Pre-election and primary polls proved accurate enough this time around to leave few big surprises on Election Night. For the most part, we got what we expected and the data reflects that. But the data also reveals some interesting geographic differences between this gubernatorial election and the last one, as well as differences in electoral preferences between white and Black voters.
This analysis by The Baltimore Banner uses the available precinct level data to show changes in voting trend in the last two gubernatorial elections. Full precinct data will be available on Nov. 18.
What changed from 2018 to 2022?
County-level data is great for understanding broad trends, but if we want to dig in to changes happening at the city level, we need precinct data.
These maps show the staggering loss of support for Republican Dan Cox relative to outgoing Republican Gov. Larry Hogan. While it’s important to remember that “land doesn’t vote,” nearly all of Maryland with the exception of the D.C., Baltimore, and their respective suburbs strongly preferred Hogan to his Democratic opponent Ben Jealous in 2018. Data for Montgomery and Kent counties is unavailable.
Map: Nick Thieme Source: Maryland State Board of Elections
This time, not only did Democratic Gov.-elect Wes Moore win an unbroken block of precincts from Waldorf to Cockeysville, but Cox’s support in many rural, generally Republican precincts dropped to not much more than 50%.
We can also see an exaggerated version of the idea that even in generally Republican areas, cities vote blue. Cities like Salisbury and Easton on the Eastern Shore, Annapolis and Aberdeen, and Frederick voted for Moore, with their outskirts and exurbs still preferring Cox.
The only region throughout Maryland where Cox’s level of support remained close to Hogan’s was the western-most tail in Appalachia.
A closer look at results from Baltimore City and Baltimore County illuminates just how overwhelmingly Moore beat Cox.
Jealous won strong support in the majority-Black Baltimore neighborhoods which scholar Lawrence Brown has named the Black Butterfly, but had middling support elsewhere in Baltimore — even losing precincts in the southeastern neighborhoods of the city. In the county, north of the city, he won just six precincts — you can almost see the city line in his support.
Map: Nick Thieme Source: Maryland State Board of Elections
Moore, by contrast, received between 75% and 100% of votes from nearly every precinct inside Baltimore City. He lost just one by a few percentage points.
This support translated into Baltimore County, where Moore expanded support to the western edge of the county and far north of the city. While the “blue horn” extending west from the city has always shown more support for Democrats, this time Moore won between 75% and 100% of the votes in many of its precincts. This region contains the most majority-Black precincts in Baltimore County — many of which went for Hogan in 2018— and only one of which went for Cox.
Even the southeastern part of the county, which borders the Chesapeake Bay and is generally a Republican stronghold, shifted strongly in favor of Moore. While Cox still won the majority of precincts there, he lost vote share relative to Hogan in all but one, and won a few precincts by only a couple percentage points.
Racial preferences for governor
Beyond just geographic trends, precinct-level results let us say things about differences in opinions across different demographics. While individuals’ votes are completely anonymous and inaccessible to the public, meaning we’ll never know an individual voter’s preference or race, precincts are small enough that combining them with census tract-level demographic data from the U.S. Census Bureau gives us access to racial differences in the aggregate.
Each dot below represents one precinct in Maryland. The further to the right the dot, the larger the percentage of Black residents living in that precinct. The height of the dot tells us the percentage of voters in that precinct who voted for Wes Moore. There is an amazingly strong trend that the more Black residents in a tract the larger Moore’s share of the vote — to the point where 100% of residents in many all-Black precincts voted for him. We can see this in the tight grouping of precincts in the top right of the chart.
This relationship breaks down a little, but not entirely, for majority non-Black precincts. At the left of the chart, dots are scattered from top to bottom, meaning that a good deal of largely-white precincts also overwhelmingly preferred Moore to Cox. However, we also see a large cluster of dots to the bottom left of the chart. This tells us that most all-white precincts voted either in favor of Dan Cox or just barely in Moore’s direction.
Broadly, the upward tilt of the entire graph — like a thin triangle — shows us that race was strongly predictive of a precinct’s gubernatorial choice. It’s important to note, though, given how geographically segregated people of different races generally are, that we may just be showing the same graph as before — not actual racial preference, but geography disguised as race. However, even after a Baltimore Banner analysis controlled for geographic location, median income, poverty rates, and overall population, race is still strongly related to Moore’s vote share.
Who wants cannabis to be legal?
Marylanders also voted to legalize marijuana for adults age 21 or older through a ballot measure simultaneous with the gubernatorial election. Using the same census-tract method as before, we can look into who voted for legalization.
This chart is a little more complicated. Again, every dot represents a precinct. The further to the right the dot, the larger the percentage of voters who wanted to legalize marijuana. We separate majority-Black and non-majority-Black precincts into two rows.The vertical location of the dot doesn’t tell us anything — it’s only there to make every precinct dot visible.
While nearly every precinct, majority-Black or non-majority-Black, supported legalizing marijuana, on average, Black precincts showed a larger preference for legalization. 73% of Black Marylanders supported the ballot measure versus 62% of non-Black residents.
The smooth grey arches superimposed over the precinct dots display the concentration of support in its most abstract. A precinct’s support in favor of legalization is most likely to be found where the grey section is widest. As they taper off to thin points, it becomes less likely that a precinct showed that level of support. This bulbous shape is what gives these charts the name “violin plots.”
Through these, we can see that between 65% and 85% of people voting in most majority-Black precincts supported passage, while somewhere between 50% and 80% of voters in most precincts that aren’t majority Black voted for passage.
Taylor DeVille contributed to this story.
nick.thieme@thebaltimorebanner.com
:quality(70)/s3.amazonaws.com/arc-authors/baltimorebanner/b7611c6c-3910-4b12-ba7d-ef76181ff331.png)
Nick Thieme designs statistical experiments and analyzes data to discover and improve stories about inequality, human rights, health care, and climate change. He has worked as a data reporter and statistician for a variety of public and private organizations, with writing appearing in the Atlanta Journal-Constitution, Slate Magazine, and elsewhere.
:format(jpg)/cloudfront-us-east-1.images.arcpublishing.com/baltimorebanner/UTOU6RCPLZEOREZ3XZNRG2ZDKY.jpg)
:format(jpg)/cloudfront-us-east-1.images.arcpublishing.com/baltimorebanner/YO2XC34J6JHOZKTPKVPUWFEYVE.jpg)
:format(jpg)/cloudfront-us-east-1.images.arcpublishing.com/baltimorebanner/5VYMTB6UOFHKXG72LMP42HMWOU.jpg)
:format(jpg)/cloudfront-us-east-1.images.arcpublishing.com/baltimorebanner/7V65VJ4Z2FCCNPYEXDDIMABSPE.JPG)