This table compares Hubbell’s advantage or deficit in the largest counties with margins for Iowa’s only successful Democratic candidates for governor in the past 50 years.

Residents of Polk, Johnson, and Story delivered record victories for a Democratic nominee for governor or president. Even Barack Obama never came out of those counties with such a large advantage.

The Dallas County number is less bad than the trendline suggests. Given the massive growth of wealthy suburbs west of Des Moines, Hubbell did well to keep Reynolds’ margin so low. Republican candidates have carried this county by thousands of votes in the recent past.

Compared to Culver and Vilsack, though, Hubbell underperformed in other urban strongholds, particularly Dubuque, Scott, Woodbury, and Pottawattamie. Unofficial results show Reynolds received about 36,000 more votes statewide than her challenger. Hubbell could have made up a third of that margin by matching previous Democratic nominees’ results in Dubuque and not getting blown out in Woodbury and Pottawattamie.

This table shows the current political make-up of the largest counties, using statistics on active registered voters from the Secretary of State’s office, as of November 1. These figures do not take into account anyone who registered to vote on election day.

In many ways, the results line up as we’d expect. For instance:

• Reynolds carried the three largest counties where Republicans outnumber Democrats.
• Hubbell did best in Johnson, which has the highest percentage of registered Democrats and the lowest percentage of Republicans.
• The share of the vote for Hubbell and Reynolds was comparable in Linn and Black Hawk counties, which have similar proportions of Democrats, Republicans, and no-party voters.

On the other hand, some oddities stand out.

• Hubbell did far better in Story than in Scott, which have comparable partisan breakdowns.
• Hubbell’s share of the vote was higher in Dallas than in Pottawattamie or Woodbury, even though Democrats make up a higher percentage of the electorate in the latter two counties.
• Hubbell underperformed in Dubuque, which has the third-highest percentage of Democrats in the electorate, after Johnson and Polk. We’d expect Polk to be a strong county for a candidate with deep family roots and connections in Des Moines. Even so, the small margin for Hubbell in Dubuque is noticeable and will increase Democratic anxiety that Donald Trump’s victory in this solid blue county was no fluke.

Since voter registration numbers don’t fully explain the disparate election outcomes, I pulled together some other statistics that struck me as potentially relevant. Most of the information in the next table comes from U.S. Census Bureau data as of July 2017. The unemployment column draws on Iowa Workforce Development numbers from September 2018.

We see evidence of the widening partisan split among white voters, depending on education level. The best counties for Hubbell (Johnson and Story) have the highest share of Iowans with a college degree. The high percentage of college-educated Dallas County residents is likely the primary reason Reynolds fared worse there than one would guess base on the GOP voter registration advantage.

Conversely, the best large-population counties for Reynolds were Pottawattamie and Woodbury, which have relatively low numbers of college-educated adults.

Hubbell’s best counties had a younger population on the whole, with a smaller percentage of senior citizens.

In other respects, this table raises more questions than answers for me.

• Why didn’t Hubbell do better in Dubuque? That county has more college-educated voters as well as a higher percentage of registered Democrats than Black Hawk, yet Hubbell carried Black Hawk by 12 points and Dubuque by less than 1.5 points.

  Is the Waterloo area’s greater racial diversity in the main explanatory factor? Or did Reynolds benefit from the national trend of white Catholics “drifting toward the Republican Party”? (I don’t have current statistics on the religious identities of Iowans, but Dubuque is among the most heavily Catholic counties.) Some Hubbell campaign commercials and direct mail pieces highlighted his longstanding support for Planned Parenthood, which could have backfired in this area.

• Linn and Scott counties are similar in terms of racial breakdown, education level, and percentage of senior citizens, so why did Hubbell win by 14 points in Linn but only by a little more than 3 points in Scott?
• Democrats often have better results in more racially diverse areas, yet Hubbell’s vote share better in Woodbury (the least white county on this list) was only a tiny bit higher than in Pottawattamie (the second most white of Iowa’s large urban counties).
• Median income, unemployment, and poverty levels do not appear to be correlated with voting patterns in these counties. Does that mean “it’s not the economy, stupid”?
• Higher-income Americans have historically favored Republican candidates. Will Democrats consolidate their gains among well-off suburbanites in Polk, Dallas, and Linn counties, or will this year be an anomaly because Trump repels so many college-educated voters?

Any relevant questions or insights are welcome in this thread.

Top image: On left, the main Polk County administrative office building in Des Moines. On right, Davenport City Hall in Scott County.