With the 2016 presidential primaries less than 12 months away, campaign season is well under way for candidates of both parties. In the last election, President Barack Obama won with about 51% of the popular vote, winning 26 states and the District of Columbia to capture 332 of 538 electoral votes. Despite the Democratic victory, an analysis of the reddest and bluest counties in each state shows stark differences between them.

Based on voting data compiled by political news organization Politico, 24/7 Wall St. measured the political leanings of county residents. Several measures were used to identify the most Republican-leaning county in each state: The percentage of county residents voting for Republican candidate Mitt Romney in 2012 and the county’s chosen representatives to the House and the Senate.. These are the reddest (and bluest) counties in each state.

Click here to see the most Republican county in each state.

Click here to see the most Democratic county in each state.

In general, people living in the reddest counties tended to be better-off financially compared to democratic-leaning areas. The reddest counties in 29 states had a median household income greater than the comparable national income of $53,046. By contrast, the median household income was higher than the national figure in only 18 of the 50 bluest counties.

While counties dominated by Republicans tended to be wealthier, residents in these areas also tended to have lower educational attainment rates compared to the bluest counties. In the five years through 2013, the percentage of adults who had attained at least a bachelor’s degree was lower than the comparable national rate of 28.8% in 36 of the 50 reddest counties. This was the case in just 17 of the bluest counties.

The racial composition of Republican-leaning counties was also significant. The vast majority of residents in most of the reddest counties identify as white. The proportions of the white populations in only two of the reddest counties were lower than the 74% of Americans who identify as white. On the other hand, 33 of the bluest counties had smaller shares of white populations compared to the national average composition.

The reddest counties almost always had the highest share of votes for the Republican presidential candidate compared to other counties in the state, and those shares were almost always a majority. However in five of the reddest counties — Honolulu county, Hawaii; Lincoln County, Maine; Plymouth County, Massachusetts; Washington County Rhode Island; and Essex County, Vermont — a majority of residents in voted to re-elect President Obama. In these counties, Romney still captured among the largest shares of votes compared to other counties in the state.

However, some of the reddest counties are not as red as those in other states. Carroll County, Maryland, for example, has sent Democratic Rep. Chris Van Hollen to Congress since 2002 yet in 2012 gave 65.7% of its vote to Romney. Rep. Collin Peterson (D-MN) was first elected to Congress from Pipestone County, Minnesota in 1990 and re-elected every two years since, but that county gave 61% of its presidential vote to Romney.

To identify the most Republican county in each state, 24/7 Wall St. reviewed the percentage of eligible counties casting votes for Mitt Romney in 2012 from Politico, a political news organization. Eligible counties were those that cast more than 2,000 total votes. Additionally, we created an index of each county’s Congressional representation in both the House of Representatives and the Senate since 2007. The highest values of our index were given to counties where Congressional representation was dominated by Republicans. The reddest county in each state is the product of our Congressional index and the share of a county’s votes cast for Mitt Romney. We also reviewed the percentage of county residents who had attained at least a bachelor’s degree, median household incomes, and the share of residents identifying as white — all from the U.S. Census Bureau’s 2013 American Community Survey five-year estimates.