Eight top presidential vote shares (from 1836 on) in every state

(Every state has now been won by better than 1.25:1 at least eight times, with Alaska having been so won the minimum number of times.)

Alabama
Franklin Roosevelt, 1936: 86.38%
Franklin Roosevelt, 1940: 85.22%
Franklin Roosevelt, 1932: 84.74%
Franklin Roosevelt, 1944: 81.28%
Strom Thurmond, 1948: 79.75%
Woodrow Wilson, 1916: 76.04%
Alton Parker, 1904: 73.35%
Richard Nixon, 1972: 72.43%

Alaska
Ronald Reagan, 1984: 66.65%
Lyndon Johnson, 1964: 65.91%
George W. Bush, 2004: 61.07%
George H. W. Bush, 1988: 59.59%
John McCain, 2008: 59.42%
George W. Bush, 2000: 58.62%
Richard Nixon, 1972: 58.13%
Gerald Ford, 1976: 57.90%

Arizona
Franklin Roosevelt, 1936: 69.85%
Franklin Roosevelt, 1932: 67.03%
Ronald Reagan, 1984: 66.42%
Franklin Roosevelt, 1940: 63.49%
Richard Nixon, 1972: 61.64%
Dwight Eisenhower, 1956: 60.99%
Ronald Reagan, 1980: 60.61%
George H. W. Bush, 1988: 59.95%

Arkansas
Franklin Roosevelt, 1932: 85.96%
Franklin Roosevelt, 1936: 81.80%
Franklin Roosevelt, 1940: 79.02%
William Jennings Bryan, 1896: 73.72%
Franklin Roosevelt, 1944: 69.95%
Richard Nixon, 1972: 68.82%
James Buchanan, 1856: 67.12%
Woodrow Wilson, 1916: 65.97%

California
Franklin Roosevelt, 1936: 66.95%
Warren Harding, 1920: 66.20%
Herbert Hoover, 1928: 64.69%
Joe Biden, 2020: 63.48%
Theodore Roosevelt, 1904: 61.84%
Hillary Clinton, 2016: 61.73%
Barack Obama, 2008: 61.01%
Barack Obama, 2012: 60.24%

Colorado
William Jennings Bryan, 1896: 84.95%
Herbert Hoover, 1928: 64.72%
Ronald Reagan, 1984: 63.44%
Richard Nixon, 1972: 62.61%
Lyndon Johnson, 1964: 61.27%
Woodrow Wilson, 1916: 60.74%
Franklin Roosevelt, 1936: 60.37%
Dwight Eisenhower, 1952: 60.27%

Connecticut
Lyndon Johnson, 1964: 67.81%
Dwight Eisenhower, 1956: 63.72%
William McKinley, 1896: 63.24%
Warren Harding, 1920: 62.72%
Calvin Coolidge, 1924: 61.54%
Ronald Reagan, 1984: 60.73%
Barack Obama, 2008: 60.59%
William Howard Taft, 1908: 59.43%

Delaware
Herbert Hoover, 1928: 65.03%
Barack Obama, 2008: 61.94%
Lyndon Johnson, 1964: 60.95%
Ronald Reagan, 1984: 59.78%
Richard Nixon, 1972: 59.60%
Horatio Seymour, 1868: 59.00%
Joe Biden, 2020: 58.74%
Barack Obama, 2012: 58.61%

Florida
Grover Cleveland, 1892: 85.01%
Franklin Roosevelt, 1936: 76.10%
Franklin Roosevelt, 1932: 74.68%
Franklin Roosevelt, 1940: 74.01%
Richard Nixon, 1972: 71.91%
William Jennings Bryan, 1900: 71.31%
William Jennings Bryan, 1896: 70.46%
Franklin Roosevelt, 1944: 70.32%

Georgia
Franklin Roosevelt, 1932: 91.60%
Franklin Roosevelt, 1936: 87.10%
Franklin Roosevelt, 1940: 84.85%
Franklin Roosevelt, 1944: 81.74%
Woodrow Wilson, 1916: 79.51%
Woodrow Wilson, 1912: 76.63%
Richard Nixon, 1972: 75.04%
John Davis, 1924: 73.96%

Hawaii
Lyndon Johnson, 1964: 78.76%
Barack Obama, 2008: 71.85%
Barack Obama, 2012: 70.55%
Joe Biden, 2020: 63.73%
Richard Nixon, 1972: 62.48%
Hillary Clinton, 2016: 62.22%
Hubert Humphrey, 1968: 59.83%
Bill Clinton, 1996: 56.93%

Idaho
William Jennings Bryan, 1896: 78.10%
Ronald Reagan, 1984: 72.36%
George W. Bush, 2004: 68.38%
George W. Bush, 2000: 67.17%
Ronald Reagan, 1980: 66.46%
Theodore Roosevelt, 1904: 65.84%
Warren Harding, 1920: 65.60%
Dwight Eisenhower, 1952: 65.42%

Illinois
Warren Harding, 1920: 67.81%
Barack Obama, 2008: 61.92%
Dwight Eisenhower, 1956: 59.52%
Lyndon Johnson, 1964: 59.47%
Richard Nixon, 1972: 59.03%
Calvin Coolidge, 1924: 58.84%
Theodore Roosevelt, 1904: 58.77%
Franklin Roosevelt, 1936: 57.70%

Indiana
Richard Nixon, 1972: 66.11%
Ronald Reagan, 1984: 61.67%
George W. Bush, 2004: 59.94%
Dwight Eisenhower, 1956: 59.90%
George H. W. Bush, 1988: 59.84%
Herbert Hoover, 1928: 59.68%
Dwight Eisenhower, 1952: 58.11%
Donald Trump, 2020: 57.02%

Iowa
Warren Harding, 1920: 70.91%
Abraham Lincoln, 1864: 64.12%
Dwight Eisenhower, 1952: 63.75%
Theodore Roosevelt, 1904: 63.39%
Ulysses Grant, 1868: 61.92%
Lyndon Johnson, 1964: 61.88%
Herbert Hoover, 1928: 61.77%
Ulysses Grant, 1872: 60.81%

Kansas
Abraham Lincoln, 1864: 79.19%
Herbert Hoover, 1928: 72.02%
Ulysses Grant, 1868: 68.82%
Dwight Eisenhower, 1952: 68.77%
Richard Nixon, 1972: 67.66%
Ulysses Grant, 1872: 66.46%
Ronald Reagan, 1984: 66.27%
Dwight Eisenhower, 1956: 65.44%

Kentucky
Horatio Seymour, 1868: 74.55%
George McClellan, 1864: 69.83%
William Henry Harrison, 1840: 64.20%
Lyndon Johnson, 1964: 64.01%
Richard Nixon, 1972: 63.37%
Donald Trump, 2016: 62.52%
Donald Trump, 2020: 62.09%
Samuel Tilden, 1876: 61.41%

Louisiana
Franklin Roosevelt, 1932: 92.79%
Franklin Roosevelt, 1936: 88.82%
Alton Parker, 1904: 88.50%
Woodrow Wilson, 1916: 85.90%
Franklin Roosevelt, 1940: 85.88%
William Jennings Bryan, 1908: 84.63%
Franklin Roosevelt, 1944: 80.59%
William Jennings Bryan, 1900: 79.03%

Maine
Calvin Coolidge, 1924: 72.03%
Dwight Eisenhower, 1956: 70.87%
Warren Harding, 1920: 68.92%
Lyndon Johnson, 1964: 68.84%
Herbert Hoover, 1928: 68.63%
William McKinley, 1896: 67.90%
Ulysses Grant, 1872: 67.86%
Theodore Roosevelt, 1904: 67.44%

Maryland
Horatio Seymour, 1868: 67.20%
Lyndon Johnson, 1964: 65.47%
Joe Biden, 2020: 65.36%
Franklin Roosevelt, 1936: 62.35%
Barack Obama, 2012: 61.97%
Barack Obama, 2008: 61.92%
Franklin Roosevelt, 1932: 61.50%
Richard Nixon, 1972: 61.26%

Massachusetts
Lyndon Johnson, 1964: 76.19%
Abraham Lincoln, 1864: 72.20%
Ulysses Grant, 1868: 69.76%
William McKinley, 1896: 69.47%
Ulysses Grant, 1872: 69.20%
Warren Harding, 1920: 68.55%
Joe Biden, 2020: 65.60%
John Frémont, 1856: 63.61%

(John Kerry is 12th, after Humphrey, Lincoln '60, and Coolidge.)

Michigan
Calvin Coolidge, 1924: 75.37%
Warren Harding, 1920: 72.76%
Herbert Hoover, 1928: 70.36%
Theodore Roosevelt, 1904: 69.51%
Lyndon Johnson, 1964: 66.70%
Ulysses Grant, 1872: 62.66%
William Howard Taft, 1908: 61.93%
Ronald Reagan, 1984: 59.23%

Minnesota
Theodore Roosevelt, 1904: 73.98%
Warren Harding, 1920: 70.59%
Lyndon Johnson, 1964: 63.76%
Abraham Lincoln, 1860: 63.53%
James Garfield, 1880: 62.28%
Franklin Roosevelt, 1936: 61.84%
Ulysses Grant, 1872: 61.27%
Ulysses Grant, 1868: 60.88%

Mississippi
Franklin Roosevelt, 1936: 97.06%
Franklin Roosevelt, 1932: 95.98%
Franklin Roosevelt, 1940: 95.70%
Franklin Roosevelt, 1944: 93.56%
Woodrow Wilson, 1916: 92.78%
Alton Parker, 1904: 91.07%
William Jennings Bryan, 1896: 91.04%
William Jennings Bryan, 1908: 90.11%

Missouri
Abraham Lincoln, 1864: 69.72%
Lyndon Johnson, 1964: 64.05%
Franklin Roosevelt, 1932: 63.69%
Richard Nixon, 1972: 62.29%
Franklin Roosevelt, 1936: 60.76%
Ronald Reagan, 1984: 60.02%
Martin van Buren, 1836: 59.98%
Harry Truman, 1948: 58.11%

Montana
William Jennings Bryan, 1896: 79.93%
Franklin Roosevelt, 1936: 69.28%
Warren Harding, 1920: 61.13%
Ronald Reagan, 1984: 60.47%
Dwight Eisenhower, 1952: 59.39%
George W. Bush, 2004: 59.07%
Lyndon Johnson, 1964: 58.95%
Franklin Roosevelt, 1932: 58.80%

Nebraska
Ulysses Grant, 1872: 70.68%
Ronald Reagan, 1984: 70.55%
Richard Nixon, 1972: 70.50%
Dwight Eisenhower, 1952: 69.15%
George W. Bush, 2004: 65.90%
Ronald Reagan, 1980: 65.53%
Dwight Eisenhower, 1956: 65.51%
Rutherford Hayes, 1876: 64.70%

Nevada
William Jennings Bryan, 1896: 81.21%
Franklin Roosevelt, 1936: 72.81%
Franklin Roosevelt, 1932: 69.41%
James Weaver, 1892: 66.78%
Ronald Reagan, 1984: 65.85%
Richard Nixon, 1972: 63.68%
Ronald Reagan, 1980: 62.54%
William Jennings Bryan, 1900: 62.25%

New Hampshire
Martin van Buren, 1836: 75.01%
Ronald Reagan, 1984: 68.659%
William McKinley, 1896: 68.655%
Dwight Eisenhower, 1956: 66.11%
Richard Nixon, 1972: 63.98%
Lyndon Johnson, 1964: 63.89%
George H. W. Bush, 1988: 62.49%
Dwight Eisenhower, 1952: 60.92%

New Jersey
Warren Harding, 1920: 67.65%
Lyndon Johnson, 1964: 65.61%
Dwight Eisenhower, 1956: 64.68%
Calvin Coolidge, 1924: 62.17%
Richard Nixon, 1972: 61.57%
Ronald Reagan, 1984: 60.09%
Herbert Hoover, 1928: 59.77%
William McKinley, 1896: 59.68%

New Mexico
Franklin Roosevelt, 1932: 62.72%
Franklin Roosevelt, 1936: 62.69%
Richard Nixon, 1972: 61.05%
Ronald Reagan, 1984: 59.70%
Lyndon Johnson, 1964: 59.22%
Dwight Eisenhower, 1956: 57.81%
Barack Obama, 2008: 56.91%
Franklin Roosevelt, 1940: 56.59%

New York
Lyndon Johnson, 1964: 68.56%
Warren Harding, 1920: 64.56%
Barack Obama, 2012: 63.35%
Barack Obama, 2008: 62.88%
Dwight Eisenhower, 1956: 61.19%
Joe Biden, 2020: 60.86%
Al Gore, 2000: 60.21%
Bill Clinton, 1996: 59.47%

(John Kerry is 12th, after Hillary Clinton, Franklin Roosevelt '36, and Nixon '72. Immediately after the 2004 election, he would have been 8th. McKinley '96 is 13th.)

North Carolina
Franklin Roosevelt, 1940: 74.03%
Franklin Roosevelt, 1936: 73.40%
Franklin Roosevelt, 1932: 69.83%
Richard Nixon, 1972: 69.46%
Franklin Roosevelt, 1944: 66.71%
Ronald Reagan, 1984: 61.90%
Alton Parker, 1904: 59.71%
Woodrow Wilson, 1912: 59.24%

North Dakota
Warren Harding, 1920: 77.79%
Theodore Roosevelt, 1904: 75.12%
Dwight Eisenhower, 1952: 70.97%
Franklin Roosevelt, 1932: 69.59%
Donald Trump, 2020: 65.11%
Ronald Reagan, 1984: 64.84%
Ronald Reagan, 1980: 64.23%
Donald Trump, 2016: 62.96%

Ohio
Herbert Hoover, 1928: 64.89%
Lyndon Johnson, 1964: 62.94%
Dwight Eisenhower, 1956: 61.11%
Theodore Roosevelt, 1904: 59.75%
Richard Nixon, 1972: 59.63%
Ronald Reagan, 1984: 58.90%
Warren Harding, 1920: 58.47%
Calvin Coolidge, 1924: 58.33%

Oklahoma
Richard Nixon, 1972: 73.70%
Franklin Roosevelt, 1932: 73.30%
Ronald Reagan, 1984: 68.61%
Franklin Roosevelt, 1936: 66.83%
Mitt Romney, 2012: 66.77%
John McCain, 2008: 65.65%
George W. Bush, 2004: 65.57%
Donald Trump, 2020: 65.37%

Oregon
Theodore Roosevelt, 1904: 67.06%
Franklin Roosevelt, 1936: 64.42%
Herbert Hoover, 1928: 64.18%
Lyndon Johnson, 1964: 63.72%
Dwight Eisenhower, 1952: 60.54%
Warren Harding, 1920: 60.20%
Ulysses Grant, 1872: 58.66%
Franklin Roosevelt, 1932: 57.99%

Pennsylvania
Theodore Roosevelt, 1904: 68.00%
Warren Harding, 1920: 65.76%
Calvin Coolidge, 1924: 65.34%
Herbert Hoover, 1928: 65.24%
Lyndon Johnson, 1964: 64.92%
Ulysses Grant, 1872: 62.07%
William McKinley, 1896: 60.98%
William McKinley, 1900: 60.74%

Rhode Island
Lyndon Johnson, 1964: 80.87%
Ulysses Grant, 1872: 71.94%
William McKinley, 1896: 68.33%
Ulysses Grant, 1868: 66.49%
Hubert Humphrey, 1968: 64.03%
Warren Harding, 1920: 63.97%
John Kennedy, 1960: 63.63%
Barack Obama, 2008: 62.86%

South Carolina
Franklin Roosevelt, 1936: 98.57%
Franklin Roosevelt, 1932: 98.03%
Woodrow Wilson, 1916: 96.71%
John Davis, 1924: 96.56%
James Cox, 1920: 96.05%
Woodrow Wilson, 1912: 95.94%
Franklin Roosevelt, 1940: 95.63%
Alton Parker, 1904: 95.36%

South Dakota
Theodore Roosevelt, 1904: 71.09%
Dwight Eisenhower, 1952: 69.27%
Franklin Roosevelt, 1932: 63.62%
Ronald Reagan, 1984: 63.00%
Donald Trump, 2020: 61.77%
Donald Trump, 2016: 61.53%
Warren Harding, 1920: 60.74%
Ronald Reagan, 1980: 60.53%

Tennessee
Franklin Roosevelt, 1936: 68.85%
Ulysses Grant, 1868: 68.43%
Richard Nixon, 1972: 67.70%
Franklin Roosevelt, 1940: 67.25%
Franklin Roosevelt, 1932: 66.49%
Donald Trump, 2016: 60.72%
Donald Trump, 2020: 60.66%
Franklin Roosevelt, 1944: 60.45%

Texas
Franklin Roosevelt, 1932: 88.06%
Franklin Roosevelt, 1936: 87.08%
Franklin Roosevelt, 1940: 80.92%
Woodrow Wilson, 1916: 76.92%
John Breckinridge, 1860: 75.47%
William Jennings Bryan, 1908: 73.97%
John Davis, 1924: 73.70%
Franklin Pierce, 1852: 73.07%

(Nixon '72 is 17th, after eight other Democrats: Wilson '12, Parker, Roosevelt '44, Cass, Tilden, Cleveland '84, Bryan '96, and Buchanan.)

Utah
William Jennings Bryan, 1896: 82.70%
Ronald Reagan, 1984: 74.50%
Mitt Romney, 2012: 72.79%
Ronald Reagan, 1980: 72.77%
George W. Bush, 2004: 71.54%
Franklin Roosevelt, 1936: 69.34%
Richard Nixon, 1972: 67.64%
George W. Bush, 2000: 66.83%

Vermont
William McKinley, 1896: 80.08%
Ulysses Grant, 1868: 78.57%
Ulysses Grant, 1872: 78.29%
Calvin Coolidge, 1924: 78.22%
Theodore Roosevelt, 1904: 77.97%
John Frémont, 1856: 77.96%
Abraham Lincoln, 1864: 76.10%
Abraham Lincoln, 1860: 75.86%

Virginia
Franklin Roosevelt, 1936: 70.23%
Franklin Roosevelt, 1932: 68.46%
Franklin Roosevelt, 1940: 68.08%
Richard Nixon, 1972: 67.84%
Woodrow Wilson, 1916: 66.99%
Woodrow Wilson, 1912: 65.95%
John Davis, 1924: 62.48%
Franklin Roosevelt, 1944: 62.36%

Washington
Theodore Roosevelt, 1904: 69.95%
Herbert Hoover, 1928: 67.06%
Franklin Roosevelt, 1936: 66.38%
Lyndon Johnson, 1964: 61.97%
Franklin Roosevelt, 1940: 58.22%
Joe Biden, 2020: 57.97%
William Howard Taft, 1908: 57.68%
Barack Obama, 2008: 57.65%

West Virginia
Donald Trump, 2020: 68.62%
Donald Trump, 2016: 68.50%
Abraham Lincoln, 1864: 68.24%
Lyndon Johnson, 1964: 67.94%
Richard Nixon, 1972: 63.61%
Mitt Romney, 2012: 62.30%
Franklin Roosevelt, 1936: 60.56%
Ulysses Grant, 1868: 58.83%

Wisconsin
Warren Harding, 1920: 71.10%
Franklin Roosevelt, 1936: 63.80%
Franklin Roosevelt, 1932: 63.46%
Theodore Roosevelt, 1904: 63.21%
Lyndon Johnson, 1964: 62.09%
Dwight Eisenhower, 1956: 61.58%
Dwight Eisenhower, 1952: 60.95%
William McKinley, 1900: 60.06%

Wyoming
Ronald Reagan, 1984: 70.51%
Donald Trump, 2020: 69.94%
Richard Nixon, 1972: 69.01%
George W. Bush, 2004: 68.86%
Mitt Romney, 2012: 68.64%
Donald Trump, 2016: 68.17%
George W. Bush, 2000: 67.76%
Theodore Roosevelt, 1904: 66.72%

District of Columbia
Barack Obama, 2008: 92.46%
Joe Biden, 2020: 92.15%
Barack Obama, 2012: 90.91%
Hillary Clinton, 2016: 90.86%
John Kerry, 2004: 89.18%
Lyndon Johnson, 1964: 85.50%
Walter Mondale, 1984: 85.38%
Bill Clinton, 1996: 85.19%

---

There is nothing magical about the eight top slots. As I mentioned, eight was the largest number of top vote-share earners we could list in every state and still ensure that all eight, in every state, had won a 'landslide' (whether that is defined the way I have defined it, or whether it is defined as 55%--#9 in Alaska is Mitt Romney, who got 54.80%). That number will tend to get larger over time, of course. 

However, it provides for a larger sample size of the top vote shares in every state than listing just the single top vote share. 

Nixon makes the top eight vote share-getters in every Southern state except South Carolina, Louisiana, Misssissippi, and Texas. (Goldwater, who did better than Nixon in Mississippi, also does not make the top eight in that state.) This establishes something of a delineation between the three specific Southern states that were consistently, or frequently, nearly unanimous in their support for the Democracy from 1896 through 1944; and the others. (In Texas, the outlier, it seems that there was a combination of factors that led to Nixon not making, or even coming particularly close to making, the top eight. For one thing, there were unusually strong antebellum Democratic performances; Texas was the only Southern state in which Nixon failed to outdo every antebellum Democrat from 1836 through 1860. For another, Texas was Nixon's second-weakest state in the Old Confederacy, ahead of just Louisiana. Apparently, Texas was home to the only congressional districts that McGovern won anywhere in the Old Confederacy [he won two in the state].) In every Southern state save Delaware, Nixon in '72 outdid Reagan in '84. In six Southern states, Nixon is the only Republican to have made the top eight (shown alongside the four in which no Republican has done so).

One interesting phenomenon is that every major party nominee from 1996 on, except for Dole and Kerry, is responsible for one of the top eight vote shares in at least one state. Kerry still is in the District of Columbia, and would have set the eighth-highest vote share ever in New York immediately after the 2004 election (but every subsequent Democratic nominee has outdone him in the Empire State). Gore is currently seventh in New York. Bill Clinton in '96 is in the top eight in two states; George W. Bush in 2000, in four states; George W. Bush in 2004, in eight states; Barack Obama in 2008, in ten states; John McCain, in two states; Barack Obama in 2012, in five states; Mitt Romney, in four states; Donald Trump in 2016, in six states; Hillary Clinton, in two states; Joe Biden in 2020, in seven states; and Donald Trump in 2020, in eight states.

Every state that is in red or blue on any of those maps, voted for that party in 2020. (The most recently that a nominee of the opposite party to the one that carried it in 2020 set a top-eight vote share in a state was 1988, when Bush Sr set a top-eight vote share in Arizona and New Hampshire--as always, unless otherwise specified, 'top eight' means, as of after the 2020 election.) However, the difference between Bush's top-eight states and Trump's is interesting. All four of the Bush's '00 top-eight states are states that haven't voted Democratic since 1964, and, at the time, seven of his eight '04 to-eight states were (all except Montana, which voted for Clinton in '92; now six of eight are, as Obama carried Indiana in 2008). Three of the '00 ones are Bryan-only-Dem-loser states; as are five of the '04 ones (and in any case, the most recently any of them voted for a losing Democrat was 1924).

Bush's top-eight states (all of his top-eight states from '00 also became top-eight states for him in '04) are predominantly in the Mountain West. The only one that could be said to be in the South is Oklahoma (where he entered the top eight only in '04). The only Bush '00 top-eight state in which Trump broke into the top eight was Wyoming. Of Bush's '04 top-eight states, he did so--in addition to Wyoming--only in Oklahoma and Indiana, in both of which his vote share is currently #8 (and lower than Bush's in '04).  

But Trump, of course, had his own top-eight states that were not shared with Bush. Of Trump's '16 top-eight states (all of which went on to be top-eight states for him in '20, along with Oklahoma and Indiana), three are last-voted-Dem-in-'64 states (the Dakotas and Wyoming); and two are Bryan-only-Dem-loser states (South Dakota and Wyoming). However, three are Southern states that voted for Bill Clinton twice: Tennessee, Kentucky, and West Virginia. West Virginia last voted for a losing Democrat in 1988, and before that had also voted for losing Democrats in 1980, 1968, and 1952. Kentucky last voted for a losing Democrat in 1952. Tennessee last voted for a losing Democrat the same year as Oklahoma--1924--but, whereas Oklahoma had last voted for a losing Republican in 1996 (as of the Bush years), and before that in 1992 and 1976, Tennessee's last three votes for losing Republicans, as of the Trump years, were in 2012, 2008, and 1960. (As of the Bush years, Oklahoma had voted for losing Republicans four times--1996, 1992, 1976, and 1960--whereas Tennessee had done so only once--in 1960.)

Even though Trump set another top-eight record in 2020 in Kentucky and Tennessee, his vote share was lower than in 2016, so his 2016 vote share will stay in the top eight longer than his 2020 vote share for those states. Bush improved on his '00 vote share in every state in which his '04 vote share is still amongst the top eight.

Trump's top eight definitely shows the shift of the centre of gravity of the GOP to the 'Scots-Irish zone'. At the same time, however, it shows the shift of the centre of gravity of the GOP away from the Mountain West and 'Nebransas' (if you will). 

The Democrats, on the other hand, seem to have intensified everywhere where they were particularly strong from the '90s on, whilst adding to the places where they are particularly strong. (The last Democrat to come close to setting top-eight records in states where the party has since demonstrably weakened would be Carter. Carter's '76 vote share in Arkansas is the ninth-highest ever in that state--although that's what it would have been immediately after the '76 election as well. And, immediately after the '76 election, his vote share in Tennesseee [which was above 55.5%] would have been the tenth-highest ever; it is now 17th, after Tilden, Romney, White, Bush Sr '88, Reagan '84, McCain, Bush '04, and Wilson '16. In West Virginia, Carter's '76 vote share actually was in the top eight--it was the seventh-highest immediately after the 1976 election [after Hoover's in '28], and remained so until 2012. It stayed in the top eight until 2016.) 

This, perhaps, explains why Kerry, uniquely out of major party nominees in the last six elections, does not hold a top-eight record in any state (although he did for four years). If Trump had improved still further on Bush in the states in which Bush set a top-eight record, he wouldn't have driven Bush's vote shares in most of them out of the top eight altogether, but he would have sped it up. Instead, Bush seems locked in, for example, in the top eight in Montana, Idaho, Utah, Nebraska, and Alaska (and Indiana) indefinitely. 

Romney's top-eight states show him to have been a transitional figure. One was a 'Scots-Irish zone' state where Bush never set a top-eight record, and which had voted for a losing Democrat as recently as 1988 (West Virginia). One is the Scots-Irish zone state where Bush did set a top-eight record in 2004 (and the only one where Trump has not bettered on Bush) (Oklahoma). One is the only Mountain West, never-Dem-since-'64 state in which Trump improved on Bush (Wyoming). And one is Utah. It should, perhaps, have been seen as somewhat alarming at the time that Romney was unable to break into the top eight in more traditionally Republican Bush-top-eight states, such as Idaho, Nebraska, or Montana (not to mention his alarming decline relative even to McCain in Alaska).

Biden holds one of the top eight vote share records in Massachusetts, which is an anomaly. No other Democrat in the last seven elections has set a top-eight record in the state. (Even Kerry--whose vote share in the state is now 12th, as it is in New York--would not have been in the top eight in the state immediately after the '04 election.) Every other state in which any of the last seven Democratic nominees holds a top-eight vote share, is one in which at least one of the other last seven Democratic nominees also does, except four for Obama in '08, in what was a nationally decisive victory (Illinois, New Mexico, Connecticut, and Rhode Island).

Speaking of the static nature of Carter's record in Arkansas, of the six states that voted twice for Bill Clinton, while Trump entered the top eight in three (Tennessee, Kentucky, and West Virginia), he did not in the other three (Arkansas, Louisiana, and Missouri). In Louisiana, this would have been unrealistic; but in Arkansas, he failed to even knock on the door of the top eight vote shares by ever doing any better than Carter had done in the state (in fact, Trump's 2020 vote share in Arkansas--the better of his two in that state--is only the 13th-highest ever, behind Carter's in '76, van Buren's in 1836, Bryan's in 1900, and Polk's in 1844). In Missouri, the lowest of the top eight vote shares is under 60% (favourite-son Harry Truman's 58.11%), so this should not have seemed unbeatable. But his 2020 vote share here is just the 12th-highest ever, after Truman's, Tilden's in 1876 (57.64%), Polk's in 1844 (56.98%), and Grant's in 1868 (56.96%). His 2020 vote share was additionally behind van Buren's in 1840 and Pierce's in 1852, although that's more understandable (Romney hadn't won a landslide in the state, and McCain had barely carried it in 2008).

Arkansas and Missouri are, along with Oklahoma (where, again, Trump has never managed to do better than Bush in '04), Tennessee, Kentucky, and West Virginia, 'Scots-Irish zone' states where entering the top eight is plausible (and has, after all, been done recently, since the dissolution of the Democratic Solid South). (Barone doesn't count Missouri as a 'Scots-Irish zone' state, to be clear.) If Trump was going to oversee (or continue to oversee, given Romney's performance in most of them) a permanent weakening of the Republican strength in traditional Republican states in the Mountain West and 'Nebransas', one might have thought he should more thoroughly demonstrate an increase in the party's strength across the Scots-Irish zone. Or, more concretely, that he should have been able to do better than Carter in Arkansas and better than Truman in Missouri.

Trump did enter the top eight vote shares ever in states worth about as many electoral votes as Bush did (in their respective re-election years, 50 vs. 43, according to the 2024-2028 Electoral College). But Bush's opponents, Gore and Kerry, had vote shares that are currently in the top eight in one and zero states, respectively (although, again, Kerry's was in the top eight in one state at the time). The one state was very large--New York--which likely largely explains Bush's difficulty in the national popular vote. But Hillary Clinton was in the top eight in the largest state, California (as well as Hawaii), and Biden's vote share is currently in the top eight in two of the four largest states (California and New York) and several others besides, some of which are not exactly small. Altogether, Biden's vote share is in the top eight in 122 electoral votes' worth of states. 

---

There are some states where a marked majority of the top eight record vote shares are from the last seven elections. This is a map where the darkest shading indicates states where at least four of the top eight vote share records from 1836 on have been set in the last seven elections; the medium shading indicates states where between one and three of the top eight vote share records from 1836 on have been set in the last seven elections; and the lightest shading indicates states where none of the top eight vote share records from 1836 on have been set in the last seven elections (with the colour being that of the party that has set the majority of vote share records in the state). Here we see a problem for the GOP. The only two states in the darkest shade for the GOP are Oklahoma and Wyoming, indicating, as mentioned above, the divide that occurred between the Bush and Trump years. There are also 'only' three states in the darkest shade for the Democrats, but two of those three are California and New York. As for states in the intermediate shading, each party has about the same number of electoral votes' worth of these states: 71 for the Democrats, 62 for the Republicans. 

The most lightly shaded states are predominantly in the Midwest and the South. The Midwest is easy to explain, since, for the most part, those are not states where either party massively runs up the score these days. 

In the case of the South, it might seem to be a case of Democratic support in the Solid South era having been so great as to produce as many as eight vote shares that no contemporary nominee could possibly hope to top. In which case it might seem unfair to include the South in the same way as the rest of the country. 

As discussed above, however, it was eminently within reach for Trump to break into the top eight in Missouri and Arkansas. The lowest of the top eight vote shares in North Carolina (59.24%) and Virginia (62.36%) are also eminently within reach of topping for the modern political parties.

In the remaining seven states (which correspond to the pre-Fort Sumter Confederacy), the lowest of the top eight vote shares is greater than 2.25:1. Along with Vermont, they are the only states in which the lowest of the top eight vote shares is greater than 2.25:1. The only times that a major party nominee has won a state by better than 2.25:1 (against his or her combined opposition), in the last six elections, are Bush in Utah in 2004, Obama in Hawaii in 2008, Obama in Hawaii and Romney in Utah in 2012, and Trump in Wyoming in 2020. (Conversely, at least one nominee has won multiple states by better than 2:1 in each of the last six elections: Bush in 2000, Bush in 2004, Obama in 2008, Romney in 2012, Trump in 2016, and Trump in 2020.) In seven states (one of which is Vermont), the lowest of the top eight vote shares is greater than Obama's vote share in Hawaii in '08 (the highest vote share achieved by a Democrat in a state in the last six elections, or indeed since 1964). In six states (again, including Vermont), the lowest of the top eight vote shares is higher than Romney's in Utah (which is the highest vote share achieved by any nominee in any state in the last six elections, or since 1984). 

Is there a way of 'normalising' these states, to provide a more meaningful compilation? Well, as can be seen here, in the cases of Alabama, Florida, Louisiana, and Mississippi, there is a rather sudden point at which Democrats (or nominees of any party) go from never having won the state by better than 2.25:1 (at least, from 1836 on), to regularly (or frequently) doing so. 

In Alabama, the top vote share from 1836 to 1860 was Buchanan's 62.08%; and the top vote share from 1836 to 1900, Cleveland's 67.00% in 1888. Then, in 1904, 1908, 1912, and 1916, Democrats won the state by better than 2.25:1 in four straight elections. In Florida, the top vote share from statehood to 1860 was Breckinridge's 62.23%; the top vote share from the end of the War of Sections to 1888 was Cleveland's 59.48% in 1888; and then Democrats proceeded to win Florida by better than 2.25:1 in three straight elections from 1892 to 1900 (with Wilson also doing so in 1912 and 1916). In Louisiana, the highest vote share from 1836 to 1860 was W. H. Harrison's 59.73% in 1840; and--if we exclude Seymour in 1868--the highest vote share from 1836 to 1884 was Hancock's 62.48% in 1880. Democrats then proceeded to win Louisiana by better than 2.25:1 in every election from 1888 to 1944. And in Mississippi, the highest vote share from 1836 to 1860 was Pierce's 60.50%; the highest from 1836 to 1884, Tilden's 68.08%; and Democrats proceeded to carry the state by better than 2.25:1 in every election from 1888 to 1944. (Tilden came pretty close to winning Mississippi by better than 2.25:1, but in Mississippi's case, we can also say that Democrats won Mississippi in every election from 1888 to 1944 by better than 2.75:1. Also, there was another rather sharp juncture, in which Democrats won the state by better than 4:1 in every election from 1896 to 1944.)

Georgia is slightly more complicated. The highest vote share from 1836 to 1860 was Pierce's 64.70%. However, Tilden was the first nominee to win it by better than 2.25:1, in 1876. But this was not followed up by Democrats consistently doing so. Cleveland did in 1888, but aside from that, Democrats wouldn't again until 1912 (which was, however, the beginning of a four-election streak of doing so, from 1912 to 1924). In other elections of this period, however, Democrats would fail to win Georgia by even 2:1 in 1880, 1884, 1892, 1896, 1904, and 1908; and by even 1.5:1 in 1892, 1896, and 1908.

In Texas--which, in its early days (when it was, of course, still a small state) seemed the closest budding analogue to Vermont for the Democrats--two regular antebellum Democrats did win Texas by better than 2.25:1: Cass and Pierce (along with Breckinridge, of course). However, from 1872 to 1900, the only Democrats to win Texas by better than 2.25:1 were Tilden and Cleveland in 1884 (and, again, there were multiple instances in which Democrats failed to win the state by 2:1 or even by 1.5:1). And from 1904 to 1916, Democrats won Texas by better than 2.25:1 in four straight elections (the same four as in Alabama). So there seemed to be a change in Texas between 1900 and 1904. 

South Carolina didn't have an antebellum popular vote. In 1872, it voted 75.73% for Grant--the Republican. Aside from that, from 1868 to 1880, the best vote share anyone got in South Carolina was Hancock's 65.70% (short of 2:1). Then, from 1884 to 1944, Democrats carried South Carolina by better than 3:1 in every election.

Except in Texas (and to a minor degree in Florida), this led to a situation where, as of October 1932, the top eight vote shares already all belonged exclusively to postbellum Democrats. (In Texas, as discussed above, and only in Texas, Nixon in '72 was unable to top at least one antebellum Democrat's vote share. In Florida, Breckinridge's '60 vote share was #8 in October 1932.) There were probably other states of which this was also true (heading into either 1932 or, in the case of historically Republican states, 1952), but it wasn't the norm--even today, James Buchanan's 1856 vote share is #7 in Arkansas, for example; and Frémont's in 1856, #8 in Massachusetts and #6 in Vermont (and, of course, W. H. Harrison's in 1840, #3 in Kentucky). In any case, on the eve of the War of Sections, these states wouldn't necessarily have seemed to have been headed in the same political trajectory. In Louisiana, for example (antebellum America's bellwether state, voting for the winner in every election from 1828 through 1856), the two top antebellum vote shares belonged to Whigs, and no Democrat had gotten so much as 52%. Whigs and Democrats had both done well in Florida and Georgia, but Mississippi and Alabama (and, of course, Texas) seemed much more staunchly Democratic even before the War of Sections.  

What we could do, is to simply excise the first and last elections for each state bracketing the period in which it generally voted for Democrats by better than 2.25:1 (and all the elections in between). Using my judgement, I would say: 1904-1948 for Alabama, 1892-1944 for Florida, 1888-1948 for Mississippi, 1912-1944 for Georgia, 1904-1944 for Texas, 1888-1944 for Louisiana, and 1884-1948 for South Carolina (and also omitting 1872 for South Carolina, 1868 for Louisiana, and 1964 for Mississippi). Except in the case of Texas, all of the eight top vote shares (with the sole exception, in some cases, of Nixon '72) for each state are from those respective timeframes.

If we did that, then the top eight vote shares for these states out of the remaining elections are as follows:

Alabama
Richard Nixon, 1972: 72.43%
Barry Goldwater, 1964: 69.45%
Grover Cleveland, 1888: 67.00%
William Jennings Bryan, 1896: 66.96%
George Wallace, 1968: 65.86%
Adlai Stevenson, 1952: 64.55%
George W. Bush, 2004: 62.46%
Donald Trump, 2016: 62.083%

Florida
Richard Nixon, 1972: 71.41%
Ronald Reagan, 1984: 65.32%
John Breckinridge, 1860: 62.23%
George H. W. Bush, 1988: 60.87%
Franklin Pierce, 1852: 60.03%
Grover Cleveland, 1888: 59.48%
Dwight Eisenhower, 1956: 57.27%
Zachary Taylor, 1848: 57.20%

Georgia
Richard Nixon, 1972: 75.04%
Samuel Tilden, 1876: 72.03%
Grover Cleveland, 1888: 70.31%
Adlai Stevenson, 1952: 69.66%
William Jennings Bryan, 1900: 66.86%
Jimmy Carter, 1976: 66.74%
Adlai Stevenson, 1956: 66.48%
Grover Cleveland, 1884: 65.92%

Louisiana
Richard Nixon, 1972: 65.32%
Winfield Scott Hancock, 1880: 62.48%
Ronald Reagan, 1984: 60.77%
William Henry Harrison, 1840: 59.73%
John McCain, 2008: 58.56%
Donald Trump, 2020: 58.46%
Donald Trump, 2016: 58.09%
Mitt Romney, 2012: 57.78%

Mississippi
Richard Nixon, 1972: 78.20%
Samuel Tilden, 1876: 68.08%
Winfield Scott Hancock, 1880: 65.08%
Grover Cleveland, 1884: 64.34%
Ulysses Grant, 1872: 63.48%
George Wallace, 1968: 63.46%
Ronald Reagan, 1984: 61.85%
Franklin Pierce, 1852: 60.50%

South Carolina
Richard Nixon, 1972: 70.58%
Winfield Scott Hancock, 1880: 65.70%
Ronald Reagan, 1984: 63.55%
George H. W. Bush, 1988: 61.50%
Barry Goldwater, 1964: 58.89%
George W. Bush, 2004: 57.98%
Ulysses Grant, 1868: 57.93%
George W. Bush, 2000: 56.84%

Texas
John Breckinridge, 1860: 75.47%
Franklin Pierce, 1852: 73.07%
Lewis Cass, 1848: 70.29%
Samuel Tilden, 1876: 70.04%
Grover Cleveland, 1884: 69.26%
William Jennings Bryan, 1896: 68.00%
James Buchanan, 1856: 66.59%
Richard Nixon, 1972: 66.20%

This isn't the cleanest 'normalisation' ever. Importantly, it is surely not really correct to simply dismiss the elections in the selected timeframes altogether. It is implausible, for example, that, disregarding whatever mechanisms kept Democratic vote shares preternaturally steadily high in the selected timeframes, Adlai Stevenson was more popular in, say, Georgia, than Franklin Roosevelt was. (And, after all, it is surely no coincidence that it is Franklin Roosevelt specifically, and not any random late-19th/early-to-mid-20th-century Democrat, who holds the record all-time high vote share in every former Confederate state save Florida.) Ideally, one would be able to 'translate' the Democratic vote shares in the selected timeframes into more 'normal' ones, instead of simply discarding them. But this is beyond my capacities.

But this does serve to show that the artifice introduced into these particular states by the mechanisms of the Solid South era really 'cheated' recent Republicans only in Alabama and Louisiana (the latter of which is the only state out of the above that voted for Bill Clinton twice). Out of the above seven states, a Republican from the last six elections shows up in the top eight only in Alabama, Louisiana, and South Carolina. In South Carolina, it is only George W. Bush, and all of Nixon in '72, Reagan in '84, and Bush Sr in '88 are ahead of him. It is also the case in Florida (as can be deduced from the above), Mississippi, and Georgia that all of Nixon in '72, Reagan in '84, and Bush Sr in '88 did better than any Republican in the last six elections. In Louisiana, Texas, and Alabama, a recent Republican has done better than Bush Sr in '88; and in Alabama, alone out of these seven states, a recent Republican has done better than Reagan in '84 as well (placing its electoral behaviour more in line with the 'Scots-Irish zone', with which Michael Barone once grouped it in 'Greater Appalachia', than with the Deep South).

Until 2016, Texas was a tremendous big-state success story for the recent Republican Party. In three of the four preceding elections (all save 2008), its nominee had exceeded Bush Sr's '88 vote share in the state. None of them made the 'normalised' top eight vote shares (and this is even leaving aside whatever FDR's 'true' vote share in the state would have looked like). But Texas is one of the few states that went from being very small (at least, from 1836 on), to being very large (it had just four electoral votes in the 1848-1860 elections, all of which are responsible for one of the top eight 'normalised' vote shares in the state). In general, one is likelier to find something approaching consensus in a smaller state than in a larger state. If we ignore the pre-Fort Sumter Confederacy, there are only four states in which the lowest of the top eight vote shares is greater than 2/3, and only five further states in which the lowest of the top eight vote shares is greater than 63.63% (which corresponds to winning a state by better than 1.75:1); the most electoral votes any of the nine total states has ever had is 11 (which Oklahoma had in the 1932-40 Electoral College). 

However, of course, Trump was well below Bush Sr's '88 vote share (and also well below McCain's vote share) in Texas in both of his elections.

In Louisiana, every Republican in the last five elections (2004-2020) has exceeded Bush Sr's '88 vote share (it was, admittedly, Bush Sr's worst former Confederate state). However, none has managed to overtake Reagan in '84, or William Henry Harrison in 1840; and it has never become part of the Republican 'fornia' this century. Of the states that voted twice for Bill Clinton and have never voted for his party again, Louisiana and the erstwhile bellwether state of Missouri are the only ones that have never become part of the recent Republican 'fornia', and the only ones in which recent Republicans have never broken 60% of the vote.

In Alabama, every Republican in the last five elections, save McCain, has exceeded Reagan's vote share in '84. That said, the high-water mark for the GOP, out of the last six elections, is Bush in '04. In every other 'Greater Appalachia' state except Oklahoma--and in every state that Bill Clinton carried twice and that has voted Republican in every election since (including Louisiana)--Trump did better than Bush in '04 in both of his elections (see here). Indeed, Alabama is part of a group of red states in which neither Romney nor Trump has ever managed to outdo Bush in '04 (again, see here)--a behaviour that puts it decidedly in contrast with the rest of 'Greater Appalachia', and more in line with states whose fading redness has attracted attention, such as Kansas, Alaska, South Carolina, and, of course, Texas.

---

There is also bad news for Democrats, though. Right now, there are five states where all eight of the top vote shares are Republican. There are none where all of them are Democratic. There have been substantially fewer Democratic landslides than Republican, but this still indicates that, rather than the parties 'switching' bases, the Republicans were able to largely keep much of their old base even as they absorbed much of the old Democratic base. (Of those five states, four voted for Trump in 2020--and not only voted for him, but gave him over 55.5%. The fifth is Vermont.)

A lot of the top eight vote shares in a lot of the states are recent, so they don't necessarily indicate states' historic tendencies. Here, I have redone this, for all the states where any of the top eight vote shares is from the last seven elections, as it would have looked as of October 1996 (except for Alaska and Hawaii). Redone in this way, there were only three states where all eight of the top vote shares belonged to Republicans as of October 1996 (Vermont, Nebraska, and Kansas), but still none where all eight belonged to Democrats.

But if we look at the states where, as of October 1996, one party was responsible for, say, at least seven of the eight top vote shares, we would get a map like this. Bush carried every state that is blue on that map in both 2000 and 2004. Of the states that are red on it, he carried eight twice, lost six twice, and carried one in '04 but not in '00.

To some degree, this does represent a bleeding of the historic Republican base. But Republicans had set a preponderance of the top eight vote shares in many more states than the Democrats had--to the point that they appeared as the natural party of government. Democrats could hardly hope to win a an election without carrying some of the states in which Republicans had set an overwhelming preponderance of the top eight vote shares. 

This becomes even more apparent if we make a map where we colour in all the states where, as of October 1996, one party was responsible for at least six of the top eight vote shares--still a fairly overwhelming preponderance (especially keeping in mind that, by October 1996, there had been three 20th-century Democratic national landslides). In blue, we now have the Old Confederacy minus Tennessee--and that's it. In red, we have almost the entire Midwest and entire Northeast, including New York. Now, both Gore and Kerry did carry a majority of the electoral votes in the states that in red on this map--but, given that it almost seems the default in the Midwest and Northeast for states to be in red on this map, it seems more important to look at how the blue states voted. (Also, Bush still did keep a good chunk of the red electoral votes.) Well, again, of the blue states, Bush carried all of them twice. So at a minimum, the Republican was keeping a fair chunk of his party's historic base--the Democrats were keeping none of theirs.

Of course, in 2020, Biden carried (of the states that are blue on the second map) Virginia and Georgia. North Carolina is now a battleground state; Florida remains contestable, and Texas is becoming contestable. But since this is coming after two Bush sweeps of the blue states on that map in close elections, it is clear that Democrats are winning--or contesting--these states for different reasons than they did in the past. 

(Incidentally, in finding the top eight vote shares in states as they would have appeared in October 1996, I found that, after 2000, Bush would have held a top-eight vote share in seven states; after 2004, in 11 states; and Romney after 2012, in six states. These show both more continuity betweeen the Bush and Trump eras, as well as giving Bush [and Romney] more credit for breaking into the 'Scots-Irish zone'. All seven states after 2000 were states which had last voted Democratic in 1964; four were Bryan-only-Dem-loser states. But one was Oklahoma, and one was South Dakota. The 11 states after 2004 were mostly last-voted-Dem-in-'64 states [plus Montana], but, this time, did include a Scots-Irish zone state that had voted for Bill Clinton twice [and that had last voted for a losing Democrat in 1952]: Kentucky. [They also included North Dakota.] And, immediately after the 2012 election, Romney's would have been amongst the top eight vote shares in the same three Clinton-twice Scots-Irish zone states as Trump's would be: Tennessee, Kentucky, and West Virginia.)

Now, conversely, over the last seven elections, the Democrats have turned California from a state where three of the top eight vote shares were Democratic, into one where five are; Maryland, from a state where five were, into one where seven are; and most strikingly, New York, from one where two were, into one where six are. (Now, it would be blue on a map on which states in which six of the eight top vote shares have been set by one party are coloured in.) So, clearly, Democrats are creating a new base, but when Republicans created their new base, they had to forgo much less of their old base to do so. The Jayhawker state has remained reliably red throughout the Southern Strategy period.

---

These are the states ordered by the size of the differential between the two top vote shares:

New Mexico: 0.03%
West Virginia: 0.12%
Nebraska: 0.13%
Oklahoma: 0.40%
Tennessee: 0.42%
South Carolina: 0.54%
Wyoming: 0.57%
North Carolina: 0.63%
Alaska: 0.74%
California: 0.75%
Texas: 0.98%
Mississippi: 1.08%
Alabama: 1.16%
Maine: 1.16%
Vermont: 1.51%
Maryland: 1.73%
Virginia: 1.77%
South Dakota: 1.82%
Ohio: 1.95%
New Jersey: 2.04%
Pennsylvania: 2.24%
Michigan: 2.61%
Oregon: 2.64%
North Dakota: 2.67%
Arizona: 2.82%
Washington: 2.89%
Delaware: 3.09%
Minnesota: 3.39%
Louisiana: 3.97%
Massachusetts: 3.99%
New York: 4.00%
Connecticut: 4.09%
Arkansas: 4.16%
Indiana: 4.44%
Georgia: 4.50%
Kentucky: 4.72%
Missouri: 5.67%
Idaho: 5.74%
Illinois: 5.89%
New Hampshire: 6.35%
Iowa: 6.79%
Hawaii: 6.91%
Kansas: 7.17%
Wisconsin: 7.30%
Utah: 8.20%
Nevada: 8.40%
Florida: 8.91%
Rhode Island: 8.93%
Montana: 10.65%
Colorado: 20.23%

And by the size of the differential between the first and eighth vote shares:

South Carolina: 3.21%
Wyoming: 3.79%
Vermont: 4.22%
Maine: 4.59%
Maryland: 5.94%
Nebraska: 5.98%
New Mexico: 6.13%
Delaware: 6.42%
Ohio: 6.56%
California: 6.71%
Mississippi: 6.95%
Pennsylvania: 7.26%
Virginia: 7.87%
New Jersey: 7.97%
Oklahoma: 8.33%
Connecticut: 8.38%
Tennessee: 8.40%
Alaska: 8.75%
Oregon: 9.07%
Indiana: 9.09%
New York: 9.09%
West Virginia: 9.79%
Arizona: 9.90%
Iowa: 10.10%
Illinois: 10.11%
South Dakota: 10.56%
Wisconsin: 11.04%
Missouri: 11.61%
Washington: 12.30%
Massachusetts: 12.58%
Idaho: 12.68%
Minnesota: 13.10%
Kentucky: 13.14%
Kansas: 13.75%
Louisiana: 13.76%
Alabama: 13.95%
New Hampshire: 14.09%
Florida: 14.69%
North Carolina: 14.79%
North Dakota: 14.83%
Texas: 14.99%
Utah: 15.87%
Michigan: 16.14%
Georgia: 17.64%
Rhode Island: 18.01%
Nevada: 18.96%
Arkansas: 19.99%
Montana: 21.13%
Hawaii: 21.83%
Colorado: 24.68%

Toward the end of this post, I discounted certain record vote shares in certain states as anomalies. I felt this made for more informative maps. I still feel this way.

This, however, is a more quantitative way of looking for outliers. 

In 30 of the 50 states, the difference between the two top vote shares is less than 4%. In 36 of the 50--over 2/3--the difference is less than 5%. These are the states where the top vote share is over 5% (darkly shaded) or between 4% and 5% (lightly shaded). 

In 31 states, the difference between the highest and eighth-highest vote share is less than 13%. In 36 states, the difference is less than 14%. These are the states where the difference is over 14% (darkly shaded) or between 13% and 14% (lightly shaded).

Where do the anomalies I listed on the other post stand? 

-New Hampshire is shaded dark on both maps. 

-Missouri is darkly shaded on the first map, but not at all on the second. Kentucky is lightly shaded on both. Neither West Virginia nor Maryland is coloured in on either.

-Kansas darkly shaded on the first map, lightly on the second.

-Colorado, Idaho, Montana, Utah, and Nevada are all shaded darkly on the first map (quite conspicuously so). On the second map, they all are again except Idaho, which is not shaded at all. (Colorado specifically is, numerically, the strongest outlier on both lists.)

-Florida is darkly shaded on both maps.

Now, I don't think this is all you should take into account. You should also take into account how anomalous the state's vote in the election in which a record vote share set was. (For example, Missouri voting Democratic in eight straight elections after a Republican, Grant, had set the record high vote share in it; or Florida being the only state in which Grover Cleveland in '92 [or in any of his elections] holds the record high vote share.) In the cases of West Virginia and Maryland, I think either or both spreads would have been significantly larger at some point in the past, although I can't compare (or don't have the will to compare) with what the spread would have been for other states. 

There are some states that are shaded on these maps that I did not mention as anomalies. Some are Southern states where Franklin Roosevelt (but only Franklin Roosevelt) got the quasi-unanimous kind of support that Democrats regularly got in South Carolina and Mississippi.

Several are states in the Upper Midwest (or partly in the Upper Midwest, in Illinois' case) where Harding set the record: Illinois, Iowa, Wisconsin, North Dakota. 

It does seem Harding did somewhat atypically well, even for a landslide-winning Republican of the party's first 100 years (or what could be called its 'Vermont era'), in the Upper Midwest. However, a few points are worth noting. First of all, Harding was winning a national landslide (unlike, say, Horatio Seymour, or Martin van Buren, or Grover Cleveland--in 1892, or ever). Secondly, all the states he set records in were typically Republican at the time. North Dakota did, narrowly, vote for Wilson in 1916; but it also, alone out of the free-soil Haga Stack states, rejected William Jennings Bryan in 1896 (and in fact, to this day, has never voted for a losing Democrat). (1920 is not the only time a party's best state was one which had voted for the other party in the previous election--Rhode Island had voted for Eisenhower in 1956, and Georgia, for Nixon in 1972, for example. Mississippi hadn't voted for the Democrat in 1968, but it hadn't voted for Nixon either, and in fact had been his worst state that year.) Wisconsin's last three Democratic votes had been in 1912, 1892, and 1852; and Illinois', in 1912, 1892, and 1856. Aside from 1912, Iowa had not voted Democratic since the founding of the Republican Party.

Furthermore, in all four states, the record-high vote share as of October 1920 belonged to a Republican. In fact, the two highest vote shares (from 1836 on) belonged to Republicans (at least--I didn't check to see if the three highest did or not). As of October 1920, the all-time high vote share in Illinois, Wisconsin, and North Dakota belonged to TR; and in Iowa, to Lincoln '64. The second-highest ever vote share belonged to McKinley '00 in North Dakota and Wisconsin, to Grant '72 in Illinois, and to TR in Iowa.

Finally, it's worth noting that Illinois, Wisconsin, and Iowa show up only on the map detailing states with large divergences between the first- and second-highest vote shares. North Dakota shows up only on the map detailing states with large divergences between the first- and eighth-highest vote shares (something it also shares in common with some other Upper Midwest Republican-record states, such as Michigan and, to a lesser extent, Minnesota). Most of the states I set aside as outliers show up on both maps. In the cases of Iowa, Wisconsin, and Illinois, this would seem to indicate that, once one gets past the gap between Harding's vote share and the next-highest vote share, the differences grow small (and that the gap between Harding and whoever is #2 is, in and of itself, not big enough to virtually ensure idiosyncrasy, as in the case--to take an extreme example--of Colorado). This seems to have been true even at the time--the difference between TR's and Grant's '72 vote shares in Illinois was 2.50%, for example; between Lincoln '64 and TR in Iowa, 0.73%; and between TR and McKinley '00 in Wisconsin, 3.15%. In Illinois, Calvin Coolidge and Herbert Hoover contributed to still further tightening the clustering of the several highest vote shares (apart from Harding's) in the two immediately following elections, with Coolidge getting 58.84% (and displacing TR as #2), and Hoover getting 56.93% (and displacing Grant '72 as, now, #4). 

In North Dakota (and also Michigan, a Coolidge-record state, and Minnesota, a TR-record state), the difference between the top two vote shares is not large, but the overall spread between #1 and #8 is. In all three, (at least) the top two vote shares are held by 'Vermont-era' landslide Republican winners: TR and Harding in Minnesota; Coolidge, Harding, Hoover, and TR in Michigan (where they were all within 5.86% of one another); and Harding, TR, and Eisenhower '52 in North Dakota. 

It seems that what happened is that sometime in the middle of the 20th century, this general area of the country became permanently less Republican. So later Republican landslide winners such as Nixon and Reagan (or even Eisenhower, in Michigan and Minnesota) were not able to approach what earlier Republican landslide winners had done in those states. At the same time, they hadn't become Democratic either (or at least, not solidly so), so certainly Franklin Roosevelt, and not even Lyndon Johnson, was able to really more tightly cluster the top vote shares in these states (and they certainly didn't become states in which recent Democrats approach the record vote share, as in California and New York). 

It sounds funny to speak of North Dakota as having become permanently less Republican, but to all appearances, it has--Ronald Reagan was never able to get within even 6% of what any of Eisenhower in '52, TR, or Harding did in the state (or within any more than 4.75% of what FDR did in the state in 1932), even though his national vote share was comparable to that of any of these (and slightly higher than TR's, Eisenhower's in '52, and FDR's in '32). As for Nixon, he doesn't even appear in the top eight, at least today.

There is another batch of states that show up on these maps that is worth addressing: Rhode Island, Hawaii, and to a lesser extent, Connecticut (the first two show up on both maps; Connecticut, only on the first map and only lightly shaded).

These states have in common that they are states in which Lyndon Johnson set the all-time record in 1964. In the case of Rhode Island and Connecticut (and also Massachusetts, another LBJ record state), the previous record had belonged to a Republican. A large majority of all top eight vote shares still belong to Republicans in Massachusetts and Connecticut, and seems likely to have done so in Rhode Island (as recently as October 2000, a majority--although not a large one--of the top eight vote shares in Rhode Island had been set by Republicans). (Hawaii, of course, is a young state, and was participating in only its second presidential election in 1964.)

What seems to have happened in these states is that, like Michigan and Minnesota, they were Republican states in the party's 'Vermont era' but then became Democratic (or Democratic-leaning--Connecticut did vote for Ford) sometime in the middle of the 20th century. (As late as 1956, Connecticut, Massachusetts, and even Rhode Island, not only voted for Eisenhower, but voted for him more strongly than the nation overall.) Unlike Michigan and Minnesota, however, they became Democratic enough that Lyndon Johnson was able to perform comparably to earlier Republican landslide winners in them. Except that, in the cases of Rhode Island and Connecticut (especially Rhode Island), it wasn't comparably--it was significantly better. Take Johnson away, and the difference between the 2nd- and 9th-best vote shares in Rhode Island (Grant's in '72 and Lincoln's in '64) is just 9.70% (although of course, it would have presumably been bigger at the time--I don't know how much bigger, although I can say that Lincoln's '64 vote share in Rhode Island, and Garfield's in '80, are the same to the tenths place). Absent Johnson--and with the same caveat--the difference between the 2nd- and 9th-best vote shares in Connecticut (Eisenhower's in '56 and Nixon's in '72) is just 5.15%. 

In Massachusetts--where Johnson didn't outstrip the top previous Republican performer by quite as much--we can know that, as of October 1964, at least the top six vote shares would have all belonged to previous Republicans, all of whom ran between 1856 and 1920. The difference between #1 and #6 (Lincoln in '64 and Frémont in '56) would have been just 8.59%; and the difference between the top two (Lincoln in '64 and Grant in '68), just 2.44%.

Why Johnson did so much better than earlier Republicans in these states, I don't know, but I can speculate. As of October 1996, we can see that the top five Republican vote shares in Rhode Island all belonged to Republicans who ran between 1864 and 1920. Four of them belonged to Republicans who ran between 1864 and 1896. None of them belonged to TR or even to Yankee Calvin Coolidge. Similarly, as mentioned above, the top six Republican vote shares (which were also the top six vote shares full stop) in Massachusetts, as of October 1964, belonged to Republicans who ran between 1856 and 1920 (and, again, five of them, to Republicans who ran between 1856 and 1896--none to either TR or to the state's former governor, Calvin Coolidge). 

So it seems that these states' level of Republicanism had likely begun waning even well before the end of the 'Vermont era'. This 1951 article dates the long-term trend of waning Republicanism in New England in general, and Rhode Island in particular, back to 1896 (although that is potentially a somewhat misleading index value, since the Northeast reacted idiosyncratically strongly against Bryan). (And of course, even though the 'Vermont era' extends from 1856 to 1956 in one sense, it can also be divided into two sub-periods, one from 1856 to 1916, when Vermont was the Most Republican State in every election save 1864 and 1912 [in the latter of which it was the least Democratic State]; and one from 1920 to 1956, when it was still the Most Republican State more often than not, but more than intermittently handed off to a free-soil Haga Stack state.) 

Connecticut, heavily influenced by the suburban vote of Fairfield County, was a little different--as of October 1964, the top five vote shares would have belonged to Republicans, but to Republicans who ran between 1896 and 1956, with the very highest, at the time, belonging to Eisenhower, who led a party that, as Menendez writes, was centred on 'the booming and prosperous suburbs of the Northeast'. Even today, Reagan's '84 vote share stands as one of the top eight in the state. But Menendez also writes that this brand of Republicanism's 'dominance in the party and country was coming to an end in 1960'. This might be an overstatement--Bergen County did hand New Jersey to Nixon in 1968--but it is true that no subsequent Republican (including Nixon in '72 and Reagan in '84) was able to match Eisenhower's '56 performance in any of Bergen, NJ, Fairfield, CT, any of Nassau, Suffolk, or Westchester, NY, or even Montgomery, PA. In the case of all of these save Fairfield, no subsequent Republican was able to match Eisenhower's performance in either '52 or '56. (In Delaware County, PA, Nixon in '72 did better than Eisenhower in '56, but Reagan did not.) Conversely, Reagan in '84 outperformed either of Eisenhower's performances in Orange and San Diego, CA; Nixon in '72, in Palm Beach, FL and Orange, FL; and both, in Maricopa, AZ, Jefferson and Arapahoe, CO, and Fairfax, VA.

That said, in 1928--the same year as Rhode Island and Massachusetts voted for a Democrat for the first time (bar 1912) since the founding of the Republican Party--Connecticut voted much more Democratic than the nation, giving Hoover a < 10% margin, even though it would go on to be one of just six states that would vote for him in 1932.

In any case, crucially, 1964 was the only landslide enjoyed by the Democracy since these states became Democratic-leaning (or, in Hawaii's case, since it became a state). Outside national landslide conditions (which, again, haven't existed for the Democracy since 1964), Democrats appear to have found it hard to approach LBJ's vote shares in these states. (The spread in Massachusetts remains [relatively] small because of the performance of Republicans prior to LBJ.) And in any case, the epicentre of the Obama-era Democracy appears to be moving subtly away from these areas anyway. Recent Democrats have approached and even set all-time vote share records in a number of states, but these have tended not to be states where the prior top record (or even top Democratic record) had belonged to LBJ. For example, California (where the all-time and Democratic record belongs to FDR '36, and which actually voted more Republican than the nation in 1964), Washington (where the all-time record still belongs to TR, and the all-time Democratic record, to FDR '36), and New Mexico (where the all-time record still belongs to FDR '32, and which, like California, actually voted slightly more Republican than the nation in 1964).

In Hawaii, Obama came within 6.91% of LBJ's record in 2008 (and within 8.21% in 2012), but, although Hawaii does appear to have become permanently more Democratic than it used to be, the ability to supercharge its blueness to better than 2.25:1 (or even to better than 2:1) appears to have been limited personally to Obama, leaving the range in Hawaii still fairly idiosyncratic. In Rhode Island (as in Massachusetts and Connecticut), the second-best ever vote share remains Republican; and in Rhode Island, no recent Democrat (including Obama in '08) has outdone either Humphrey or Kennedy. In Connecticut, Obama's '08 vote share is the current seventh-best ever in the state (and second-best for a Democrat)--but he failed to overtake, not only a slew of Vermont-era Republicans, but also Reagan in '84. 

In any case, none of the states that potentially appear as additional outliers, seem like anomalies to me. One thing to note about all the states we have gone over is that they have come in batches of states that have behaved broadly similarly; and the record-holder is a national landslide winner. And the record, in every case, fits in with the state's political history (at least as of the time when it was set). (In LBJ's case, he was--up to at least October 2000--the only Democrat amongst the top eight vote shares in one state where he set the all-time record [Connecticut], as Bryan still is in Idaho today [and as van Buren was in New Hampshire from December 1924 through October 1964]. And he was the only Democrat amongst the top seven in another, Massachusetts. But in the case of Massachusetts, all of the second- through seventh-best vote shares belonged to Republicans prior to him. In the case of Connecticut, all of the second- through fifth-best vote shares belonged to Republicans prior to him, and the sixth- and eighth-best belonged, at least, to national landslide-winning Republicans. And he was himself winning a national landslide, of course--and in any case, this is no longer true of any of the states in which he holds the record for highest vote share.)

---

This is a map of the states by how high the eighth-highest vote share in them was (on the premiss that this is likely a better indication of their general inclination toward extreme partisanship, than their highest vote share). The shading corresponds to: between 1.25:1 and 1.5:1; between 1.5:1 and 1.75:1; between 1.75:1 and 3:1; and better than 3:1. The colour is the colour of the party that has set a majority (or a plurality, if applicable) of the top eight vote shares. (If there is a tie, then the colour used is the colour of the party that set the highest vote share.) That didn't, perhaps, lead to the most balanced distribution (although I found there were very few states where the eighth-highest vote share was between 7/11 and 2/3), although it does show an interesting distinction between the Dakotas and Nebransas. 

If we weren't concerned with covering every state, we would see more differentiation in the South (as indicated above). For example, here's a similar map, but with the following shadings: 8th-highest vote share between 2.25:1 and 2.75:1; between 2.75:1 and 3.5:1; between 3.5:1 and 5:1, and > 5:1 (with the rest of the states simply left blank). (It does have to be admitted that, apart from the outlier of Vermont--whose eighth-highest vote share is nevertheless higher than the eighth-highest in any Southern state save South Carolina, Mississippi, and Louisiana--all these states are in the South, albeit, more specifically, in the pre-Fort Sumter Confederacy.)