This blog post began as one in a series of investigations about reproductive behavior in Southeast Michigan. In particular, it was looking initially at the separate influences of race and ethnicity on variables such as the number of births, the ages at which women have children, and adequacy of prenatal care. However, the inquiry took a different turn when we focused not on their influence on reproductive behavior but on the intersection of Hispanic identification and racial identification in four areas of Southeast Michigan: Detroit; Wayne County outside of Detroit (Out-Wayne); Oakland County; and Macomb County.

We found, for example, that among Hispanic-identified women in Detroit and Oakland County who gave birth in the years 2017-2022, the percentage identified as White increased over the six years. However, this was not the case in other areas of Southeast Michigan. Was this an indication of differences among the areas in Hispanic self-identification? This blog post lays out the evidence and then discusses it as a cautionary tale.

Our data come from the Michigan Department of Health and Human Services (MDHHS) birth certificate database of all births to women residing in Michigan, and for this study specifically in Wayne, Oakland, and Macomb Counties in 2017-2022.

Measures
Hispanic identification was measured by a binary “yes” “no” response from the birth certificate. Race, also from the birth certificate, was coded into 20 categories, which are listed in the Appendix. We collapsed the 20 categories into five: White, Black, Asian/Pacific Islanders, Multiple Races, and Other Race, grouped as shown in the Appendix. Grouping them into these general categories of similar characteristics simplifies the analysis while still providing relevant results. Other Race is a catchall phrase selected by those who don’t find any of the options provided adequately represent their identity, meaning the races represented in Other Race may be different between the different geographic areas. In the four areas in this study, over 90% of those identifying as Hispanic were identified as White or Other Race.

Results
We started our exploration by asking whether there are differences in ethnicity-race combinations among the four areas. We also wanted to know whether the ethnicity-race combinations changed systematically over the years 2017-2022. The four charts below show only the percentages for those mothers identifying as Hispanic and either White or Other Race.

 

Figure 1.

The chart titled "Percentage of Adequate Prenatal Care by Area of Residence 2017-2022" displays the percentage of adequate prenatal care in four regions: Detroit, Out Wayne County, Oakland County, and Macomb County, over six years. Detroit shows a gradual increase in the percentage of adequate prenatal care from around 50% in 2017 to just above 60% in 2022. Out Wayne County maintains a relatively high percentage around 70% throughout the years. Oakland County consistently shows high percentages, around 80%, with minimal fluctuations. Macomb County also demonstrates high and stable percentages, starting at approximately 75% in 2017 and reaching just a high 85% by 2021. The data is sourced from the Geocoded Michigan Birth Certificate Registry, Division for Vital Records & Health Statistics, Michigan Department of Health & Human Services, and Data Driven Detroit.
Click to view figure 1 as a table
Detroit Hispanic Mothers’ White and Other Race Identification, 2017-2022 2017 2018 2019 2020 2021 2022
Hispanic and White
22% 33% 27% 31% 34% 38%
Hispanic and Other Race
71% 60% 67% 62% 61% 58%

 

In these charts the green bar represents the percentage of mothers identifying as Hispanic who were identified as Other Race. Clearly, Other Race is the predominant category for Detroit, although the percentage identified as White (orange bar) gained on it in the most recent years.

Figure 2.

The chart titled "Percentage of Intermediate Prenatal Care by Area of Residence 2017-2022" depicts the percentage of inadequate prenatal care in four regions: Detroit, Out Wayne County, Oakland County, and Macomb County, over six years (2017 to 2022). Each year is represented by a different color: orange for 2017, green for 2018, yellow for 2019, teal for 2020, blue for 2021, and purple for 2022. Detroit shows the highest percentages of intermediate prenatal care among the regions, with a high of 31% in 2022 and low of 26% in 2022. Out Wayne County has a small variability in percentages, ranging from a low of 19% to a high of 24%. Oakland County maintains the lowest percentages, around 15% throughout the period. Macomb County displays an increase from 19% in 2017 to 25% by 2022. The data sources include the Geocoded Michigan Birth Certificate Registry, Division for Vital Records & Health Statistics, Michigan Department of Health & Human Services, and Data Driven Detroit.
Click to view figure 2 as a table
Out-Wayne County Hispanic Mothers’ White and Other Race Identification, 2017-2022 2017 2018 2019 2020 2021 2022
Hispanic and White
51% 55% 53% 44% 52% 53%
Hispanic and Other Race
42% 40% 41% 47% 43% 41%

 

In Out-Wayne County, about half of the Hispanic mothers identify as White with around 40-43% Other Race.

Figure 3.

The chart titled "Percentage of Inadequate Prenatal Care by Area of Residence 2017-2022" depicts the percentage of inadequate prenatal care in four regions: Detroit, Out Wayne County, Oakland County, and Macomb County, over six years (2017 to 2022). Each year is represented by a different color: orange for 2017, green for 2018, yellow for 2019, teal for 2020, blue for 2021, and purple for 2022. Detroit shows the highest percentages of inadequate prenatal care among the regions, with a decreasing trend from 21% in 2017 to 12% in 2022. Out Wayne County consistently shows low percentages, around 8% across the years. Oakland County also maintains very low percentages, around 5% throughout the period. Macomb County displays a noticeable decrease from around 13% in 2017 to below 4% by 2022. The data sources include the Geocoded Michigan Birth Certificate Registry, Division for Vital Records & Health Statistics, Michigan Department of Health & Human Services, and Data Driven Detroit.
Click to view figure 3 as a table
Oakland County Hispanic Mothers’ White and Other Race Identification, 2017-2022 2017 2018 2019 2020 2021 2022
Hispanic and White
57% 57% 58% 65% 71% 69%
Hispanic and Other Race
39% 39% 36% 29% 24% 26%

 

In Oakland County, more Hispanic mothers were identified as White than as Other Race, with the gap growing from 18 percentage points in 2017 to 47 points in 2021.

Figure 4.

The chart titled "Percentage of Inadequate Prenatal Care by Area of Residence 2017-2022" depicts the percentage of inadequate prenatal care in four regions: Detroit, Out Wayne County, Oakland County, and Macomb County, over six years (2017 to 2022). Each year is represented by a different color: orange for 2017, green for 2018, yellow for 2019, teal for 2020, blue for 2021, and purple for 2022. Detroit shows the highest percentages of inadequate prenatal care among the regions, with a decreasing trend from 21% in 2017 to 12% in 2022. Out Wayne County consistently shows low percentages, around 8% across the years. Oakland County also maintains very low percentages, around 5% throughout the period. Macomb County displays a noticeable decrease from around 13% in 2017 to below 4% by 2022. The data sources include the Geocoded Michigan Birth Certificate Registry, Division for Vital Records & Health Statistics, Michigan Department of Health & Human Services, and Data Driven Detroit.
Click to view figure 4 as a table
Macomb County Hispanic Mothers’ White and Other Race Identification, 2017-2022 2017 2018 2019 2020 2021 2022
Hispanic and White
64% 68% 58% 69% 69% 73%
Hispanic and Other Race
29% 26% 35% 19% 24% 19%

 

Of the four areas, Macomb County’s percentages varied more than those of the other three areas. Except for 2019, mothers identifying as Hispanic also identified as White ranged from 64% to 73%.

To answer the first question of whether the four areas differ in the combination of race and ethnicity, the data clearly show differences. Within each of the four areas have the ethnicity-race combinations changed systematically over the years 2017-2022? The answer appears to be “yes” for Detroit and Oakland County. In both geographies, the percentage of birthing mothers identifying as Hispanic who were also identified as White generally increased as the percentage of Other Race declined. In Detroit, the percentage of White rose from a low of 22% in 2017 to 38% in 2022. In Oakland County, it rose from 57% in 2017 to 71% in 2021 before declining to 66% in 2022.

Discussion

While it’s tempting to trumpet these findings, emphasizing those of Detroit and Oakland County, as insights into Hispanic identity, there are multiple reasons to be cautious.

  1. The data from birth certificates were not collected to address the question of Hispanic identity or for research purposes but for administrative purposes. Hospitals, where most babies in the United States are born, have staff whose responsibility it is to gather information from the parents that they don’t have from other sources. Race is a particularly slippery concept. Do the 20 categories of race as coded in the birth certificate data accurately reflect individuals’ conceptions of their race? How often is race decided not by parents but by hospital staff? Are race and ethnicity even distinct concepts? Without certainty around these questions and collection methods, we should be cautious when using these data.
    To that point, the Census Bureau has collapsed race and ethnicity into a single question for the 2030 census instead of asking about Hispanic or Latino origin as a separate ethnicity question. Additionally, the Census and Office of Management and Budget have approved adding a new race-ethnicity category labeled Middle Eastern or North African (MENA), with a deadline for the specifics set for 2027. This category would join the other major categories of White; Black or African American; Hispanic or Latino; Asian; American Indian or Alaska Native; and Native Hawaiian or Pacific Islander. According to the Office of Management and Budget, its engagement with the MENA and Hispanic/Latino communities has shown “the need for demographic and socioeconomic statistics about its population to inform policy decisions, health research, civil rights monitoring and enforcement, and many other needs.”
  2. While the race-ethnicity combinations differed among the four geographical areas we cannot assume that the combinations within each area are fixed and static or that the Hispanic population in Southeast Michigan is static. What has been happening with that population? Is it growing and if so, where? Are people identifying as Hispanic moving from Detroit to Out-Wayne County, Oakland County, and Macomb County? Are trends different for younger and older people who identify as Hispanic?

    Related: A Guide to Conscious Data Consumption

  3. Disaggregating (or breaking down) geographies may reveal important differences. People identifying as Hispanic in northern Oakland County may differ from those identifying as Hispanic in southern Oakland County.
    Related: Gaining Insight by disaggregating geographic areas
  4. The data for this study covered the years 2017-2022. The trends in Detroit and Oakland County looked fairly convincing, but that was only six years of data, including three years of the COVID pandemic. Will the trends hold up when we add the data from 2012-2016? This is an important question that will be answered in a subsequent blog post.

More significantly, it is easy to jump from finding intriguing ethnicity-race combinations to concluding that we have unveiled something fundamental regarding Hispanic identity. Maybe there is some truth in that, but an individual’s identity is multifaceted, and this data from MDHHS is not enough for us to draw conclusions from. The two variables of ethnicity (Hispanic/non-Hispanic) and race are not measured consistently across studies, as illustrated by the evolving Census categories in the first point above, nor across time as illustrated below; and the data from birth certificates do not allow us to make such a sweeping conclusion. However, the data reviewed here could be a starting point for a more comprehensive understanding of Hispanic identity.

Race as a Changing Social Construct

Race is a social construct, and our understanding of it changes over time—both at a societal level and an individual level—and we change how we measure it to reflect that. The Census Bureau has used many different identifiers since the first census in 1790, and a corollary to the current “Hispanic, Latino or Spanish Origin” first showed up in 1930 as “Mexican.” This identifier was removed in the 1940 census, to reappear in 1970 as “Origin or Descent: Mexican; Puerto Rican; Cuban; Central or South American; Other Spanish,” and it has remained on the Census since then in similar variations. Hispanic is not the only identity with a complex history—as the chart below demonstrates—even those that started out with a simple classification by the Census often get more complex over time as the Census responds to evolving views of racial and ethnic identities. These complexities always need to be considered when we are looking at race/ethnicity data over time.

As the classifications change over time in response to both power dynamics and societal norms, the consequences of the classifications for individuals and communities also change. While the racial categories used by MDHHS and the Census are helpful for us to do a data analysis on subsets of the population to understand how different communities experience life in different ways, we should remember that it’s the impossibility of a perfect survey question that perfectly represents the complexity of the people represented in the data.

Related: Data Isn’t Color Blind

A bar chart titled "US Census Race and Ethnicity Across the Decades: 1790-2030" displaying the changing race and ethnicity categories used by the US Census in each decade from 1790 to 2030. The chart starts with only three categories: Slaves, Free White Females & Males, and All Other Free Persons. Variations of those categories remain until All Other Free Persons are dropped in 1850, and Indian and Chinese are added in 1860. The White category remains fairly simple over time, while each other category gets additions and changes consistently over decades. Mexican is added only for 1930, and then added back with more options in 1970. Hawaiian is added in 1960, and more specificity added most decades through 2030. Middle Eastern is the most recent category that will be added in 2030. Data sources include: Gibson, Campbell, and Kay Jung. 2002. “Historical Census Statistics on Population By Race, 1790 to 1990, and By Hispanic Origin, 1790 to 1990, For The United States, Regions, Divisions, and States.” Humes, Karen, and Howard Hogan. 2009. “Measurement of Race and Ethnicity in a Changing, Multicultural America.” Humes, Karen R., Nicholas A. Jones, and Roberto R. Ramirez. 2011. “Overview of Race and Hispanic Origin: 2010.” Office of Management and Budget. 1978. “Statistical directive no. 15: Race and ethnic standards for federal agencies and administrative reporting.” Office of Management and Budget. 1997. “Revisions to the standards for the classification of federal data on race and ethnicity.” U.S. Census Bureau History Questionnaires. (2014, March 31).

Appendix

Racial Categories used by MDHHS from the birth certificate

1 = WHITE
2 = BLACK
3 = AMERICAN INDIAN
4 = ASIAN INDIAN
5 = CHINESE
6 = FILIPINO
7 = JAPANESE
8 = KOREAN
9 = VIETNAMESE
10 = OTHER ASIAN
11 = NATIVE HAWAIIAN
12 = GUAMAN,CHAMORRO
13 = SAMOAN
14 = OTHER PACIFIC ISLANDER
15 = OTHER RACE
21 = WHITE MULTIPLE RACE
22 = BLACK MULTIPLE RACE
23 = AMERICAN INDIAN MULTIPLE RACE
24 = ASIAN/PACIFIC ISLANDER
MULTIPLE RACE 99 = UNKNOWN

These were grouped as follows:
Category 1 = White
Category 2 = Black
Categories 4-14 = Asian/Pacific Islander
Category 3 and 15 = Other Race
Categories 21-24 = Multiple Race

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!

Copyright © 2022 Data Driven Detroit. All Rights Reserved.
TERMS OF USEPRIVACY POLICYCONTACT US