of the decline in total urban mortality in this period, the results do not show that
the total death rate responded as one might expect from the expenditures on water,
sewers, and refuse. Hereafter, consideration of the effects of urban sanitation expen-
ditures will focus on their impact on the waterborne disease death rate. These
results are reported in Table 1.4.
For all cities taken together, the existence of a waterworks and the addition of
filtration works during this period (WATKALL) had an important effect on reducing
waterborne deaths. On the other hand when we examine the impact of operating
expenditures (WATERAV3), we see that cities with higher waterborne disease death
rates either spent more on disinfection or the addition of disinfection marginally
increased deaths. Examining these results by the various city types puts the findings
into sharper relief. The negative coefficient on the water capital variable is signifi-
cant only for saltwater cities that must travel the greatest distance to find a source of
10 L. P. Cain and E. J. Rotella
fresh water. The expected negative effect is present in both types of river cities, but
the coefficient is in fact positive for freshwater lake cities. While this coefficient is
not statistically different from zero, the positive sign may be interpreted as a conse-
quence of these cities customarily drawing water from the same water source into
which they deposit their wastes. The differences in the size of the coefficients are
1 Urbanization, Sanitation, and Mortality in the Progressive Era, 1899–1929 11
also worthy of note. An additional dollar spent on water capital in saltwater cities
has ten times the impact as in major river cities, and expenditures in major river cit-
ies have three times the impact as in minor river cities.
The only city type in which annual expenditures for the water department show
a positive coefficient is minor river cities, those which rely most heavily on ground-
water supplies.6 Since this city type is the only one for which refuse capital expen-
ditures do not have a significant negative effect, where the effect is in fact positive,
one might conclude that refuse could be leeching into groundwater supplies. It is
only in freshwater lake cities where disinfection has a significant negative effect,
and it seems likely chlorination was important in combatting impurities in the
wastewater that was discharged into these cities’ water supply sources.
The two sewage variables both have a statistically significant negative effect on
the waterborne disease death rate in the regression for all cities taken together. In
each type of city, the expenditure on sewer construction and sewage treatment works
has a negative effect. It is significant for all but the freshwater lake cities, who, given
the interdependency between water supply and wastewater disposal, have to spend
a great deal more to get the same effect on death rates as the other city types. The
variable means reported in Table 1.5 do not indicate, however, that these freshwater
lake cities spent a great deal more for sewage capital than did the other city types.7
Annual expenditures on sewers have a more complex pattern across city types.
While the all city regression has a significant negative coefficient, the same is true
only for the minor river cities. Lake cities also have the expected negative coeffi-
cient, but those for saltwater cities and major river cities are in fact positive. None
of these latter three coefficients is statistically significant. In almost every case,
minor river cities use the river as their disposal source, but they may have to convey
their wastewater several miles downstream to a point where the river is sufficiently
large to handle the city’s volume. It is a simple matter of disposal strategy, which is
less of a problem for the other three types. Both saltwater cities and major river cit-
ies locate their water supply sources in such a way as to minimize the cost of waste-
water disposal. They have located their sewage works such that sewage services
have an effect on the death rate, but the annual expenditures on sewer maintenance
get short shrift relative to the need to maintain the purity of their water supplies.
The solid waste variables, which involve many fewer dollars than the other two
as the table of variable means (Table 1.5) documents, prove to have a statistically
significant negative effect in almost every city type and for all cities taken together.
Given the close connection between waterborne and foodborne diseases, the regular
removal and disposal of food wastes in such a way as to remove them from water
supply sources has important consequences. The only exception to the significant
6
Some, such as Denver, constructed systems reminiscent of saltwater cities to tap distant sources.
7
In the case of the Sanitary District of Chicago, more than 10 years of capital expense led to a large
decrease in cholera deaths in the first year after the Main Channel was opened. Unfortunately, the
District is a supra-governmental body and, therefore, not included in the Financial Statistics of
Cities. When the North Side Treatment Works opened in the mid-1920s, Chicago had spent more
on sewage treatment than the expenditures in the next ten largest cities combined.
12 L. P. Cain and E. J. Rotella
negative effect for outlays of refuse disposal capital is minor river cities, where the
potential that decomposing refuse may pollute groundwater supplies is a possible
explanation for the (statistically insignificant) positive effect observed for those cit-
ies. The only exception for annual expenditures on refuse disposal is freshwater lake
cities, which had a significant positive effect. This coefficient is a puzzle.
Inasmuch as the fixed effects model controls for variation between cities, and
since the regressions are estimated on a pooled cross-section, time-series basis,
three variables are used to control for variation across time in all the regressions.
The first such variable is YEAR, which has the expected negative coefficient and is
statistically significant in all cases. These years saw tremendous increases in medi-
cal knowledge and education, as well as important changes in food preparation with
canning, dehydration, and refrigeration producing large changes in the way the typi-
cal household confronted meal planning and preparation.
WAR is a dummy variable for the years 1917–1920 during which three major
events may have had a positive effect on waterborne death rates. The first is the
effect that wartime controls and postwar inflation might have on expenditure levels,
conceivably postponing some sanitation expenditures to postwar years when infla-
tion caused deferred expenditures to be more expensive. The second is rapid popu-
lation growth in some urban areas, much of it due to the migration of agricultural
workers from the south to urban industrial jobs in the north, and to the growth of
cities with large military installations. The crowding this created, both during and
after the war, put pressure on existing sanitation systems and may have led to
increases in waterborne death rates. In fact, there is a marked slowing in the rate of
decrease in these diseases revealed in Table 1.1, followed by an acceleration in the
early 1920s. The third effect, the influenza pandemic of 1918–1919, may have
increased the disease environment so that low levels of waterborne contamination
affected more people. The coefficients on WAR are all positive and statistically
significant for all cities other than lake cities.
1 Urbanization, Sanitation, and Mortality in the Progressive Era, 1899–1929 13
The third of the variables controlling for variation across time is LATE20, a
dummy for the period 1925–1927 during which many cities consistently overesti-
mated their population given the levels reported in the 1930 Census. Thus, the death
rates would tend to be underestimated.
Finally, ASSDPC and LANDAREA are included to control for the fact that, dur-
ing these years, many cities grew by annexation and consolidation, although the
great age of annexation was ending just as the study period begins (Cain 1983).
Since annexed areas could be either healthy or unhealthy, either well-endowed with
sanitation capital or not, the diverse pattern of coefficients should not be surprising.
Their inclusion in these regressions is attributable to the fact annexation is not con-
sidered to be a fixed effect ex ante.8
These regression results only pertain to cities with municipal water supplies. In
an attempt to assess whether the results of Tables 1.3 and 1.4 might also apply to
cities with private water supplies, two additional regressions were run. The results
are reported in Table 1.6. Because the number of cities without municipal supplies
is relatively small, it is not possible to run regressions by city type. The first regres-
sion simply reruns the sample for all cities reported in the earlier tables without the
water variables. The second regression repeats this specification for the cities with
private supplies.9 The final column of Table 1.6 reports the results of a simple
8
Regressions excluding these two variables indicate that the loss of what in the all cities case are
two statistically significant coefficients does not affect the overall results reported here.
9
It should be noted that the distribution of cities by type for these two cases is approximately equal.
It should also be noted that the cities included in the private regression are those that had no
municipal works at the start of the study period. Inasmuch as several of these cities shifted to
municipally owned works during the period, they are also included in the other regression for
those years.
14 L. P. Cain and E. J. Rotella
statistical test as to whether the coefficients of the first equation are in the confi-
dence intervals of the second equation. By this test, three of the four coefficients for
the sewage and refuse variables in the cities with municipal waterworks are in the
confidence intervals for the cities with private works. Six of the nine variables
included in the equation meet this test. This provides support for a conclusion that
the effects of sanitation expenditures on the waterborne disease death rate are simi-
lar in cities with municipal and private waterworks.
This paper seeks to answer two questions. First, was there a payoff to cities’ expen-
diture on sanitation works, and how big was that payoff? As Table 1.7 documents,
cities received a big payoff to expenditures on waterworks, sewer systems, and
refuse collection and disposal in the form of reduced deaths from waterborne dis-
eases.10 The second question is whether there were observable differences in the
four city types. As Table 1.7 illustrates, the answer to that question is yes. This study
further demonstrates that the mechanisms which do a good job of explaining the
decline in waterborne disease death rates (Table 1.4) do not perform anywhere near
as well in explaining the decline in the total death rate (Table 1.3). Indeed, the cor-
relation between the two grows smaller over time, suggesting that additional study
is needed to explain the decline in overall urban mortality in the early twentieth
century.
The total per capita expenditures that appear in the first row of Table 1.7 are the
sum of the variable mean expenditures listed in Table 1.5. The greater expenditures
for saltwater cities are attributable to the high expenditure on water capital in those
cities. The rest of Table 1.7 lists the annual decrease in the number of deaths attrib-
utable to waterborne diseases that would result from a one-percent increase in per
capita expenditures on each of the six categories. Over all the cities in the pooled
sample, a one-percent increase in each of the six categories would have saved 27
lives annually in a city of average size.
In Table 1.7, we see substantial differences between the types of cities. A one-
percent increase in expenditures on water capital in saltwater cities would have
averted almost 24 deaths, a much greater effect than elsewhere. A one-percent
increase in annual expenditures on water, interpreted as expenditures on disinfec-
tion, would have had its greatest effect in freshwater lake cities, averting over 11
deaths. Increased expenditures on sewer capital had their greatest potential impact
10
In 1902, typhoid deaths were, on average, 26% of all deaths from waterborne diseases; by 1929,
this had fallen to 16%. Thus, deaths from typhoid had fallen faster than those from intestinal/diar-
rheal diseases.
1 Urbanization, Sanitation, and Mortality in the Progressive Era, 1899–1929 15
in river cities, particularly minor river cities, where the number of deaths attribut-
able to waterborne diseases would have been reduced by an annual average of more
than 20. An additional one-percent increase on annual operating expenditures for
sewers in minor river cities would have saved an additional ten lives. The 16 reduced
deaths in freshwater lake cities that would have resulted from a one-percent increase
in expenditures on refuse capital could have been countered by 13 more deaths from
increased annual expenditures on refuse collection and disposal, which worked to
reduce mortality in the other three city types.
The differences we see in Table 1.7 are consistent with the sketch of each city
type appearing in the first section of this essay. To reiterate with the broadest brush-
strokes, the capital expenditures of saltwater cities on water supply and wastewater
works helped reduce waterborne disease deaths. While most cities adopted disinfec-
tion of their water supplies during this period, disinfection proved to have a signifi-
cant effect only in the freshwater lake cities. The intelligent location of sewage
works was important to both major and minor river cities, and the annual operating
expenditures of the sewer system were of additional importance to minor river cit-
ies. This study incorporates refuse collection and disposal as part of sanitation, and
the effects are as we expected with the exception of the positive effect of annual
operating expenditures in freshwater lake cities. Finally, and tentatively, the com-
parison of cities with municipal versus private waterworks presented in Table 1.6
suggests the analysis of cities with municipal waterworks derived from Tables 1.3,
1.4, and 1.7 applies generally.
Acknowledgments We are grateful for the research assistance of Ashish Aggarwal, Supriya
Mathew, and Stacey Tevlin and for the financial assistance of the Center for Economic History, the
Balzan Foundation, and Loyola University Chicago. We thank George Alter; John Brown; Stanley
Engerman; Joel Mokyr; Tom Weiss; participants in the International Economic History Congress
in Leuven, Belgium; and participants in seminars at the University of Illinois, Indiana University,
and Northwestern University.
16 L. P. Cain and E. J. Rotella
Louis Cain: I first met John at an academic conference, perhaps he was still in grad
school. In any event, it was about a quarter century ago when our friendship began.
We got to spend an extended time talking about our work when John came to give
a seminar at Northwestern in 2000. In the Fall of 2005, I joined Bob Fogel’s
Center for Population Economics at the University of Chicago and began lobbying
to have John come and give a seminar there. Six years later, he came. By then,
Bob’s health was failing, and we often didn’t know until an hour or so before the
seminar began whether he would appear. I knew Bob would enjoy John’s topic, but
Bob’s assistant called just before lunch to say that he was not going to make it. I
apologized for Bob’s absence, but John was just happy to have been invited. He
had booked a flight that left several hours after the seminar so we had a lot of time
to talk about the books each of us was finishing. As luck would have it, those
books were both the subject of an “author-meets-critics” session at the 2012 Social
Science History Association meetings in Vancouver. I was grateful and apprecia-
tive that one of our critics was John Murray. His comments were always to the
point but delivered with the kindness that was the hallmark of this gentle and
generous scholar.
Elyce Rotella: I got to know John at meetings of the Economic History
Association and the Social Science History Association but did not come to know
him well until after I relocated to the University of Michigan in 2007. Because John
was at the University of Toledo – less than an hour away from Ann Arbor – he came
regularly to our weekly Economic History Seminar. He was an active seminar par-
ticipant with valuable comments to offer for every paper. It was a pleasure to see
him regularly which gave me the opportunity to develop a personal relationship that
typically involved sharing stories of our musical daughters. We had many research
interests in common. One of my treasured possessions is an autographed copy of his
book on health insurance that he gave to me.
In addition to being a highly productive scholar, John was the very definition of
a good citizen. He was a stalwart of the Economic History Association, the Social
Science History Association, and the Cliometric Society – serving in leadership and
service positions for these groups and their journals. When a colleague was elected
President of the Social Science History Association, he immediately approached
John for the big job of chairing the Program Committee. John had to turn down the
offer because he had already done that job a few years earlier.
References
Alsan M, Goldin C (2019) Watersheds in child mortality: the role of effective water and sewerage
infrastructure, 1880–1920. J Polit Econ 127(2):586–638
Cain LP (1977) An economic history of urban location and sanitation. Res Econ Hist 3:337–389
1 Urbanization, Sanitation, and Mortality in the Progressive Era, 1899–1929 17
Cain LP (1983) To annex or not? A tale of two towns: Hyde Park and Evanston. Exp Econ Hist
20(1):58–72
Chadwick E (1842) Report of the sanitary condition of the labouring population of Great Britain,
reprinted in Flinn, MW (ed) (1965) University Press. Edinburgh
Condran GA, Crimmins-Gardner EM (1978) Public health measures and mortality in US cities in
the late nineteenth century. Hum Ecol 6(1):27–54
Condran GA, Crimmins-Gardner EM (1983) Mortality variation in US cities in 1900. Soc Sci Hist
7(1):31–59
Cutler D, Miller G (2005) The role of public health improvements in health advances: the twentieth-
century United States. Demography 42(1):1–22
Cutler D, Miller G (2006) Water, water everywhere: municipal finance and water supply in
American cities. In: Glaeser E, Goldin C (eds) Corruption and reform: lessons from America’s
history. University of Chicago Press, Chicago, pp 153–184
Ferrie JP, Troesken W (2008) Water and Chicago’s mortality transition, 1850–1925. Exp Econ
Hist 45(1):1–16
Galishoff S (1980) Triumph and failure: the American response to the urban water supply prob-
lem, 1860-1923. In: Melosi MY (ed) Pollution and reform in American cities, 1870–1930.
University of Texas Press, Austin, pp 35–57
Glaab CN, Brown T (1967) A history of urban America. Macmillan, New York
Meeker E (1972) The improving health of the United States, 1850-1915. Exp Econ Hist
9(3):353–373
Meeker E (1974) The social rate of return on investment in public health, 1880-1910. J Econ Hist
34(3):392–421
Mohl R (1985) The new city: urban America in the industrial age, 1860–1920. Harlan Davidson,
Arlington Heights, IL
Mokyr J (1983) Technological progress and the decline of European mortality. Am Econ Rev
83(2):324–330
Murray J (2007) Origins of American health insurance. Yale University Press, New Haven
Preston SH, Van de Walle E (1978) Urban French mortality in the nineteenth century. Pop Stud
32(2):27–297
Schultz SK (1989) Constructing urban culture. Temple University Press, Philadelphia
Shattuck L (1850, 1948) Report of the sanitary commission of Massachusetts, 1850. Harvard
University Press, Cambridge
Szreter S (1988) The importance of social intervention in Britain’s mortality, c, 1850-1944: a rein-
terpretation of the role of public health. Soc Hist Med 1(1):1–38
Tarr J, McCurley J, Yosie TF (1980) The development and impact of urban wastewater technol-
ogy: changing concepts of water quality control, 1850-1930. In: Melosi MY (ed) Pollution and
reform in American cities, 1870–1930. University of Texas Press, Austin, pp 59–62
U.S, Bureau of Labor Statistics (1899–1902) Bureau of labor statistics bulletin #24, 30, 36, 42
U.S. Bureau of the Census (1900–1929) Mortality statistics of cities
U.S. Bureau of the Census (1904) Census bulletin #20
U.S. Bureau of the Census (1905–1929) Financial statistics of cities
U.S. Bureau of the Census (1910, 1916) General statistics of cities, 1909, 1915
Wohl A (1984) Endangered lives: public health in Victorian Britain. Methuen, London
Woods R (2000) The demography of Victorian England and Wales. Cambridge University Press,
Cambridge
Chapter 2
The Continuing Puzzle of Hypertension
Among African Americans: Developmental
Origins and the Mid-century
Socioeconomic Transformation
G. T. Senney (*)
Office of the Comptroller of the Currency, Washington, DC, USA
e-mail: [email protected]
R. H. Steckel
Ohio State University, Columbus, OH, USA
e-mail: [email protected]
2.1 Introduction
Fig. 2.1 Hypertension prevalence in 2011 of African Americans 18 years and older. (Source:
BRFSS, CDC. Note: Prevalence denotes the share of individuals who are being treated for hyper-
tension or who have been told by a physician, nurse, or healthcare professional that they have
hypertension. Because some individuals have not been examined for hypertension, this measure
underestimates the true rate)
1
www.cdc.gov/dhdsp/maps/national_maps/stroke_all.htm and www.cdc.gov/diabetes/pdfs/data/
diabetesbelt.pdf.
2 The Continuing Puzzle of Hypertension Among African Americans: Developmental… 21
calories, such as recreational exercise, made limited headway in the South. With
industrialization women increasingly worked outside the home, and while benefi-
cial for income, led to unsupervised eating habits of children who consumed more
snack foods, perhaps while watching television after school, and to greater purchase
of less-nutritious prepared foods for family meals. In the context of the southern
diet, which featured starch, fat, and salt, these behavior patterns contributed to obe-
sity, which is a major risk factor for hypertension (Hall et al. 2015; Jiang et al. 2016;
Leggio et al. 2017).
Because our data and methods lead to conclusions that are only suggestive, an
important goal is to motivate the collection of intergenerational household-level
data for African Americans that could provide a rigorous evaluation of this method-
ology. Most useful would be intergenerational evidence on household income and
socioeconomic status combined with measures of lifestyle behaviors and
hypertension.
2.2 Background
The literature discusses numerous possible explanations for the disparity in hyper-
tension prevalence, including obesity, diet, quality of medical care, stress related to
socioeconomic change, poor access to health insurance, socioeconomic status, salt
retention, and substance abuse (Centers for Disease Control and Prevention 2010).
Several studies report, however, that the disparity in hypertension prevalence per-
sists even after adjustment for a wide range of socioeconomic, behavioral, and bio-
medical risk factors (Redmond et al. 2011). Despite large interventions to eliminate
hypertension disparities, evidence such as that shown in Table 2.2 indicates that
these differences have actually grown over the past few decades (Geronimus et al.
2007), suggesting that unrecognized factors are important in driving inequalities in
hypertension (Fuchs 2011).
The developmental origins hypothesis provides a mechanism for understanding
the origins of hypertension and other noncommunicable diseases found in later
adult life. Some 30 years ago, David Barker and colleagues proposed the develop-
mental origins approach (Barker and Osmond 1986; Barker 1990), after Barker
noticed that older adult deaths from heart disease in England were correlated with
infant mortality rates and birth sizes in their cohorts and geographic locations of
birth. Although there were many skeptics, discovery of the relationship stimulated a
search for possible mechanisms built around the idea that early life conditions influ-
ence adult susceptibility to cardiovascular disease (Lackland 2004). Subsequently,
medical researchers proposed and refined ideas (see, e.g., discussions in Kuzawa
and Pike (2005) and Kuzawa and Sweet (2009)), and a strong following developed
among economists, demographers, and numerous medical researchers (Lackland
et al. 1999, 2003; Hanson and Gluckman 2008; Skogen and Overland 2012; Barker
and Thornburg 2013; Steckel 2013; Lackland 2014). This type of work also engages
economic history (Fogel and Costa 1997; Bleakley 2007), environmental econom-
ics (Deschenes et al. 2009), and family decision-making (Del Bono and Ermisch
2 The Continuing Puzzle of Hypertension Among African Americans: Developmental… 23
2009; Del Bono et al. 2012). The approach also has the advantage of providing a
mechanism for understanding the interconnection with stroke and diabetes.
Mounting evidence suggests that fetuses and infants respond to poor nutrition
and stress by compromising organ integrity and degrading biological processes that
regulate physiological systems in later life (Gluckman et al. 2008; Barker and
Thornburg 2013). Evidence shows that individuals are predisposed to hypertension
if the heart, vascular tree, kidneys, and pancreas are modified in the womb in
response to maternal social stress and poor nutrition. If people rendered vulnerable
face an energy-rich diet or elevated stress in later life, pathophysiological processes
are set into motion that might significantly increase their likelihood of hypertension.
During their developmental stages, humans are able to accommodate stresses
and environmental changes by pleiotropic gene expression patterns that promote
survival. The adaptations made to the stressful environment change the structure
and function of organs before birth and might be passed on to future generations
through epigenetic mechanisms (Aiken and Ozanne 2014). When the fetus is opti-
mized for a lean world, but instead must process a lush load of net nutrition as an
adult, there is a mismatch between expectations and reality. This unexpectedly rich
nutrition actually proves harmful in certain ways to the individual in the longer term
(De Boo and Harding 2006; Swanson et al. 2009). Given this mounting evidence,
there is a need for determining the mechanisms that underlie the observation and the
generality of the finding for other noncommunicable diseases (Jasienska 2009;
Kuzawa and Sweet 2009).
Study of the Helsinki birth cohort, a longitudinal study of 13,517 men and women
who were born in Helsinki University Hospital from 1924 to 1944, shows that low
birth weight, especially when followed by obesity in early adolescence, is associated
with later life hypertension (Barker et al. 2002). Figure 2.2 shows the excess burden
of hypertension as a function of these factors. All births of low weight (<3000 g) had
elevated odds ratios of adult hypertension, but the risks were greatest (odds = 2.5) for
individuals born with weights under 3000 g and a BMI at 11 that exceeded 17.6.
Our goal is not to test or evaluate all of the suggested explanations for elevated
hypertension among African Americans, which would be a considerable task, but
rather to integrate social science research with medical knowledge to advance the
understanding of this puzzle. Our efforts are warranted by the persistence of the
puzzle and the lack of generally accepted explanations. With our data and methods,
we cannot “prove” that the mechanism of the hypothesis operated through rapidly
changing socioeconomic conditions, but we can achieve the important goal of mak-
ing a plausible case for additional study.
The Socioeconomic Transformation Figure 2.3 shows that the American South
was relatively poor for several decades following the Civil War. Regional income
per capita in New England was roughly three times that of the South (Kim and
Margo 2003). Conditions drastically improved in the middle of the twentieth cen-
24 G. T. Senney and R. H. Steckel
Fig. 2.2 Odds ratios of hypertension in adults as a function of birth weight and BMI at age 11.
(Source: Barker et al. 2002)
Fig. 2.3 Relative regional income per capita, 1840–1990 (USA = 100). Legend: ne New England,
ma Middle Atlantic, enc East North Central, wnc West North Central, sa South Atlantic, esc East
South Central, wsc West South Central, mt Mountain, pc Pacific. (Source: Kim and Margo 2003)
2 The Continuing Puzzle of Hypertension Among African Americans: Developmental… 25
tury, as regional industrial structures as well as income per capita converged dra-
matically. Southern per capita incomes grew significantly faster than the national
average (Kim and Margo 2003). This was a remarkable achievement because the
quarter century following 1950 was the strongest period for economic growth in the
twentieth century. Not only did the South gain relative to the rest of the country in
mid-century, but African Americans gained relative to whites. Between 1940 and
1980, the real incomes of white men grew 2.5 times, while that for African Americans
grew fourfold (Smith and Welch 1989). As a percentage of white male wages, those
for African Americans averaged 43.3 in 1940 and 72.6 in 1980. Opinions differ on
the sources of this progress, but schooling, civil rights legislation, and south-north
migration were all part of the mix (Heckman 1990; Donohue and Heckman 1991;
Margo 1993).
Pointing toward the importance of civil rights legislation in creating new labor
market opportunities, the median income of African American men in the South
relative to the 25th percentile of southern white men grew by 34% points between
1960 and 1990 (Card and Krueger 1993). Therefore, we argue that southern African
American adolescents of the 1960s and 1970s were particularly vulnerable to hyper-
tension as adults because their parents and older ancestors were poor, and this gen-
eration realized dramatic improvements in net nutrition beyond an age when
biological adaptation to rapidly improving circumstances was limited or impossible.
Income growth per se created vulnerabilities, but as discussed below, this variable is
also a proxy for many changes affecting the diet, work effort, and lifestyles of the
African American population, especially those living in the South. Among these
factors was desegregation of hospitals and fair housing laws, which may have had
independent, beneficial effects on hypertension.
In our analysis, the developmental origins hypothesis predicts that the children of
generationally poor parents who were born under rapidly improving conditions
would have higher rates of hypertension as adults. This paper considers how these
changes translated into greater prevalence of hypertension for African Americans
by relying on the timing of socioeconomic change and its differential impact across
states to identify forces that influence this disease. The empirical strategy has
acknowledged limitations, but if the hypothesis is powerful, one would expect to
find elevated prevalence rates in states and among ethnic groups with this dynamic
environmental history.
Earlier work using state-level data has found that long-term poverty followed by
rapid economic improvement increased the risk for type 2 diabetes at the state level
(Steckel 2013). Given the aforementioned related study and following the develop-
mental origins hypothesis, we suspect that African American families, especially in
the South, who were persistently and severely poor until undergoing significant
income growth after the middle of the twentieth century will have suffered high
rates of hypertension.
Circumstantial evidence suggests an association between income growth and
prevalence. Figures 2.4 and 2.5 provide data on the socioeconomic transformation
in the South and its relationship to the geographic prevalence of hypertension.
26 G. T. Senney and R. H. Steckel
3.0
Growth of Median Houseold Income (Relative to 1940)
2.5
2.0
Nation
[29, 37)
1.5
[37, 39)
[39, 42)
[42, 46)
1.0 [46, 54]
0.5
0.0
1940 1950 1960 1970 1980 1990 2000 2010
Census Year
Fig. 2.4 Growth of black median household income by hypertension region (population weighted).
(Sources: Census 1940, Census 1950 V2 Detailed Characteristics Table 87 and 56 for AL and HI,
Census 1960 V1 Chapter D: Detailed Characteristics Table 133, Census 1970 V1 Chapter D:
Detailed Characteristics Table 192, Census 1980 V1 Chapter D: Detailed Characteristics Table 243
and 244 for AL, Census 1990 V1 CP-2 Table 53, Census 2000 Summary File 4, Census 2010 ACS)
Figure 2.4 shows the growth of median African American household income from
1940 to 2010 ranked by the prevalence of hypertension in 2011 and organized by
clusters of states having similar levels of prevalence. The group of states having the
highest prevalence of hypertension in 2011 were also the states in which income
growth was most rapid. For example, in the cluster of states where prevalence was
in the range of 46%–54%, median household income grew the fastest, by a factor of
2.45 from 1940 to 2010. Figure 2.5 shows that the states with the highest prevalence
rates in 2011 were also the poorest in 1940. These results indicate that rapid growth
out of poverty may have triggered the rise in the prevalence of hypertension.
2.4 Controls
We recognize that many variables other than those associated with fetal origins are
linked to hypertension, and they must be recognized in the empirical analysis.
Among these are smoking, educational attainment, current income, and exercise.
2 The Continuing Puzzle of Hypertension Among African Americans: Developmental… 27
45,000
40,000
Median Household Income (2010 Dollars)
35,000
30,000 Nation
[29, 37)
[37, 39)
25,000 [39, 42)
[42, 46)
[46, 54]
20,000
15,000
10,000
1940 1950 1960 1970 1980 1990 2000 2010
Census Year
Fig. 2.5 Black median household income by hypertension region (population weighted). (Sources:
Census 1940, Census 1950 V2 Detailed Characteristics Table 87 and 56 for AL and HI, Census
1960 V1 Chapter D: Detailed Characteristics Table 133, Census 1970 V1 Chapter D: Detailed
Characteristics Table 192, Census 1980 V1 Chapter D: Detailed Characteristics Table 243 and 244
for AL, Census 1990 V1 CP-2 Table 53, Census 2000 Summary File 4, Census 2010 ACS)
2
We would consider measuring growth from an earlier period, say 1930, but unfortunately the
median household income data are unavailable prior to 1940.
2 The Continuing Puzzle of Hypertension Among African Americans: Developmental… 29
2.6 Results
The coefficients in the table of results denote the marginal effect of each indepen-
dent variable on the dependent variable, holding other variables constant. As indi-
cated earlier, the value of the variable rMedian income used in the regression
depends upon the age group in which the individual is located. To capture the effect
of changing socioeconomic circumstances on hypertension, we measure income
growth from the time period of birth to young adulthood.
It is well-known that hypertension increases with age, and for this reason, we
include age dummies as regressors that identify birth cohorts. As the age of the
30 G. T. Senney and R. H. Steckel
groups declined relative to the period of rapid growth, individuals had greater
opportunity to adapt to change, and consistent with this observation, the coefficients
on the dummy variables for age declined. Notably the coefficient for the age group
65+ (1.536) was 4.1 times greater than that for the age group 35–44. This result is
consistent with our expectation, derived from the developmental origins hypothesis,
that individuals who were younger during a time of great change had more opportu-
nity to adapt.
The results for the other coefficients are as expected. Higher median household
income in 2010 lowers the prevalence of hypertension because richer households
are better able to afford medical care. A related income measure, living in poverty,
raises it. Many studies have noted that health improves with the level of education
and in our specification people with lower education (high school or less) have a
greater prevalence. Several studies also report that the prevalence of hypertension
increases with smoking, obesity, and lack of exercise, all of which are confirmed in
Table 2.2.
It is reasonable to ask whether this number is large or small. In making this deter-
mination, we note that reported hypertension underestimates actual hypertension
prevalence, especially among minority groups. A recent study found that fewer than
50% of adults with hypertension controlled their blood pressure in 2007–2008,
which approximately doubles the impact on health of the income coefficient in
Table 2.2. The coefficient on median income in 2010 is negative and significant in
two specifications and is marginally significant in a third one. The direction of the
effect (negative) is intuitive because larger current income enables households to
better provide healthcare for their children.
The regression results suggest that intergenerational poverty followed by rapid
socioeconomic improvement elevates the risk of hypertension, which describes the
experience of African American adults born after World War II. This analysis is
consistent with developmental origins hypothesis as states with the larger income
growth, controlling for other factors, tended to have larger prevalence of hyperten-
sion. Below we offer interpretations of the variables that control for current
conditions.
Lower education suggests that the individual is less informed about the impor-
tance of regular health maintenance or less able to locate resources to assist in
obtaining healthcare. In line with established research, we find that the coefficient is
positive and statistically significant. Stress associated with poverty can cause hyper-
tension. It is well documented that potential stresses include job, financial, and fam-
ily distress (Kulkarni et al. 1998). The coefficient is positive and significant; our
result is well in line with the documented fact that low-income families tend to have
generally poorer health than wealthier families (Marmot 2002).
Medical research has shown that excess body fat is associated with higher levels
of hypertension and mortality (Faeh et al. 2011; Zheng et al. 2013). Consistent with
this pattern, obese individuals were significantly more likely to be hypertensive in
all cohorts. Similarly, exercise reduced the chances of hypertension, although the
variable is marginally significant in only one specification.
2 The Continuing Puzzle of Hypertension Among African Americans: Developmental… 31
2.7 Discussion
In recent decades, social scientists and medical researchers have studied the upward
trend in obesity rates, and collectively they have put forward several explanations.
All begin, however, with some type of energy accounting, i.e., that the growth of
calorie consumption outpaces the growth of physical activity. Among the ideas put
forward are a rise in the cost of time-intensive, home-prepared meals associated
with women working outside the home (Cutler et al. 2003; Hamrick and Okrent
2015), technological change that made work less demanding (Philipson and Posner
2003; Lakdawalla and Philipson 2009), changes in diet featuring processed foods
that replaced home-prepared meals (Devine et al. 2006), stress-induced eating cre-
ated by managing the challenges of socioeconomic change (Torres and Nowson
2007), and the proliferation of fast-food restaurants that conveniently provided calo-
ries at low cost (Chou et al. 2004; Schlosser 2012). Many of these arguments apply
to African Americans, especially those who lived in the South.
Table 2.1 shows the trend in obesity by race from 1959–1962 to 2015–2016. In
all years, obesity rates of blacks exceeded those of whites, and on average were 39%
higher. Here we discuss pathways by which the transformation of African American
socioeconomic conditions, particularly in the South, ultimately promoted obesity,
which in turn contributed to hypertension in a vulnerable population by reducing the
physical activity of daily life in an environment of a rich diet and little recreational
exercise. The major components of our analysis are the mechanization of agricul-
ture, lack of recreational exercise, the spread of automobiles, women’s employment
outside the home, and the continuation of a rich diet.
The Mechanization of Agriculture Based on the 1950 Census, agriculture was
the dominant industry in the South (U.S. Bureau of the Census 1952). In the swath
of states that extend from Texas and Oklahoma to North Carolina, the average share
of African Americans employed in agriculture was 31.8%. It exceeded 35% in North
Carolina, South Carolina, Arkansas, and Georgia. By 1980, however, the average
share in the 11 states had declined to 3.4% and slightly exceeded 5% in only three
states—Florida, Arkansas, and Mississippi (U.S. Bureau of the Census 1983).
Relief from field labor came late to the South relative to other regions (Hurt
1989). Mechanization of the harvest was difficult to accomplish in the region’s most
important crops of cotton and tobacco. Even today, the latter requires extensive
hand labor and thus mechanization contributed little to productivity in this crop.
Therefore, we focus on the predominant crop, cotton.
A 1939 study of man-hours per acre in cotton production in the Mississippi Delta
showed that that vast majority of time (62.9%) was devoted to the harvest, while
cultivating, thinning, and weeding the crop absorbed an additional 30.9% (Holley
2000, p.134). Picking cotton by traditional methods required long hours of stoop
labor, and unlike grain for which mechanical harvesters had existed for over a cen-
tury, cotton harvesting faced two challenges: irregularly spaced bolls and bolls that
ripened at different times on the same plant (Holley 2000). Development of new
32 G. T. Senney and R. H. Steckel
Table 2.2 Explaining the prevalence of hypertension across individuals within age cohorts with
30-year income gap after birth
Fig. 2.6 Diffusion of the cotton picker: percent of the crop picked by machine within regions.
(Source: US Department of Agriculture (1974), Table 185)
varieties that ripened bolls at about the same time solved the latter, but it took some
engineering to build a machine that was effective at removing bolls with little
destruction of fibers while also eliminating plant debris.
The diffusion of the cotton-picking machine during the 1950s and 1960s nearly
eliminated hand labor in picking by the early 1970s. Figure 2.6 demonstrates the
extent of change. In 1950, approximately 5% of the crop was picked by machine
and by 1970 the figure had risen to virtually 100%. Although mechanical cotton
pickers largely replaced hand labor between the late 1940s and the 1960s, hand
methods persisted on small farms for a decade or more (Heinicke and Grove 2008;
Logan 2015).
The diffusion of the tractor was a second important change that eased the burden
of physical labor in the South (Fig. 2.7). Relative to other regions, farmers were
slow to adopt this machine, and mules lingered on small farms operated by older
farmers until the 1960s (Ellenberg 2007). Southern customs were fashioned by a
long history of physical labor in the fields that welcomed rest at the end of the work-
day and discouraged work on Sunday. These habits persisted after diffusion of the
tractor and the mechanical cotton harvester, thereby adding to weight gain in an
environment where people maintained a rich diet and eschewed recreational exer-
cise (Church et al. 2011)
The South was not a region where habits of recreational exercise and health club
memberships readily replaced a decline in caloric expenditure associated with a
Another random document with
no related content on Scribd:
“Harrow Wanderers v. Gentlemen of Cheshire.—T.
S. M. Trentham, not out, 205.”
“We have it from a reliable source,” says the Athletic
News, “that the authorities at Old Trafford are making
strenuous efforts to induce Mr. T. S. M. Trentham, this
year’s captain at Harrow, and the youngest member of the
famous brotherhood whose name he bears, to qualify for
Lancashire. As doubtless our readers are aware, the
authorities at Old Trafford have always been justly
celebrated for their generous appreciation and
encouragement of the cricketing talent of other counties,
and in the case of young Mr. Trentham there is something
peculiarly appropriate in the benevolence of their present
attitude, as it is rumoured that Mr. Trentham once had an
aunt who lived near Bootle.”