Sunday, November 29, 2009

Public Health Measures and Mortality in U.S. Cities in the Late Nineteenth Century

Public Health Measures and Mortality in U.S. Cities in the Late Nineteenth Century Author(s): Gretchen A. Condran and Eileen Crimmins-Gardner Source: Human Ecology, Vol. 6, No. 1 (Mar., 1978), pp. 27-54 Published by: Springer Stable URL: http://www.jstor.org/stable/4602436 Accessed: 29/11/2009 14:48 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=springer. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. Springer is collaborating with JSTOR to digitize, preserve and extend access to Human Ecology. http://www.jstor.org
Human Ecology, Vol. 6, No. 1, 1978 Public Health Measures and Mortality in U.S. Cities in the Late Nineteenth Century Gretchen A. Condran2 and Eileen Crimmins-Gardner3 In this paper we examine the decline in mortality rates by cause of death in U.S. cities during the last decade of the 19th century. Causes of death are grouped according to their probable relationship to specific public health mea- sures. The reduction which occurred in the death rates from some diseases, e.g., typhoid and diarrheal diseases, can probably be attributed in part to the pro- vision of sewers and waterworks. Large declines also occurred in the death rates from tuberculosis and diphtheria, but the relationship between the declines in these diseases and public health practices designed to combat them is more am- biguous. We therefore conclude that public health measures had some impact on the decline in mortality, but that these measures do not provide a complete explanation of the mortality decline. KEY WORDS: mortality; public health measures; urban areas. INTRODUCTION In the United States, high population density, inadequate water supplies, and inadequate sewage disposal systems aided in the spread of infectious diseases and contributed to high urban mortality levels for most of the 19th century. Ur- ban mortality began to decline toward the end of the century; however, the exact date of inception of the decline is hard to determine. While certain diseases (e.g., cholera, typhus, and smallpox) declined earlier than others, urban mortality in The research on which this paper is based was supported by NICHD Grant 1-ROl-HD- 05427. A version of this article was presented at the meetings of the American Sociological Association, New York, August 30-September 3, 1976. 2Department of Human Ecology and Social Sciences, Cook College, Rutgers University, New Brunswick, New Jersey. 3Department of Sociology, University of Illinois at Chicago Circle, Chicago, Illinois. 27 0300-7839/78/0300-0027 $05.00/0 ? 1978 Plenum Publishing Corporation
28 Condran and Crimmins-Gardner general appears to have remained high and may even have been rising until late in the century. By 1890, mortality levels were probably declining in many cities (Sydenstriker, 1972; Easterlin, 1977; Meeker, n.d.; Gore, 1904; Taeuber and Taeuber, 1958). The decline in mortality - rural and urban combined - which occurred during the 19th century in England and Wales and the U.S. has been attributed largely to changes in levels of living (McKeown and Record, 1962; Meeker, 1972). However, in urban areas new public health practices are also be- lieved to have contributed to the decline in mortality. Meeker (1972), writing about the period from 1880 to 1915, argues that: The improvement in the state of health was due to a rising level of living and to novel public health measures. The former refers especially to better diets and housing, the latter mainly to the installation of sewers and filtered, or otherwise improved, central water supplies. (p. 367) Thus, Meeker suggests that the public health measures which affected mortality rates involved quite specific practices. In this paper, we examine the urban mor- tality decline in the last decade of the 19th century in relation to several of these public health measures. The public health movement in general encompassed a large number of goals and procedures, some of which probably had an impact on mortality, others of which may have had little or no effect. Our analysis con- centrates on measuring the impact of specific public health measures on those causes of death which were likely to be affected by them. THE DATA All the data reported in this paper are taken from the federal censuses. Data on the number of deaths were collected in the regular decennial enumera- tions from 1850 through 1900. However, by 1880 the census mortality statistics were a combination of the enumeration data and the returns from the registra- tion systems in several states and cities. Urban death rates in this paper are based solely on the city death registration data as published in the federal censuses. Most of the city registration systems had been in operation for several years prior to 1890, and therefore the data are considered to be fairly reliable from that date on, and certainly a better source than the census enumeration. In 1880 the number of registration cities is small and therefore our research is based on data from 1890 and 1900.4 Base population figures used in the cal- 4The research reported in this paper is part of a larger study in which the changes in morta- lity from 1850 to 1900 and the regional variation in mortality in the second half of the 19th century are being examined in conjunction with the levels and trends in fertility. We are using the 1890-1900 decade for this examination of urban mortality because it is the first time period for which reasonably reliable data are available. An extension of this examination into the 20th century is also being undertaken and should provide further evidence on many of the issues raised in this paper.
Public Health Measures and Mortality 29 culation of mortality rates were collected in the federal censuses of 1890 and 1900. Coverage in these censuses was not perfect and any inadequacy in these data affects our results. Age-specific and cause-specific death rates have been calculated from data for 28 cities in 1890. The death rates for some of these cities are very low, probably as a result of poor registration. When data for individual cities are shown, all 28 cities have been included in the tables. When data are aggregated, only 26 cities with infant mortality rates over 120 per thousand were used. Data for 1900 are available for 127 cities. All correlations, however, have been calculated using only 115 cities with infant mortality rates greater than 120 per thousand because of the possibility that low rates result from poor registration and that their inclusion would bias the results. This criterion has eliminated many Midwestern cities from the analysis. EVIDENCE OF DECLINING URBAN MORTALITY Table I contains the total mortality rates and the age-specific mortality rates for 28 cities for which registration data are available for 1890 and 1900. Most of the cities in Table I show declines in some mortality rates. The declines in mortality are not uniform across all cities and, in most cities, are limited to the younger age groups. There are, of course, dangers in inferring the trend in mortality from data at only two dates. Mortality rates may fluctuate a great deal, especially when mortality levels are high and when deaths from infectious and epidemic diseases constitute a large proportion of the total deaths. The mortality statistics have been collected on a yearly basis for many of the 28 cities in Table I as well as for a number of other cities. Yearly mortality rates were calculated using population estimates based on an assumption of a geo- metric change in the population between 1890 and 1900. Five-year moving averages of the death rates of 04-year-olds for a number of cities are plotted in Fig. 1. In most cases, these figures show a downward trend in mortality over the decade. The trend in the mortality rates shown in Table I and in Fig. 1 depends on both the adequacy of the death registration system and the reliability of the population estimates used in calculating the rates. A particular problem for our analysis is that the population is thought to have been more poorly enumerated in the 1890 census than in either the preceding or following censuses. We must therefore be cautious about inferring a decline in mortality. If the registration of deaths was fairly constant from 1880 to 1900 but the population enumeration was especially poor in 1890, artificially high mortality rates in 1890 may be sug- gesting a decline in mortality which actually did not occur. The available evidence strongly suggests that while it is hard to determine exactly when mortality began to decline, death rates were actually dropping in cities during the last decade of the 19th century. The raw figures on number of
30 Condran and Crimins-Gardner o' 0- ~o 0 O O- '- 90 o a C- 00 - :t 00 "kD Wm r 00 -- C ^m 4 o W) 00 C- W) ,t r-00 s-C . 00 0- r)00 ? O C :t O en 00 _ -z n O= Ol 00 D? 0-) ) 00 LO) 0O ?-) 0e 9 0 rO1t-mC t- t, oo o6 tO ,0 c, u6 o6 kr Oj c6 ,0 tr oO. cr 4 oo 4 4 6 c^ t-- 4 w;r l 0 m en en en en en n m c O tt m "m 001 m ? ^0 m Cl o t 00 00 N C 00 a-0, - O 00 en00 C,, 00 % M m 0) 0_ O N00 00 V- - t V- CD 0 r- - al OO oo v) cI, Nt e#~ 00 ON ut) CO co-.t 00o -0 tn -D -- -4 -t m n 00 CD ON cn aN 00 C' cn cn "t - t O~~~~~~~~~~~~- -- .-I _ 4 _ - -4_ --I_ 4 ll o - _ crn c>l to~~~V) 00 o x tO en O n c Om ?o kn 0 N On O- > CO 00 r- stf u 00 v~~~~~~~~~~~~~~ 06 O- o6 oo6 oo~ 'i cr oo C6 O. r_ C) O- ,6 ?-. oo 0 t- t- W tC c O0 00 00 C- " I - 0 C O 00 00 l ^ 0 o O 0 C.) ONa t v- oo r-e en^ CE M C) -4 ct " I' O ?- o M r- - ol " ON ootO_ 0 ) ~ - o ~ 00 O~ 00 O~ 00 O 00 ~C00a- t Ol 00 O0 0 O0 O 0 O 00 O" 00 1 t 0 r- 00 0q - ---O t- - - ---n ---- - t- - - v) -n c :O C" 00 CN C) CN en n m :t cn C> M O" o- D 00 -m ID m 00 ~~~ Q Q Q - .8~~~~~~~~~~~ " M C, rl M N cn q n eq I m- c" en "o en r 00 00 rl -o C) c n - cl r- r- - c ^o - rl ON - r- 'It C) CD -o 00 ON tn co>?o n r- cn .o al 'I C) v) - o n C) m CD -4 C t C) C) CN m en v) 0 00 all o t n tn oO r- 0t cro - C) -q 0 CD cq 0 0c -4 C4 Cl -4mN m b -4 C4 C4 -4 C4 C- -- -4 -- -- CCD N 0 ON )ON C)0- C= 0- C) N C) CN C) ,- N ON ) CN ) CN CD 0~C N O 0 aY O0 CN 0 CN 00 0- 0 C -0 ON 00 ON 00 X 00 ON 0 N O0 Ol 00 C,, 0N 00O *~~~~~~~~ 0 Y 000000000000000000~~~~~~~~~~~~~~~~~~~~~~~~~~~l00000 000 ~~~~~~~~~c w t _ $4 _1 _ . E~~~~~~~~' - '. 0 3 oc d 3=>
Public Health Measures and Mortality 31 n ?O r- 1t o) " 4 m m. tn tn o- N N 11 ?O m ON C N (7- oo N m oo 'I N ~ c1 o stoo or1mNottmob ---- - - -- - , - CD ?D 00 "- ~O 00 N CD N M 1- r- "t r- 1, O N CD NNm N N I - ? O tn -t -N n t ----- - - T N N m - - --t -T 0 - O -zt oo -t - o t -o N ,-O - - -- oc ON -'>N n dD en N w _N eq N N N N H- tn o N cn 'oo oo0 W) 1% D CD It W Ioo 'I 'oo 1 ) oo u) 'I 'I CD O u)- r- oIT ON m - - -- r-_ - - - -- - -4 -4 ce H -~~~~~~~~~~~~~~~~~~~~~~~~~~~~* - a-. - CD-4 ".0af) 00/0 00 C '-4O O NK '-IO 00 0 - 0 -f)" ) 040 W4C 'af CD l O ' 0 O 00 00 a-, CD 00 0O ON C , O n o o O O > N C N 0 ^ O O W _l 66 - ;4 6 _. 4 0~6cl 00 16 -V,It' 't l el;C - v 0u)1. tn 'I -- m r- tl V- Od IT 00 0) N O ?? m m 1. 01 O N N N% ^- ~ ON 0 _C5 oo a 110 00 0c"ItGo ONo- . 000- rW) NON t~ tOl- ~Cl~af)~-~Cl000000Cl0 K'?00Cl~.0Cl00r-0 O o - -y -- t O t- O - ---N O --- - v t C ---- O ^ tNt^mb^N^>ttnv)^NN^^n^^^mmt>~~~~C CO I u) U .-I N O CO M o ,* 0 O (= cn d o C)0 t0 0 D oo~ ro xoq v- ,O m m W) m O n tn m tn o ^ , qo W) o oo m % O ri CD c - 0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C ON Z > N Z t0 - F M t C N N ^ o o6 6 i -; D aQ N -4 d e N M :t en -- -- n 'It en " N - N N N N-I C14 ct M :c t m N 0 o oo 11. W) ur b N -f 00 N W) N %- W) 1. (7 N N 00 -t en CD 0 ? r- on _s _ ) % o W\ I N o oen o= N -4 t- t)1 -.o en 00o t~- _/ N cq N N N _ 6 N eq N N > _ N _ N CI N _ tn N t- -it N - t_ r- r_ r- 1 t_- N 00 00 00 cl O 00 N 'I N 00 [ O Q Q) * ~ -; 6 6 o6 r- 4 4 c- * - 4 * - * - * * 6 tr * ~ * t* * ~ w * 6 * . v @ 4 < o=" ON" o=" a-">X2 :
32 Condran and Crimmins-Gardner - 0)~~~~~~~ t gqOZ { 5 R ? : =~~~~~~~~~~~I r 00000'~~~~~~~~~~~~~~ c/i 00 CZ k ~ ~ 0CO 0 0~~~~~~~~~~~~~~~~~~~~~~~~~~~~0 VI .0) 00 V)~~~~~~~~~~~~~~~~~~~~~~~~~C 0~~~~~~~~~~' Co~~~~~~~~~~~~~~~~~~~~~~~~~~~~C J' '11 "0) a o. a* >C',
Public Health Measures and Mortality 33 Table II. Disease-Specific Death Rates for 26 Cities: 1890-1900 (Rates per 100,000 Population) 1890 1900 Measles 18.0 17.7 Scarlet fever 17.0 15.4 Diphtheria and croup 116.9 59.2 Whooping cough 17.6 13.2 Typhoid fever 53.0 33.8 Malarial fever 19.4 6.9 Pneumonia 223.2 241.7 Consumption 272.9 218.1 Diarrheal diseases and cholera infantum 216.1 153.1 Cancer and tumor 53.5 65.8 Heart diseases and dropsy 124.0 134.1 Liver 28.0 25.8 Diseases of the nervous system 260.8 206.2 Diseases of the urinary system 84.9 120.0 Old age 32.9 45.7 Unknown 9.8 8.6 All other 637.9 567.3 Total 2185.5 1932.6 Source: 1890 - U.S. Census Office (1896), Table V, pp. 410-655. 1900 - U.S. Census Office (1902b), Table XIX, pp. 285-555. deaths used in calculating the annual death rates in Fig. 1 show that for many cities the actual number of deaths was declining. Hence, even with more moderate rates of growth of the population than indicated in the 1890 and 1900 census returns, a mortality decline would be evident. In addition, the decline in morta- lity was not uniform for all causes of death but, as expected during a period of ini- tial mortality decline, occurred largely because of a decrease in the deaths from in- fectious and parasitic diseases. Table II contains the cause-specific death rates ag- gregated over 28 cities in 1890 and 1900. Mortality rates from many infectious and parasitic diseases were declining, while the rates of death from cancer, heart disease, etc. were not. If the observed decline in mortality were a result of a shift in the quality of the statistics, there is little chance that this pattern of decline in cause-specific rates would occur. We therefore conclude that a real decline in mortality was occurring in many U.S. cities by the end of the 19th century. PUBLIC HEALTH MEASURES AND THEIR RELATIONSHIPS TO MORTALITY DECLINE By 1890 the public health movement included a broad range of activities designed to combat both endemic and epidemic diseases. The stated aims of the movement in most cities included first and foremost the provision of central
34 Condran and Crimmins-Gardner water supplies and sewage systems. In addition, the movement often included plans for general street cleaning, improvements in slum housing, inspection of food and milk, use of quarantine and disinfection practices, and the distribution of diphtheria antitoxin (Rosen, 1971; Duffy, 1974; Meeker, 1972). The actual practice was often far removed from these aims. Detailed studies of the develop- ment of public health practices in local areas point out the failure to meet goals, sometimes because of lack of technology, sometimes because of political dis- agreements, and often because of corrupt city officials and contractors (Duffy, 1974; Rosenkrantz, 1972). We can find examples of specific cities in which public health measures appear to have reduced mortality from certain diseases. Whipple (1908) compares mortality rates from typhoid and diarrheal diseases in Albany, New York, and Troy, New York. Albany began filtering its water in 1899, Troy did not. Albany's decline in both typhoid and diarrheal diseases far exceeded that in Troy (Meeker, 1972). The drop in infant death rates in the City of New York between 1885 and 1915 from 273 per thousand live births to 94 per thousand live births has been attributed to the improvement of water and milk supplies (Rosen, 1971). Indeed, in reports to the Census Bureau in 1900, local health authorities clearly indicated public health measures as the major source of declines in mortality rates. The Census Bureau requested a statement from health officials of the local causes operating to diminish the death rate. The replies varied in length and detail, but in every case the decline was attributed to some public health measure. These replies, as reported in the 1900 Census, are summarized in Table III. One is, of course, tempted to believe all these reports and to generalize from the clear cases of decline in some specific diseases in local areas to the over- all decline in mortality in cities. Clearly, the various activities included under the general heading of the public health movement could have an effect on mortality only inasmuch as the goals were translated into actual practice. It is quite clear -that different aspects of the public health programs in cities were likely to affect some diseases and not others. Some causes of death - cholera infantum, diar- rheal diseases, and typhoid - should certainly have been directly affected by such public health measures as sewers and waterworks. Other infectious dis- eases - tuberculosis, diphtheria, measles, influenza, scarlet fever, whooping cough, etc. - were less related to these specific public health measures. However, they may have been affected by general improvements in cleanliness and by quarantine and disinfection practices and, in the case of diphtheria, by the de- velopment and use of antitoxin. Finally, the etiology of other diseases like cancer and heart disease suggests little or no connection to any public health measures of the time. A breakdown of diseases into categories reflecting their relationship to public health practices provides a way of assessing the importance of these practices in explaining the urban mortality decline.
Public Health Measures and Mortality 35 DIARRHEAL DISEASES, CHOLERA INFANTUM AND TYPHOID: SEWERS AND WATERWORKS Table IV contains the mortality rates from diarrheal diseases and cholera infantum combined and from typhoid for 28 cities in 1890 and 1900. Diarrheal diseases and cholera infantum show a fairly large decline in a number of cities and at least a moderate decline in all but five cities. Typhoid declines in all but two cities. The size of the decline in typhoid is very impressive in some cities, particularly in those cities which started out with very high rates in 1890. These findings suggest that sewers and waterworks may have contributed to the decline in mortality in these cities. Data indicating the extent of sewers and waterworks in these cities are very limited. While some data for selected cities are published in the reports of public health officers, most data are unusable for comparison across cities or for observation of changes in practices for more than a few cities. The 1890 Census and 1901 Statistics of Cities published by the Department of Labor provide some information on these phenomena. In 1890, data on miles of sewers, num- ber of sewer connections, and gallons of water consumed were published for the 28 large cities for which mortality data were published. For 1900, the available data include miles of water mains, expenditures up to 1900 on construction of waterworks, and the amounts spent on yearly maintenance of sewers and water- works. Because the data for 1890 and 1900 are not comparable, changes in public health facilities cannot be calculated from these sources. Our use of these data is limited, therefore, to an analysis of the cross-sectional relationships between public health measures and mortality rates. Because 1890 mortality data are available for only a small number of cities, we have undertaken this analysis only for 1900. The expenditure variables - expenditures up to 1900 on the construc- tion of waterworks, and the expenditures on the yearly maintenance of sewers and waterworks - were calculated per capita. Only the relationships between the expenditures up to 1900 on the construction of waterworks and mortality rates are shown in Table V. The results using the other data relating to water- works are similar to those in Table V. The expenditure on maintenance of sewers gener4lly shows no significant relationship to mortality rates. Two variables were calculated using the data on miles of waterworks - the miles of waterworks per person and the miles of waterworks per acre in each city. In Table V we are pre- senting only the miles of waterworks per person. Similar results were obtained when the other variable, miles of waterworks per acre, was correlated with the death rates. All the correlations in Table V were calculated controlling for city size, percent Black, percent foreign-born, and persons per dwelling unit. Expenditures on the construction of waterworks and the miles of water- works per person show significant negative correlations with mortality rates in
36 Condran and Crimmins-Gudner ^~~~~~~~, o o~~~~~~~~~~~~~~~~W ct v ::~~~~~~~~~~~~~~~~~~~~4- E ~ ~~~~~~~~~~~~~~~ 0 .O~~~~~~~~~~~~~~~~~~~~t Ot ux O~~~~~~~~~~~~~~~c < << *i :^~~~~~~~~~~~~~~C r 3 o xx x x~~~~~~~~~~~~~~- Y r~~~~~~~~~ o 4-o o cn's C' Ev41. 0 ~ ~~~~ < UUc t Y
Public Health Measures and Mortality 37 O 4-4 Q~~~~ > 0 bo _ t o vz cn 45~~~~~~ X X X X Xx 00 X ~~~~~~ Cd 0~~ O~ Q ~~~~~~~~ 0 ~~~~~~~~~~ ~~ 0 -4A >44) ~~~~~~ ~~~~)0 ~0 0 ~ 4) 0 0 C 0 Z O z z 04V tsZr
38 Condran and Crimmins-Gardner Table IV. Death Rates from Diarrheal Diseases and Cholera Infantum and Typhoid Fever, U.S. Cities 1890 and 1900 (Rates per 100,000 Population) Diarrheal diseases Typhoid and cholera infantum fever Allegheny 1890 171.0 182.4 1900 172.4 103.2 Baltimore 1890 307.1 46.5 1900 185.7 34.8 Boston 1890 199.1 38.8 1900 132.1 27.3 Brooklyn 1890 234.4 24.1 1900 209.9 21.0 Buffalo 1890 233.5 31.3 1900 133.4 25.0 Chicago 1890 254.3 72.2 1900 150.4 27.7 Cincinnati 1890 140.8 50.9 1900 127.6 32.5 Cleveland 1890 204.7 62.8 1900 111.3 48.2 Denver 1890 203.4 217.4 1900 77.7 42.6 Detroit 1890 236.2 19.4 1900 151.2 18.9 Indianapolis 1890 138.5 54.1 1900 104.0 45.5 Jersey City 1890 198.8 82.2 1900 181.7 22.3 Kansas City 1890 143.9 39.9 1900 121.5 40.9 Louisville 1890 107.4 75.7 1900 98.2 73.3 Milwaukee 1890 180.0 29.8 1900 137.7 18.2 Minneapolis 1890 156.0 57.1 1900 58.7 40.2 Newark 1890 253.0 99.5 1900 149.1 10.6 New Orleans 1890 294.6 18.6 1900 268.2 66.5 New York 1890 301.3 23.0 1900 175.6 15.4 Omaha 1890 89.0 44.9 1900 85.8 40.0 Philadelphia 1890 153.0 73.6 1900 118.3 37.8 Pittsburgh 1890 192.8 127.4 1900 197.4 147.4 Providence 1890 166.5 40.1 1900 163.4 29.6
Public Health Measures and Mortality 39 Table IV. Continued Diarrheal diseases Typhoid and cholera infantum fever Rochester 1890 182.2 39.6 1900 90.4 23.4 St. Louis 1890 118.4 32.1 1900 108.5 33.0 St. Paul 1890 227.6 69.1 1900 52.1 22.1 San Francisco 1890 87.6 55.5 1900 89.0 39.4 Washington, D.C. 1890 257.0 86.8 1900 175.4 80.7 Source: 1890 - U.S. Census Office (1896), Table V, pp. 410-655. 1900 - U.S. Census Office (1902b), TableX MX, pp. 285-555. Table V. Partial Correlations Between Mortality Rates and Two Measures of the Public Health Activities Cost of waterworks to 1900 per person Miles of water mains per person North- North North- North Total east Central South Total east Central South Age N=91 N=46 N=16 N=11 N=96 N=48 N=16 N=11 Total -.16 -.18 .44 -.12 -.24 -.32a .27 -.23 Under 1 .04 -.01 .28 .11 .05 -.09 .45 *49a 1-4 .00 -.14 .19 .11 -.12 -.22 .19 -.45 5-9 .00 -.09 .43 .18 -.02 -.13 .50 -.39 10-14 -.06 -.01 .33 _.34 -.17a -.23a .36 -.06 15-19 -.10 -.22 -.03 .15 -.22a _.36a -.12 -.18 20-24 -.16 -.14 .49a -.39 -.19Ga -.27a .22 -.40 25-29 _.40a -.34a -.20 -.65a _.40a -.31a -.41a -.74a 30-34 -.17a -.13 .48a -.18 -.18a -.24a .34 -.35 35-44 -.29a -.30a .09 -.29 _ 33a -.42a .18 -.22 45-54 -.25a -.36a .53a -.08 _.33a -.45 .41a -.41 55-64 -.14 -.23 .37 -.16 -.20a -.32a .31 -.11 65+ .05 .01 .09 .16 -.04 -.07 .10 -.09 Diarrheal diseases -.05 -.02 .07 -.45 -.08 -.07 .28 -.54a Typhoid -.17a -.15 -.01 -.41 -.14 -.19 .11 -.07 aCorrelations are run controlling for city size, percent foreign born, percent Black, and persons per dwelling unit. Similar results occurred when persons per acre was substituted for persons per dwelling unit. Sources: Miles of water mains - U.S. Department of Labor (1901), Table 15. Cost of waterworks - U.S. Department of Labor (1900), Table 12. Death rates by cause - U.S. Census Office (1902b), Table xix, pp. 285- 555. Deaths by age - U.S. Census Office (1902c), Table xix, 285-555. Population - U.S. Census Office (1902a), pp. 122-149.
40 Condran and Crmmins-Gardner the middle to old age groups. There is variation in the pattern of these relation- ships from region to region. In the Northeast the relationships are negative; the more a city spends on sewers and waterworks, the lower the mortality rates. The South shows generally similar relationships to those in the Northeast, al- though the correlations are somewhat weaker. Finally, the North Central shows a significant positive relationship between public health measures and some age-specific mortality rates. We have no conclusive explanation for this ususual pattern of relationships by region. The positive relationship in the North Central region may be the result of poor public health administrations in some cities leading to both the underreporting of mortality and little or no expenditure on waterworks and sewers. Tentatively, however, we hypothesize that in the North Central region only cities with very high mortality had installed water and sewer systems. High mortality is therefore related to having begun the construction of sewers and waterworks. In the Northeast sewers and waterworks have a longer history and are much more widespread. In this region, therefore, public health facilities are negatively related to mortality rates in a number of age groups. As these findings indicate, there are difficulties in predicting even the sign of cor- relation between sewers and waterworks and mortality rates. On the one hand, if sewers and waterworks reduce the number of deaths from certain disease cate- gories, the mortality rates should have a negative relationship with these public health measures. On the other hand, if high mortality leads cities to make major expenditures on sewers and waterworks, the relationship with mortality may well be positive. Other problems in using these measures exist. The introduction of central water systems and sewage systems does not always have an immediate impact on disease. Provision of a central water supply, for example, is not synonymous with the provision of uncontaminated water. In some cases, early systems pro- bably continued to supply contaminated water and, in other cases, they may even have replaced uncontaminated water with contaminated water. Filtration of water supplies had a much more clear-cut effect on mortality reduction. Data on the initiation of filtration in individual cities are not available for the cities used in our analysis. While water supply systems were first built in some cities just after the beginning of the 19th century, it appears that it was much later be- fore filtered water was provided. According to Winslow (1923), in 1880 only about 30,000 people in U.S. cities over 2,500 in population were supplied with filtered water. Sewage systems should have operated to reduce mortality by inhibiting the spread of a number of diseases. Sewage systems were fairly widespread by 1890. "Of the 96 cities of 10,000 and over reporting whether they had sewers, only 26 had no sewers at all" (Meeker, 1972). The building of a sewer system was, however, a long-term project for a town. Portions of the town were gradually added to the system. In 1890 the cities reporting a sewer system had, on the average, less than
Public Health Measures and Mortality 41 one-half of their dwellings connected to the system (Meeker, 1972). Consequently, data reporting the existence of sewers or a year's expenditures on sewers are somewhat inadequate to deal with the questions we are asking. Table V also contains the correlations between diarrheal diseases and ty- phoid and the two public health measures. The correlations are largely insigni- ficant. A regional pattern similar to that observed using age-specific death rates occurs. The relationships between these two diseases and the costs and miles of waterworks are very weak. Indeed, the significant relationships between age- specific mortality rates and these public health measures are largely explained by disease categories other than the two in Table V. These unexpected results may be due to many of the problems with these measures outlined above. Deaths from diarrheal diseases and typhoid, two disease categories which should have been directly affected by the construction of sewers and waterworks, did contribute significantly to the overall decline in mortality in these cities. This result may indicate that sewers and waterworks are important in explaining the mortality decline. The lack of correlation between these particular disease categories and the expenditures on the construction of waterworks and the miles of waterworks per person casts some doubt on this conclusion. The decline in diarrheal diseases and typhoid may well have been due to other factors which affect other disease categories as well. However, because of the limitations of the data and the problems inherent in making longitudinal inferences from cross- sectional data, we hesitate to place too much emphasis on this result. PUBLIC HEALTH COMPAIGNS AGAINST TUBERCULOSIS AND DIPHTHERIA Table VI contains the mortality rates from tuberculosis and diphtheria and croup5 from 1890 to 1900 in 28 cities. Diphtheria and croup declined in all but one city. Deaths from these diseases were moderately high, but were often only half as high as the deaths from tuberculosis, which was the leading killer in many cities. Tuberculosis declined dramatically in many cities and hence was a major contributor to the overall mortality decline. The relationship of the incidence of tuberculosis to public health practices, or for that matter to any factor influencing mortality, is very hard to establish. Since we have no general data on public health programs designed to reduce tuberculosis in all of these cities, we must make inferences from examples of campaigns against tuberculosis in a few places. 'The available data made it impossible to separate this disease category and to look at the changes in diphtheria alone.
42 Condran and Crimmins-Gardner Table VI. Death Rate from Tuberculosis and from Diphtheria and Croup, U.S. Cities, 1890 and 1900 (Rates per 100,000 Population) Tuberculosis Diphtheria and croup Allegheny 1890 147.2 121.6 1900 160.9 28.5 Baltimore 1890 293.0 55.9 1900 220.3 73.5 Boston 1890 375.7 142.3 1900 245.0 75.6 Brooklyn 1890 288.3 169.4 1900 224.0 72.2 Buffalo 1890 186.2 86.1 1900 131.7 26.4 Chicago 1890 175.9 140.5 1900 178.6 64.8 Cincinnati 1890 280.2 164.7 1900 235.3 33.1 Cleveland 1890 158.8 147.3 1900 131.8 51.6 Denver 1890 240.8 178.1 1900 325.7 33.6 Detroit 1890 162.2 174.9 1900 125.0 46.9 Indianapolis 1890 283.6 138.5 1900 202.2 104.0 Jersey City 1890 271.8 198.8 1900 237.4 181.7 Kansas City 1890 179.3 143.9 1900 204.0 121.5 Louisville 1890 281.1 107.4 1900 233.5 98.2 Milwaukee 1890 183.9 180.0 1900 149.3 1 37.7 Minneapolis 1890 153.0 156.0 1900 130.2 5.8.7 Newark 1890 326.7 253.0 1900 234.1 149.1 New Orleans 1890 343.8 294.6 1900 357.7 268.2 New York 1890 387.5 301.3 1900 252.6 175.6 Omaha 1890 67.6 89.0 1900 107.3 85.0 Philadelphia 1890 279.6 80.6 1900 231.1 97.0 Pittsburgh 1890 149.2 189.4 1900 135.6 38.9 Providence 1890 303.5 93.8 1900 226.7 29.0
Public Health Measures and Mortality 43 Table VI. Continued Tuberculosis Diphtheria and croup Rochester 1890 213.6 45.6 1900 153.7 28.3 St. Louis 1890 184.6 61.8 1900 203.0 52.7 St. Paul 1890 125.4 104.4 1900 126.3 31.3 San Francisco 1890 378.3 58.9 1900 308.9 26.5 Washington, D.C. 1890 359.0 83.3 1900 305.3 75.3 Source: 1890 - U.S. Census Office (1896), Table V, pp. 410-655. 1900 - U.S. Census Office (1 902b), Table XIX, pp. 285- 555. In the detailed history of the public health movement in New York City, Duffy (1974) describes the antituberculosis program there. In 1889, several con- sulting pathologists for the Board of Health published a "Report on the Preven- tion of Pulmonary Tuberculosis." The authors urged the dissemination of the knowledge of the contagious characteristic of tuberculosis and the concomitant danger of human contact. They also urged the disinfection of the premises which had been occupied by tuberculous individuals, and recommended the inspection of cattle, inasmuch as diseased cattle were a source of the disease. The informa- tion contained in this pamphlet was printed in a circular for distribution in several languages. Its distribution was fairly limited, and a program of general education was deferred until 1893, when a circular, "Information for Consump- tives and Those Living with Them," was printed in English, German, Italian, and Hebrew, and sent to tenements in the city. The Health Department started a campaign for the laboratory diagnosis of suspected tuberculosis cases, and for the reporting of tuberculosis cases diagnosed by private physicians. The latter was met with opposition from physicians, who had also been directly hostile to the education campaign. Physicians were influenced by their patients not to report deaths from tuberculosis because many insurance companies were syste- matically negating policies in the event of death from tuberculosis. Opposition from the medical profession, ostensibly on the grounds that the Health Depart- ment was interfering in the doctor-patient relationship and that tuberculosis was not a communicable disease, continued until the turn of the century. Ef- forts to make tuberculosis a reportable disease were blocked by the medical profession until 1900, when the New York medical profession finally recognized the value of the Health Department recommendation. During the decade 1890 to 1900, a period in which we have seen substantial reduction in the disease, the campaign against tuberculosis was very limited. Certainly there was little in the
44 Condran and Crimmins-Gardner description of the campaign to suggest that it was the sole or even the most im- portant factor in the decline in the deaths from tuberculosis. Similar evidence on the campaign against tuberculosis in Massachusetts (Rosenkrantz, 1972) is even more ambiguous as to its likely effect on mortality rates. Indeed, in Mas- sachusetts no substantial activity against the disease began before the turn of the century. Diphtheria, a disease which shows the most consistent decline over our set of cities, has a somewhat ambiguous relationship to the public health move- ment. According to Duffy, the major innovation affecting diphtheria was the introduction of an antitoxin. The value of the antitoxin was established in January and February of 1895, when it was used to end an outbreak of diph- theria in the New York Infant Asylum. The following years saw the Division of Laboratories steadily expanding. In 1896, six additional bacteriologists and two more laboratory attendants were added to the staff. Through constant experimentation and by refining existing processes, the quantity and quality of diphtheria antitoxin was improved and production cost reduced.... The introduction of more effective antitoxin and better methods of administration reduced the mortality rate by 50 per cent within the space of six months, and by the end of 1897, the Health Department was well on its way to controlling diphtheria. (Duffy, 1974) The decline in diphtheria, at least in New York, appears to be closely related to public health activities. With so little evidence, we have difficulty generalizing about the impact of the campaigns against tuberculosis and diphtheria on the decline in death rates from these diseases. Because we can be fairly sure that these public health cam- paigns were restricted to urban areas, we have compared the declines in mortality from these causes of death in rural and urban parts of registration states. The re- sults are shown in Table VII, which contains the death rates from tuberculosis and diphtheria for 1890 and 1900, as well as the percent decline in these rates. Declines in both disease categories occurred in the rural as well as in the urban parts, and the urban declines are not consistently higher than those in rural areas. These results differ from those for diarrheal diseases, also shown in Table VII for comparison. This disease category, for which the relationship to specific public health measures seems more likely, does show very small declines in rural areas and larger declines in cities. Rural declines in diphtheria and tuberculosis pre- sumably occurred independently of public health activities. From this compari- son, we conclude that tuberculosis and diphtheria may well have been declining independently of public health campaigns in cities as well. THE DECLINE IN OTHER DISEASES The other causes of death categories for which data are available seem un- likely to have any direct relationship to any public health measures. Death rates
Public Health Measures and Mortality 45 Table VII. Selected Cause-Specific Death Rates, Rural and Urban Parts of Registration States, and Percent Decline, 1890 and 1900 (Rates per 100,000 Population) Diphtheria Diarrheal and croup Consumption diseases Cities Rural Cities Rural Cities Rural Connecticut 1890 132.7 72.3 270.5 209.6 172.8 141.4 1900 41.6 30.9 192.4 147.7 144.9 136.0 % decline 68.6 57.3 28.9 29.5 16.1 3.8 Massachusetts 1890 112.9 61.4 282.0 227.9 192.5 98.7 1900 51.9 24.8 193.7 162.5 154.3 90.1 % decline 54.0 59.6 31.3 28.7 19.8 8.7 New Hampshire 1890 143.0 63.2 191.9 194.3 211.8 125.2 1900 41.1 18.7 173.0 142.4 161.8 103.9 % decline 71.3 70.4 9.8 26.7 23.6 17.0 New Jersey 1890 147.0 55.4 269.4 193.4 222.9 117.6 1900 58.4 36.4 202.3 151.6 147.4 114.8 % decline 60.3 34.3 24.9 21.6 33.9 2.4 New York 1890 126.7 41.7 308.8 148.8 246.8 95.9 1900 56.4 21.7 220.3 138.0 164.3 82.6 % decline 55.5 48.0 28.7 7.3 33.4 13.9 Rhode Island 1890 86.0 76.3 294.9 227.6 212.4 198.7 1900 29.4 29.9 213.9 165.1 189.0 178.6 % decline 65.8 60.8 27.5 27.5 11.0 10.1 Vermont 1890 44.0 86.2 241.9 195.7 206.6 91.7 1900 16.5 18.5 155.9 152.1 142.7 66.7 % decline 62.5 78.5 35.5 22.3 30.9 27.2 Source: 1890 - U.S. Census Office (1896), Table V, pp. 410-655. 1900 - U.S. Census Office (1902b), TableXIX, pp. 285-555. from these causes are given in Table VIII. No disease category shows a decline in every city; for every disease there are at least one or two cities which show a rise in the rate between 1890 and 1900. Small changes in disease categories may be due to annual fluctuations or to changes in the diagnosis and classification of some diseases and therefore should not be regarded as evidence of a significant change. A number of diseases - measles, whooping cough, and scarlet fever - do not constitute a large proportion of the deaths in any city. Consequently, even though they show large proportionate declines in a number of cities, their contri- bution to the decline in total mortality in these cities is small. These diseases also have rising rates in a number of cities. They remain a fairly small proportion of the overall mortality rate in the cities showing an increase. The remainder of the diseases in Table VIII, with the exception of nervous system diseases, generally show little or no decline. Pneumonia, a major killer in these 28 cities, showed very small declines in some cities and slight rises in others.
46 Condran and Crimmins-Gardner cO 0o oo o - 00 oo- D o oO C10 cl- m - 000 z CD (-- C) on "t m co t N 00 t m 0000 0 S o00 ClO -0 00> 0 ' WC ') o C Coo o z > 0C 0e6 Ni 00 000 0 aO 00'r) -''0- C-C * C00 _N 00 0 m- 00 0 ~~~ '1-~~~~r- ~ 00- C'- t 'r)'- 0-C' 00ZrW ') 0- 0'~ C)~ _o c C) o C' _ - _ 1.4CA C) (o O C't - Clef Cl0 00 000 OCf)0 OC' 'tor 0 COC) e " --4 - ~ ~ Cf)~~f C''-00 0r- (C) '0 00 '~ ~ C)r f0 0) CA 'a + 0 C>-00 00 -- O m 0-0-m 0fC) '-C c7N O 0 -C co C-I-') 0- C- C W-I 0? C;O C,;I I ct &:> o ^ m rl \D m vo o0 o-. 0_ CD oom x a- 00 ~ ~ ~~~0 :t 0%0 Cf)n 0~0 '0 0''C) 0 Cf) cf) c0~ Or-0 0'Cf CZ W s v ? O ?~~~~~~~~~~0 t- 00 r- Ol 0 0 00 DO Up 14 06 06 ooi r- O e,; 0 00 CO ~~ 0)5 00'C)0 CD0 CD -,- CIr-2 oOO C C DCl'--0- lc CD'C)0 O- X . '(cf) E a- 5 o C-C- a _ ' 0N _0C O CD c7N ' 0rN O O\ oo(- oo a-, oo o- 0 oo oo (ON U o a- oo > 00 C1 > 0 CJ ) 9 = ~0oo -C- '0-0 C) 00- t-0 r---- Cr)0 jC C:-I - ef ?? ? - - - ? dw - C5 _ CO U > s- 0) U m' m v v
Public Health Measures and Mortality 47 4 l - 00 m - -- I C It t- -4 co -e N orl 0oo N oo T I N ON t 00 h-n Clt WC Cl - tl'- q rl --q '- c--' e Ic 00 e I Ciir om c or eO r-0 d0 N 00N 00 m e . - ov 0oON ri . .; .6 . . . .4 . * .; . .; .,6 .- . . . . c oot N m 00 C O~ Ct I n o ; -- C,; 00C ~ 4,QC Q 0l O 00 C N 0OD -CI CICI (~CI '-4- CIIN c-4 4-4 ( Cl n oo ( OR I-o r-O 00CCC) W 0I C cl.- 000 0 0 co 0000 10 On 00 "C0 C - 0 0-t N It r O N r- 00 O N q cq cq c N cq cq c N I" N cn N- ~~40 C~~~40 ~r C1 f) r- a 'IO ON C --D 4040 C1C1 lf e '00 0 0 C0 n O0 O Cr-tf~ 0 -0 O 0 r -4 t- C1 - -D CI - m - ? ? ? I ? ' m o Q 0 o ^ocEo . o ?o ~oo O>40 N o oo Wo oo m ' > oo ? o c C cn tn t- It It N Wco cq cq C - N N W) m w- -4 '0- - { -4 - - -4 1- - - 4-4 -4 - -2- - ? o Cl C, Ol 0o Ot 00 I00 cl C^ Co C1 rl) It C tt 00 00 m <7> - cn ?O r- 'I 'I %DF? N cn co N tl ? - c7 _- 1-4 N -4 'I U C)t W It lo m 't 0 0) -s o o # o o U o o z 4 o s o o S o C C o z z 0) *,00 <> ss 00 cy 00 mo< oo o cr oo o 30 C, 00 ON 0 cr 3o 00 m Ce 00 csN n 3 * 2 z S S ~~~~~~~~~~~~~~~~~~~~z z z o
48 Condran and Crimmins-Gardner CO o a'Os o (t I:t m t' 00Io- 00 0) ~ -4~1 CTh ~-4C' l ll - - ~ C 06 ( C') - 0i Oa 00 tC c m 00 rN-4 oo . 00 oo 00 c1' eNr er' e _1 - _-L rN' ;4 ON I" N ?>?uot 'IC CO C06( 000 4 -t - -~ - ri - ri It -0 -' C 0 ^ 00 0 OD O cr n C me > CmN tr t6M6 0> .,.4 -W0V -0 cr' 1) 00 c lt , ) C.) c C.) ._ un > tn 00 ? o t u O N 01 ? C 9 ca O 0 o t0 t0 00 0 ' 0 00 00 00 - 0- 0 0 c ' o - C- - - - C - 4.4 cli 06 C O 0 4f X; v ~~~~~~~~~o o o CD CD 0 C) 0 o c oo ON oo ON oo or 9L0 os co cr- m0C ^~~~~~~~~~~~~~~~~~~~~~~- --- Cl -- mN m o o io
Public Health Measures and Mortality 49 c ' 06 0 000 ~-0 0 00 6 c-~ C5 ON0 '-4 00 0'Lr W; C i o o co C, t -.. co a,- m o N o oN on o ' ur 4 ? O co 00 00OC -- C 0O e4r ?0 C0 cr c o oo- O 0000 ( O=>O 0 C e 0D 00 0 C= t ON cn cli C7N craxo oom m o o "C O \D ci 00 o o- 'N u)0 o06 ic-i mo> -: (^c 0? b O0 cN N O 0 xD0 O r > c t-00 cicW N c - N - a)~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~~~~~- W N FF ~~~to O F C, cn C,, rl- 00 C - m tn co 1 tc m e - ci*' j- m (I) 0 0 r nu t00~~~C co 00 NO rq el Itt Ct C), bO c-ic-- 00~~~ 000 00~~0 Oci ~-0 cf- 0i'9 0~ ~ II -0~ 0 c C,; c i o o 00 00 ci o - t'o N o- ON C 'IO 00 0 t '-400 elO r- -- 0 e o - ;0 C, \- , C,;r ~ r_ -- k 0 \ 6 - --4ci i -4 -4 -;0 C)~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~ )C 000N - N LON 0 CF C) -4- _ -0a ci L >o ' ci ci o o LO t '- -4 ci co oo ON oo ?, oo coo a- > co 0 ooN c0 O 0 000~ ^ -0 0 - -0 0 00-4 -0 Oy x ooQ o ?o Q Q X
50 Condran and Crimmins-Gardner olc C en O t o-n - ,o eo 0m ot 00 Mt _ w; o o 6o 00 0&4 4? 6 o 6 - o o0 o ^ do 0ci r0- ef I'* r-O e 000 t ?o tn C) c bb 00 W) en o) r- CY- ?o oo C:: O ) 00 00 It aN 00 C1 db- _% Rt t dON 00' 0 o6 O o ; o6 00 o 60 o' en en c Rt O WN I O -4 C ^- ti , O N 'C O n 0 0 N Ot (4 ') tn W) t- W) t n It cn - W) co cc r- v co o r Q 0 en O Fo 0000 r- o Wo me -I> em ro0 ~ o o-m- oo 0 6 oR e-i 6r- oC6 6C el; > m CD to Q tmo - Th e e e e - n 0- , oo . FOt tn cl C14 cn cn co ? t aEn cli C) '0 0) - 0 to 0) te -- *)00 o0 -. 64 e-~~~r ~o'o- ~ r-=- o4 a; c oRoR el -id oR cd C1 N _0 W) _ l 'I It as 00 _) o 0) ~~ 64 ~~C-i 6 c el; a' ' ; 64 CL Q 0) - - - 0) 0) 0)c0 ' N - O _ m ON d 0 o o o *oo 3o o i 0o o '0 o o 3 CO ' 0' ' '' O ' 0) ' 0) a 0) _z - -) ~~~~~~~~~~~~~~ ~~ Z Z Z
Public Health Measures and Mortality 51 t rr -^ 00 C) N ooc O rv^oo o D 1- cro cn r-- _-- 11. oo _14 0i r r-m m 000 n a,, e I --~ 00 -04 cm c- rl- itc~c ~ enr14 - C1C (1~c'1 C (C cC 'IO oo) 00 0 ?r elC O N oo an 0e0 o- ) o 00 m cc; Ce 0 10t c N00 Om1) 00e'1 NO " -N- 'I 1. 0C c 00 aC4I, 0 n C-C l t I'l Ct-- C~-C W ~'r r-0 CC) OIQ%C-0 or- q " oce - rOC r-. C-LO 0)C0 a,4 C- - - (1~~~cl -.-I- -- -00 00 0 m 0 -- C m c 'O CD -q " ) en w; c; 6 o6 co r- co N c- ,6 t -4 4 o6 wi ,6 w en Ov) 1.0 vwv) 00 cq tS cko bm 0 r- 'IC ^ O- 'IO -4 - - oo - --q 0l; m cn~~~~~~ > -O0 ,0 00 d m 00C-- 00 E- O 00 . _- ( ( n t_ O C. 0 00 al m cn W) 00 00 ~ ~ ~ ~ 0 4~ ~~~ ~~~~~~~ tn %1 C)C ) -1 ) _1 CN~ (1 - O o~~~~~~o r 0 00 O ao 0 C 0 z 0 _ 0 0 00 2?- d C) ~~~~~~~~~~~~~ - - 4 ONCD O Ia, a,- ' C o o ,, o o NWoo r o o N o o o o a-, ooWo '~~00 ~~~00+ 0 00 0O 00 00 a,,00 N00 aN - - -- 0 -O o - -- cl - - 0 X P P 0
52 Condran and Crimmins-Gardner Deaths from cancer and tumors, heart disease and dropsy, and diseases of the urinary system, as well as from old age, actually showed a substantial rise in many cities. This finding is not unexpected at the onset of mortality decline; people who live into old age are subject to just these diseases. Diseases of the liver show little change. Influenza shows a marked rise. While it is possible that an epidemic disease like influenza might show this pattern of change over time, the classification of this disease also changed over the period, and the rise may be a result of the change in classification. Diseases of the nervous system show a marked decline. This category of diseases is hard to interpret. In 1890 when data by age and cause are available, the data show a high rate of infant and early childhood mortality from this cause. The particular disease within this category which affects infants and young children is convulsions. Exactly what set of symptoms was classified as convulsions is hard to determine, and therefore de- clines in this category, while they may be important in accounting for declines in infant and childhood mortality, are particularly hard to interpret. CONCLUSIONS Previous literature had suggested that mortality declines in urban areas in the 19th century were affected by public health measures as well as by other fac- tors, such as changes in levels of living, which affected both rural and urban mortality declines. We have argued that the concept of the public health move- ment in general has little analytic value, and that analysis of the causes of the mortality decline must be done in terms of specific public health practices and their impact on specific causes of death. We have therefore looked at the con- struction of sewers and waterworks in relation to the decline in diarrheal dis- eases, cholera infantum, and typhoid, which were often transmitted through impure water. We have also considered the limited available information on anti- tuberculosis and antidiphtheria campaigns in some cities. Diarrheal diseases, typhoid, tuberculosis, and diphtheria constitute a large proportion of the mortality decline in these cities. Almost 80% of the total change in mortality in these cities is accounted for by these four disease cate- gories. If most of the decline in these diseases could be attributed to public health practices, then the role of the public health movement in the decline in urban mortality would be very large indeed. However, our analysis suggests that the impact of specific aspects of the public health measures on these dis- ease categories is by no means clear. The relationship between mortality rates and the expenditures on and the extent of sewers and waterworks is negative, but these public health measures are only weakly related to mortality rates from diarrheal diseases and typhoid, the disease categories which were most likely to be affected by them. If these measures are related to diseases whose
Public Health Measures and Mortality 53 transmission is unlikely to be affected by sewers and waterworks, then alter- native hypotheses are needed. The possibility that both mortality rates and the construction of sewers and waterworks were affected by some other variable, such as per capita income, cannot be overlooked. There are, however, severe limitations in the existing data on public health practices. Further analysis, care- fully tracing developments in the construction of sewers and waterworks and changes in the death rates in individual cities, is needed. The decline in tuber- culosis appears to be unrelated to public health activities during the 19th cen- tury. Similarly, the decline in diphtheria cannot be unambiguously related to public health campaigns. Again, additional research on programs against these diseases in individual cities is necessary. Our research has shown that the decline in urban mortality was probably somewhat affected by specific public health measures. However, the examina- tion of the declines in specific causes of death has increased our understanding of the limited role of public health measures in the decline in urban mortality. ACKNOWLEDGMENTS Valuable assistance in this research was provided by George Clark, Fran Ryan, and Richard Oliver. We are especially grateful to Richard A. Easterlin for his valuable comments. REFERENCES Crimmins-Gardner, E., and Condran, G. A. (1976). Nineteenth Century American Mortality. Paper presented at the meetings of the Population Association of America, April 1976. Dublin, L. I., Lotka, A. J., and Siegelman, M. (1949). Length of Life. Ronald Press, New York. Duffy, J. (1974). A History of Public Health in New York City, 1866-1966. Russell Sage Foundation, New York. Easterlin, R. A. (1977). Population issues in American economic history: A survey and critique. In Gallman, R. E. (ed.), Recent Developments in the Study of Business and Economic History: Essays in Honor of Herman E. Krooss, Johnson Associates, Greenwich, Connecticut. Gore, J. K. (1904). On the improvement in longevity in the United States during the nine- teenth century. In Proceedings of the Fourth International Congress of Actuaries, Vol. I, The Actuarial Society of America, New York. Meeker, E. (1972). The improving health of the United States, 1850-1915. Explorations in Economic History 9: 353-374. Meeker, E. (n.d.). Public health and medicine. In Dictionary of American History, Charles Scribner's and Sons, forthcoming. McKeown, T., and Record, R. G. (1962). Reasons for the decline of mortality in England and Wales during the nineteenth century. Population Studies 16: 94-122.
54 Condran and Crimmins-Gardner Rosen, G. (1971). Historical trends and future prospects in public health. In Medical His- tory and Medical Care: A Symposium of Perspectives, Oxford University Press, New York, pp. 57-81. Rosenkrantz, B. G. (1972). Public Health and the State: Changing Views in Massachusetts, 1842-1936. Harvard University Press, Cambridge. Smith, W. (1964). Cities of Our Past and Present. New York. Sydenstriker, E. (1972). Health and Environment. Arno Press and The New York Times, New York. Taeuber, C., and Taeuber, I. B. (1958). The Changing Population of the United States. New York. U.S. Bureau of the Census (1904). Twelfth Census, 1900: A Discussion of the Vital Statis- tics. Bulletin 15, U.S. Government Printing Office, Washington, D.C. U.S. Census Office (1896). Eleventh Census, 1890, Vital and Social Statistics. Part II, Vital Statistics. U.S. Government Printing Office, Washington, D.C. U.S. Census Office (1902a). Twelfth Census, 1900: Population, Part II. U.S. Census Office, Washington, D.C. U.S. Census Office (1902b). Twelfth Census, 1900: Vital Statistics, Part I - Analysis and Ratio Tables. U.S. Census Office, Washington, D.C. U.S. Census Office (1902c). Twelfth Census, 1900: Vital Statistics, Part II - Statistics of Death. U.S. Census Office, Washington, D.C. U.S. Department of Labor (1900). Bulletin of the Department of Labor, Statistics of Cities, No. 30. U.S. Department of Labor (1901). Bulletin of the Department of Labor: Statistics of Cities, No. 36. Whipple, G. C. (1908). Typhoid Fever, Its Causation, Transmission, and Prevention. Wiley, New York. Winslow, C. E. A. (1923). The Evolution and Significance of the Modern Public Health Campaign. Yale University Press, New Haven.

No comments:

Post a Comment