In an area with diverse groups of people, it can be particularly helpful to analyze these different groups separately. While exploring the birth data in Wayne County for 2017-2022 from the Michigan Department of Health and Human Services, we decided to look more closely at trends in Detroit. This served as a great case study for the importance of disaggregating data to smaller geographic regions, something D3 advocates for regularly.
The differences in data between cities and even neighborhoods within a city often tell stories that can help direct resources and programming more efficiently. Comparing Detroit to Out-Wayne County (all of Wayne County except the city of Detroit) showed us that Detroit women generally gave birth at younger ages than the rest of the county, a trend that is entirely masked by the volume of births overall in Wayne County.
Figure 1.
Figure 1 shows the number of births each year from 2017 to 2022 in Wayne County, Michigan by the age of the mother (in 5-year groups). Overall, the numbers of births have decreased over this time period for women ages 20-29. Births increased for 30-34-year-old women and for 35-49-year-olds, although to a much smaller extent for the oldest two groups. A strong impression from this figure is that the age groups with the greatest numbers of births are the mid-to-late 20s, followed by the early 30s, and then the early-to-mid 20s.
However, when we look at the numbers for Detroit alone, a different pattern emerges (Figure 2).
Figure 2.
While 25-29-year-olds had the highest concentration of births among Detroit women, the dominance of that age group was not so pronounced as it was for the county as a whole. Moreover, in contrast to the county birth patterns, births to women in Detroit ages 20-24 were greater in numbers than to women 30-34 (until the last two years of data), consistent with a pattern of births at younger ages for Detroit women.
Figure 3.
The birth data in Out-Wayne County present a sharply different profile from the Detroit profile. The most striking difference between the two geographies is the concentration of births to women in the somewhat older ages of 25-29 and 30-34 in Out-Wayne County, as Figure 3 illustrates. This contrasts sharply with births to Detroit women. One should also note the greater number of births to women ages 40-49 than to women ages 15-19 in Wayne County outside of Detroit. There are also some similarities between the two regions; namely, in that the numbers of births to women younger than 30 decreased over the six years and increased for women 30 and older in both Detroit and Wayne County outside of Detroit.
Discussion
These age-related differences in the two different geographies have implications for programs directed to the women living in them. For example, programs addressing the issues facing young, pregnant women would likely be needed more in Detroit than in Out-Wayne County where pregnant women tend to be older on average.
Disaggregating is a process of unhiding; that is, of unhiding the factors that drive behavior. To continue with the birth data example: Let’s say that there is evidence that there are significant numbers of pregnant women who start prenatal care not in the first trimester of pregnancy but in the second trimester or later or not at all. Designing a program to encourage visiting a physician in the first trimester would first need to identify where women live who don’t attend early prenatal appointments. Like an archeologist, the “disaggregator” could drill down into successively deeper layers: county, city, census tract/ZIP code, block group (given enough observations per unit) This step can reveal geographic patterns and point the researcher to other key factors to investigate.
Once the areas of greatest need have been identified, the analyst can examine factors related to the timing of prenatal care. These factors could include characteristics of the women (e.g., educational attainment, marital status, age, race, ethnicity) that are included in the birth files as well as information derived from external databases about, for example, the availability of public and personal transportation, proximity to medical facilities, and income adequacy.
The crucial step, however, remains disaggregating the data to a smaller, more manageable geography for planning a program.
Head over to our State of the Child to our birth and family related data. Can’t find your answer there? AskD3 for free to get started!