Death And Spending: Did Urban Mortality Shocks Lead To Municipal Expenditure Increases?

Louis P. Cain, Loyola University of Chicago and Northwestern University and Elyce J. Rotella, Indiana University

In an earlier paper we demonstrated that municipal expenditures on water delivery and treatment, sewers, and refuse collection dramatically reduced mortality in US cities from typhoid, dysentery, and diarrhea in the period 1899-1929. We estimated that a 1% increase in sanitation expenditures was associated with an annual saving of 27 lives in the average American city. Having established that expenditure increases caused mortality decline, we investigate here the reverse causation, i.e., whether changes in municipal expenditures on sanitation were motivated by mortality shocks. Such a supposition is consistent with anecdotal evidence.

In the summer of 1879 heavy rains caused the Chicago River to discharge into Lake Michigan for thirty consecutive days, polluting the city's water supply. Under normal conditions, the sewage-laden river would have discharged into the Illinois and Michigan Canal away from the city's Lake Michigan water supply. The Citizens' Association of Chicago sponsored a committee "to devise a plan to dispose of the sewage of the city without contaminating the city water supply." Nothing came of this as a series of dry seasons removed the impetus for further action. Then, in August 1885, the city was deluged by an unprecedented storm. "And a Flood Came" read a Chicago Tribune headline after more than 5 1/2 inches of the rain had fallen on the city in 19 hours. Not only did the sewers prove totally inadequate, but the Des Plaines River overflowed into the canal, the canal into the Chicago River, and vast amounts of filth were carried into Lake Michigan. The Chicago Daily News reported, "The rainfall ... is... carrying out filth unspeakable and polluting the water far beyond the crib. This is what the majority of the people of Chicago will have to drink for days to come." This episode caused a substantial increase in sickness and mortality leading to calls for government action.

Among those calling for action was the Citizens' Association committee which released a report proclaiming that the South Branch was "in an abominable condition of filth beyond the power of the pen to describe." This report "amplified and urged" the committee's earlier recommendation (a new, larger canal with dimensions comparable to those of the Chicago River), but added a new twist by proposing a study by "a commission of experts" to put "a stop for all time to the unsanitary condition which then existed." This report, as well as pressure from Chicago's leading citizens, forced Chicago's City Council to create a Drainage and Water Supply Commission in the spring of 1886. Their preliminary report, issued in January 1887, essentially concurred with the new canal proposal. The Sanitary District Enabling Act of May 29, 1889, was a direct result of the Drainage and Water Supply Commission's recommendations. This story, and others that could be told from other cities, have contributed to the argument that, as the demand for urban infrastructure increased dramatically in late 19th century, municipalities responded positively only when faced with a crisis.

The most commonly held theory of municipal budgeting is incrementalism which argues that the best predictor of this year's allocation for a department is last year's allocation. The incrementalist approach implies that municipal expenditure patterns do not in general respond to specific conditions or to public pressures. As MacDonald and Ward argue in the introduction to their collection The Politics of Urban Fiscal Policy, "the local public sector was routinely both insulated from the crises generated by socioeconomic development and autonomous from the demands of specific groups in the urban environment." This relationship was broken only when there was a dramatic crisis, but not by every such crisis. During an acknowledged epidemic in Albany, city fathers rationalized a lack of response by noting that their new filtration works had halved the death rate in comparison to the pre-filtration days and that the epidemic was widespread, affecting more cities than Albany.

In this paper we make use of mortality and expenditure data for 50 large cities to test whether shocks do matter for municipal expenditures. Annual data on mortality and municipal expenditures were collected for the period 1899-1929. The sample was defined to include all cities having a municipal water supply and populations over 100,000 in the 1920 Census and for which we could obtain nearly complete data on mortality experience and sanitation expenditures.

The mortality data used in this study were collected from the Mortality Statistics of Cities which annually published death-by-cause statistics. We have constructed a waterborne death rate (WDR) series that includes deaths attributable to typhoid fever, diarrhea, and dysentery. In this paper we follow the convention of referring to this group of diseases as "waterborne," even though water is not the exclusive means of transmission. Typhoid, dysentery, and diarrheal diseases were spread by impure water and food, and by contact with feces and other filth. We expect, as did contemporaries, that these diseases were controlled by programs to deliver clean water and to remove and treat waste water. Only a small minority of cities with 1920 populations over 100,000 did not have municipal waterworks and had to be excluded from our sample. While historical evidence on death-by-cause is notoriously problematic because of changing definitions of diseases and changes in diagnoses, those studied in this paper were well identified in this period.

The data on municipal sanitation expenditures used here were published in various bulletins to 1903 and in Financial Statistics of Cities beginning in 1905. We use data on annual operating costs and capital acquisition costs of water and sewage works. Not every series was reported every year. Few direct figures are available for 1904, and Financial Statistics of Cities was published in 1913, 1914, or 1920. For the regressions discussed below, missing observations were filled in by interpolation.

We ran regressions which attempted to explain current expenditures as a function of lagged expenditures and various measures of mortality shocks in the current period and the recent past. Separate analyses were done for water and sewer expenditures, for operating and capital expenditures, and for all sanitation expenditures together. Regressions were run for each of 30 cities and for all cities together. This produced a total of 210 city-level regressions and 63 multiple city regressions.

Did cities respond to shocks from an increase in relevant death rates by increasing their expenditures? Unfortunately, the regression results give us little insight into the answer to this question. The mortality shock variable was statistically significant in only 20 of the city-level regressions and in none of the multiple city regressions.

We do not, on the basis of these disappointing regression results, reject the hypothesis that cities responded to mortality shocks by changing their expenditures. Many examples of such responses appear in the historical literature, however it appears that nonresponse was also common. What we conclude from the statistical analysis is that cities did not regularly and systematically exhibit expenditure responses to mortality shocks.

At this point, the more interesting question is: what determines whether cities will respond to a mortality shock or not? We begin the search for an answer by taking a straightforward counting approach in which we count the number of mortality shocks that were closely followed by a notable expenditure increase. We defined a mortality shock year as one in which the actual waterborne death rate is more than one standard error above its trend for 1899-1929. Using that definition, there are 187 mortality shocks with heavy concentrations in the years 1906-10, quite a few in the war years, and almost none in the 1920s. We defined a "response" to be a situation in which a shock in the expenditure series (i.e., an expenditure more than one standard error above trend) occurred within three years of a shock in the mortality series. Once a response was identified, we turned to the Engineering News to see if we could identify the event and determine the actual nature of the expenditure.

The waterborne death rate shocks were grouped into 104 episodes (some lasting more than one year). In 32 of these episodes, there was no response from the affected city. Otherwise, an above trend expenditure occurred within three years. We therefore count responses in 72 (69.2%) of the 104 mortality episodes. Of these, 13 are cases in which the only response recorded is in the same year as the shock.

Many of these episodes could be found in the Engineering News. Without going into detail, the following headlines are all tied to episodes identified by the counting process. Each is illustrative of a response:

"The Typhoid Fever Epidemic at Columbus, O." (2-11-04)
"Typhoid Fever History and the Water Supply of Pittsburg, Allegheny and Vicinity" (2-25-04)
"A Filtration Plant for Pittsburg, PA." (3-3-04)
"The Sewage Testing Station at Columbus, O." (10-20-04)
"Progress on the Sewage Works and Water Softening and Purification Works of Columbus, Ohio" (9-21-05)

Through the Engineering News we are also able to identify failures to respond and, in some cases, the reasons why there was no response. An example is the failure of Albany to respond to an increase in the typhoid death rate in 1901. A story about Albany's new water filtration plant, then the largest in the country, appeared in the January 14, 1900 issue. In the August 9, 1900 issue an article appeared giving the typhoid death rates before and after filtration. The rate averaged 85 per annum for the period 1890-98. The data for the first eight months of 1899 are similar, but the total number of deaths for the last four months dropped from 24 to seven. This suggests the rate would be about 25 per annum after the opening of the filtration plant. A third article appeared in the June 27, 1901 issue, a WDR shock year in our data, reporting that Albany experienced 39 typhoid-related deaths through the first 11 months of the plant's operation, a rate that was unexpectedly high. The article noted that no response was forthcoming for two reasons: first, the State Board of Health reported that typhoid was unusually prevalent in the state that autumn, and, second, physicians reported 14 of the 39 cases were contracted elsewhere.

We know that cities did respond to many mortality crises by increasing their sanitation expenditures even though the results of our regression analyses tells us that such responses were far from universal. We are currently examining the history of specific mortality crises in an attempt to discover the factors which governed whether and how cities responded.