Each party's best raw vote margin states

The nominee who netted the most vote out of any single state is listed first; the overall national winner is in bold. If a state had the most electoral votes of any state at the time of the election, it is bolded.

1828
Andrew Jackson: Pennsylvania (28) (50,694)
John Quincy Adams: Massachusetts (15) (23,817)

1832
Henry Clay: Kentucky (15) (36,249)
Andrew Jackson: Tennessee (15) (26,731)

1836
Martin van Buren: New York (42) (28,247)
Whigs: Tennessee (15) (9,857) (0.66%)

1840
William Henry Harrison: Kentucky (15) (25,872)
Martin van Buren: Missouri (4) (7,015)

1844
Henry Clay: Massachusetts (12) (14,023)
James Polk: Illinois (9) (12,941)

1848
Zachary Taylor: New York (36) (98,086)
Lewis Cass: Ohio (23) (16,414)

1852
Franklin Pierce: New York (35) (27,201)
Winfield Scott: Vermont (5) (9,129)

1856
James Buchanan: Pennsylvania (27) (83,400)
John Frémont: New York (35) (80,126)

1860
Abraham Lincoln: Pennsylvania (27) (89,159)
John Breckinridge: Texas (4) (32,110)
Stephen Douglas: New Jersey (7) (4,523)

1864
Abraham Lincoln: Massachusetts (12) (77,997)
George McClellan: Kentucky (11) (36,514)

1868
Ulysses Grant: Massachusetts (12) (77,276)
Horatio Seymour: Kentucky (11) (76,323)

1872
Ulysses Grant: Pennsylvania (29) (137,548)
Horace Greeley: Missouri (15) (32,238)

1876
Samuel Tilden: Georgia (11) (79,624)
Rutherford Hayes: Iowa (11) (59,205)

1880
Winfield Scott Hancock: Texas (8) (98,535)
James Garfield: Iowa (11) (78,082)

1884
Grover Cleveland: Texas (13) (132,168)
James Blaine: Pennsylvania (30) (86,019)

1888
Grover Cleveland: Texas (13) (146,461)
Benjamin Harrison: Kansas (9) (80,159)

1892
Grover Cleveland: Texas (15) (139,460)
Benjamin Harrison: Pennsylvania (32) (63,747)

1896
William McKinley: Pennsylvania (32) (295,072)
William Jennings Bryan: Texas (15) (202,914)

1900
William McKinley: Pennsylvania (32) (288,433)
William Jennings Bryan: Texas (15) (136,791)

1904
Theodore Roosevelt: Pennsylvania (34) (502,951)
Alton Parker: Texas (18) (115,958)

1908
William Howard Taft: Pennsylvania (34) (296,997)
William Jennings Bryan: Texas (18) (151,636)

1912
Woodrow Wilson: New York (45) (200,086)
William Howard Taft: Utah (4) (5,521)

1916
Woodrow Wilson: Texas (20) (221,515)
Charles Evans Hughes: Illinois (29) (202,320)

1920
Warren Harding: New York (45) (1,089,929)
James Cox: Texas (20) (174,229)

1924
Calvin Coolidge: Pennsylvania (38) (2,144,850)
John Davis: Texas (20) (354,582)

1928
Herbert Hoover: Pennsylvania (38) (987,796)
Al Smith: Louisiana (10) (113,495)

1932
Franklin Roosevelt: Texas (23) (662,389)
Herbert Hoover: Pennsylvania (36) (157,592)

1936
Franklin Roosevelt: New York (47) (1,112,552)
Alf Landon: Maine (5) (42,490)

1940
Franklin Roosevelt: Texas (23) (697,282)
Wendell Willkie: Kansas (9) (124,444)

1944
Franklin Roosevelt: Texas (23) (630,180)
Thomas Dewey: Kansas (8) (154,638)

1948
Harry Truman: Texas (23) (520,768)
Thomas Dewey: Pennsylvania (35) (149,771)

1952
Dwight Eisenhower: New York (45) (848,214)
Adlai Stevenson: Georgia (12) (257,844)

1956
Dwight Eisenhower: New York (45) (1,589,571)
Adlai Stevenson: Georgia (12) (224,442)

1960
John Kennedy: Massachusetts (16) (510,424)
Richard Nixon: Ohio (25) (273,363)

1964
Lyndon Johnson: New York (43) (2,669,597)
Barry Goldwater: Mississippi (7) (303,910)

1968
Hubert Humphrey: Massachusetts (14) (702,374)
Richard Nixon: Indiana (13) (261,226)

1972
Richard Nixon: New York (41) (1,241,694)
George McGovern: Massachusetts (14) (220,462)

1976
Jimmy Carter: Georgia (12) (495,666)
Gerald Ford: Michigan (21) (197,028)

1980
Ronald Reagan: California (45) (1,441,197)
Jimmy Carter: Georgia (12) (236,565)

1984
Ronald Reagan: California (47) (1,544,490)
Walter Mondale: Minnesota (10) (3,761)

1988
George H. W. Bush: Florida (21) (962,184)
Michael Dukakis: New York (36) (266,011)

1992
Bill Clinton: California (54) (1,490,751)
George H. W. Bush: Texas (32) (214,256)

1996
Bill Clinton: New York (33) (1,822,685)
Bob Dole: Texas (32) (276,484)

2000
Al Gore: New York (33) (1,704,323)
George W. Bush: Texas (32) (1,365,893)

2004
George W. Bush: Texas (34) (1,694,213)
John Kerry: New York (31) (1,351,713)

2008
Barack Obama: California (55) (3,262,692)
John McCain: Texas (34) (950,695)

2012
Barack Obama: California (55) (3,014,327)
Mitt Romney: Texas (38) (1,261,719)

2016
Hillary Clinton: California (55) (4,269,978)
Donald Trump: Texas (38) (807,179)

2020
Joe Biden: California (55) (5,103,821)
Donald Trump: Tennessee (11) (708,764)

---

The parties' best raw vote margin states don't generally tell us as much as their best raw vote margin counties, or as much as their best states (by vote share).

There are four cases, arguably, of a single state having been a party's best raw vote margin state on a relatively consistent basis for an extended period of time. From 1860 through 1948, Pennsylvania was the Republicans' best raw vote margin state in 12 elections (out of 23), and for five straight elections from 1892 through 1908. (It has never been the Republicans' best raw vote margin state again.) It was the best raw vote margin state at least once for Lincoln, Grant, Blaine, Harrison, McKinley, Theodore Roosevelt, Taft, Coolidge, Hoover, and Dewey. At no time was it the nation's largest state.

From 1992 through 2016, Texas was the Republicans' best raw vote margin state for seven elections in a row, more elections in a row than any other state has been (although still not more than Pennsylvania total). Curiously, Texas first took this status amid a defeat, in 1992, when it was the largest state Bush Sr was carrying (and he was carrying it by only 3.48%). It was the best raw vote margin state at least once for George H. W. Bush, Bob Dole, George W. Bush, John McCain, Mitt Romney, and Donald Trump.

If we count Breckinridge as the Democratic nominee in 1860, Texas was the Democrats' best raw vote margin state in 16 elections (out of 23) from 1860 through 1948. (If we don't count Breckinridge, then it was the Democrats' best raw vote margin state in 15 elections [out of 18] from 1880 through 1948.) It was the Democrats' best raw vote margin state eight elections in a row, from 1880 through 1908, the most times in a row that a single state has been a party's best raw vote margin state. It was the best raw vote margin state at least once for Breckinridge, Hancock, Cleveland, Bryan, Parker, Wilson, Cox, Davis, Franklin Roosevelt, and Truman. 

Finally, California has been the Democrats' best raw vote margin state for five straight elections, as well as in 1992, and looks increasingly likely to be the party's best raw vote margin state indefinitely. Of course, it is also the largest state. It has been the best raw vote margin state at least once for Bill Clinton, Barack Obama, Hillary Clinton, and Joe Biden.

Some other states that have been a party's best raw vote margin state multiple times:

Massachusetts for the Republicans, 1864 and 1868 (and also for the National Republicans/Whigs, 1828 and 1844)

Iowa for the Republicans, 1876 and 1880

Kansas for the Republicans, 1888, 1940, 1944

Massachusetts for the Democrats, 1960, 1968, 1972

There has been one occasion on which, if DC were counted as a state (see here for why I don't), it would have been a party's best raw vote margin 'state': in 1984, Mondale netted 151,399 votes out of DC, well over the number he netted out of Minnesota. 

---

One interesting thing to look at is the differential in the amount of raw vote different nominees netted out of their best raw vote margin states. Sometimes, in a close election, one nominee nets discernibly more vote out of his best raw vote margin states, than the other nominee does out of his. This is the kind of situation that can lead to a relatively small Electoral College victory, or to an outright mismatch between the Electoral College and the popular vote, as noted respectively by 'Foggy Mtn Breakdown' and Nate Silver, respectively:

[P]lainly the South was the sine qua non for the victory to the incumbent, both as to the small presumptive electoral victory and the relatively small popular vote victory... 

The result, not surprisingly therefore, of a victory based on such a narrowly circumscribed population base is that the usual offset to a relatively slender popular vote margin, a much more substantial electoral victory, is not present in this election, with the maximum final margin being 34 electoral votes, 286 to 252. 

---

[I]n the first half of the 20th century...Republicans persistently had an Electoral College advantage because Democrats racked up huge margins in the South, yielding a lot of wasted votes as far as the Electoral College is concerned. This nearly led to a massive [sic] Electoral College-popular vote split in 1948...

The question is whether Democrats are re-entering something akin to the “Solid South” era, except with their votes concentrated in more urban coastal states instead of the South. In this respect, California — where Clinton leads by 28 percentage points, more than Obama’s 23-point margin in 2012 — represents lots of wasted votes, at least in terms of the Electoral College.

'Foggy Mtn Breakdown' is writing shortly after the 2004 election (so shortly that the final results appear not to have been finalised); Silver, after the 2016 election. Interestingly, Silver is implying that there was no remarkable concentration of popular vote for one party as recently as 2004, when 'Foggy Mtn Breakdown' is arguing there was such a concentration for the GOP. 

In discussing state raw vote margins, 'Foggy Mtn Breakdown' distinguishes Bush's margin out of Texas, and Kerry's out of New York and California, from all the others as the only ones comparable to one another; he then establishes further demarcations at 400,000 votes and 300,000 votes. He may have been choosing thresholds that felt natural to him, but it just so happens that in the 2004 election, these thresholds very closely approximated 1%, 0.3%, and 0.25% of the national vote, respectively. Let's compare the states that were won by such margins in the close elections of the 20th and 21st centuries. (Number of electoral votes and the percentage of the national vote represented by the margin in the state are included. Lines delineate the states that fall between each pair of adjacent thresholds, and each election is accompanied by a colour-coded map.)

(It should be mentioned that in an earlier version of this post, I mentioned that 'Foggy Mountain Breakdown' and George Will both defined 'the South' in a way I did not necessarily agree with, but which I use at times for reference--namely, as these 14 states. Otherwise, certain references below won't make sense.) 

1896 (map)
Pennsylvania (32) (295,072) (2.12%) (60.98%-36.27%)
New York (36) (268,469) (1.93%) (57.58%-38.72%)
Texas (15) (202,914) (1.46%) (68.00%-30.75%)
Massachusetts (15) (173,265) (1.24%) (69.47%-26.32%)
Illinois (24) (141,517) (1.02%) (55.66%-42.68%)
Colorado (4) (134,734) (0.97%) (84.95%-13.86%)
Wisconsin (12) (102,612) (0.74%) (59.93%-37.00%)
New Jersey (10) (87,840) (0.63%) (59.68%-36.02%)
Alabama (11) (74,625) (0.54%) (66.96%-28.61%)
Arkansas (8) (72,591) (0.52%) (73.72%-25.12%)
Iowa (13) (65,552) (0.47%) (55.47%-42.90%)
Missouri (17) (58,727) (0.42%) (53.96%-45.25%)
Mississippi (9) (58,536) (0.42%) (91.04%-6.92%)
Michigan (14) (56,170) (0.40%) (53.77%-43.47%)
Louisiana (8) (55,138) (0.40%) (76.38%-21.81%)
Minnesota (9) (53,768) (0.39%) (56.62%-40.89%)
Connecticut (6) (53,545) (0.38%) (63.24%-32.54%)
South Carolina (9) (49,488) (0.36%) (85.30%-13.51%)
Ohio (23) (48,494) (0.35%) (51.86%-47.08%)
Maine (6) (45,816) (0.329%) (67.90%-29.21%)
Vermont (4) (40,487) (0.29%) (80.08%-16.66%)
New Hampshire (4) (35,794) (0.26%) (68.66%-25.88%)

McKinley's raw vote margin (of 4.33%) could be accounted for in a minimum of three states, such as these three, or these three (which could allow us to say that his margin came from the Northeast or from the Mid-Atlantic, respectively).

1916 (map)
Texas (20) (221,515) (1.20%) (76.92%-17.45%)
Illinois (29) (202,320) (1.09%) (52.56%-43.34%)
Pennsylvania (38) (182,039) (0.98%) (54.26%-40.22%)
New York (45) (119,812) (0.65%) (51.53%-44.51%)
Georgia (14) (107,062) (0.58%) (79.51%-7.03%)
Ohio (24) (89,408) (0.48%) (51.86%-44.18%)
Colorado (6) (76,508) (0.41%) (60.74%-34.75%)
Mississippi (10) (76,169) (0.41%) (92.78%-4.91%)
Louisiana (10) (73,409) (0.40%) (85.90%-6.95%)
Alabama (12) (70,747) (0.38%) (76.04%-21.92%)
Arkansas (9) (63,332) (0.34%) (65.97%-28.73%)
Iowa (13) (61,470) (0.33%) (54.57%-42.55%)
South Carolina (9) (60,296) (0.33%) (96.71%-2.42%)
New Jersey (14) (57,964) (0.31%) (54.40%-42.68%)
Virginia (12) (53,456) (0.29%) (66.99%-31.83%)
Michigan (15) (52,322) (0.28%) (52.09%-44.05%)
Oklahoma (10) (50,880) (0.27%) (50.59%-33.21%)
North Carolina (12) (47,493) (0.26%) (58.10%-41.71%)

Wilson's national popular vote margin of 3.12% could be accounted for in six Southern states (such as these six, or these six, or these six, or these six, etc.).

1948 (map)
Texas (23) (520,768) (1.07%) (65.96%-24.29%)
Missouri (15) (262,276) (0.54%) (58.11%-41.49%)
Massachusetts (16) (242,418) (0.50%) (54.66%-43.16%)
Minnesota (11) (209,349) (0.43%) (57.16%-39.89%)
North Carolina (14) (200,498) (0.41%) (58.02%-32.68%)
Oklahoma (10) (183,965) (0.38%) (62.75%-37.25%)
Georgia (12) (169,591) (0.35%) (60.81%-20.31%)
Pennsylvania (35) (149,771) (0.31%) (50.93%-46.92%)
Kentucky (11) (125,546) (0.26%) (56.74%-41.48%)

1960 (map)
Massachusetts (16) (510,424) (0.74%) (60.22%-39.55%)
New York (45) (383,666) (0.56%) (52.53%-47.27%)
Ohio (25) (273,363) (0.40%) (53.28%-46.72%)
Indiana (13) (222,762) (0.32%) (55.03%-44.60%)
Kansas (8) (198,261) (0.29%) (60.45%-39.10%)
Georgia (12) (184,166) (0.27%) (62.54%-37.43%)
Louisiana (10) (176,359) (0.26%) (50.42%-28.59%)

Kennedy's extremely narrow national margin of 0.17% could be accounted for in any one of the blue states on the above list, as well as in Pennsylvania.

1968 (map)
Massachusetts (14) (702,374) (0.96%) (63.01%-32.89%)
New York (43) (370,538) (0.51%) (49.76%-44.30%)
Indiana (13) (261,226) (0.36%) (50.29%-37.99%)
California (40) (223,346) (0.31%) (47.82%-44.74%)
Michigan (21) (222,417) (0.30%) (48.18%-41.46%)
Florida (14) (210,010) (0.29%) (40.53%-30.93%)
Minnesota (10) (199,095) (0.27%) (54.00%-41.46%)

Nixon's national popular vote margin could be accounted for by the Nixon-Humphrey margins in these three Southern states, or in these three. This was at the same time as he became the first Republican since 1928 to carry North Carolina at all.

Humphrey was the first close-election Democrat for whom none of his best raw-vote margin states was in the South. His best raw vote margin in the South (including, incidentally, Maryland) was 66,536, or 0.091% of the national vote, out of West Virginia.

1976 (map)
Georgia (12) (495,666) (0.61%) (66.74%-32.96%)
Massachusetts (14) (399,199) (0.49%) (56.11%-40.44%)
New York (41) (288,767) (0.35%) (51.95%-47.52%)
Minnesota (10) (251,045) (0.31%) (54.90%-42.02%)
Arkansas (6) (230,861) (0.28%) (64.94%-34.93%)
Michigan (21) (197,028) (0.24%) (51.83%-46.44%)
Tennessee (10) (191,910) (0.24%) (55.94%-42.94%)
North Carolina (13) (185,405) (0.23%) (55.27%-44.22%)
Indiana (13) (169,244) (0.21%) (53.32%-45.70%)
Florida (17) (166,469) (0.20%) (51.93%-46.64%)
Utah (4) (155,798) (0.19%) (62.44%-33.65%)
Alabama (9) (155,100) (0.19%) (55.73%-42.61%)

Ford carried no states by 0.25% of the national popular vote, so I arbitrarily extended the list until the first 'natural' break outside the very top states (the next-largest raw vote margin for either nominee was 139,960 out of California for Ford--and each nominee's next-largest raw vote margin was out of a state won by less than 4% [California for Ford, Texas for Carter]).

Carter drew 0.28% of the national vote out of Arkansas, a state with just six electoral votes. The only times a major party nominee has drawn at least 0.25% of the national vote out of a state with six or fewer electoral votes (out of the elections under consideration) are 1916 (when Wilson did so out of Colorado) and 2000, 2004, and 2012 (when the Republican did so out of Utah).

Carter's best raw vote margin states in the South were Georgia, Arkansas, Tennessee, and North Carolina. Of these four, two voted for Bill Clinton twice (Arkansas and Tennessee). These two, along with West Virginia, were, interestingly, three of the four states characterised by Karl Rove as the 'four historically Democrat[ic] border states' that it was part of Bush's strategy to win in 2000. (The fourth was Kentucky, which, while it did not vote for Carter all that powerfully in 1976, was one of eight former slave states that voted for Reagan in 1980 by less than 2.5% and with only a plurality. [Reagan's next-narrowest margin in a former slave state that year was 5.45%, in Louisiana, where he won a majority.] Of these eight states, only Kentucky, Tennessee, and Arkansas [and Delaware] voted for Clinton twice [or indeed ever].) 

Arkansas, Tennessee, and West Virginia were also the only former slave states that gave Carter over 55% of the vote that went on to vote for Clinton twice (the others were North Carolina, Alabama, South Carolina, and Carter's home state of Georgia).

Carter's national popular vote margin over Ford (2.06%) could be accounted for in nine Southern states, some possible combinations of which are these, these, these, and these. 

1992 (map)
California (54) (1,490,751) (1.43%) (46.01%-32.61%)
New York (33) (1,097,801) (1.05%) (49.73%-33.88%)
Illinois (22) (719,254) (0.69%) (48.58%-34.34%)
Massachusetts (12) (513,613) (0.49%) (47.54%-29.03%)
Pennsylvania (23) (447,323) (0.43%) (45.15%-36.13%)
Michigan (18) (316,242) (0.30%) (43.77%-36.38%)
Maryland (10) (281,477) (0.27%) (49.80%-35.62%)
Minnesota (10) (273,156) (0.26%) (43.48%-31.85%)
Washington (11) (261,803) (0.251%) (43.41%-31.97%)

1992 was not particularly close, but it is interesting because it was the first time that a Democrat won the election without netting at least 0.25% of the national vote out of a Southern state. (For these purposes, I am excluding Maryland, although by 2020, we shall also have to exclude Virginia.) Clinton's best raw vote margin state in the South (as defined by George Will) was 168,499 (or ~0.16% of the national vote) out of Arkansas. (If we include Missouri, it was 242,714, or ~0.23% of the national vote, out of that state.)

George H. W. Bush's best raw vote margin state was Texas, which gave him a margin of 214,256 votes, or ~0.21% of the national vote. This was followed by Indiana, and then--if we take the Bush-Clinton margin--Utah, and then Nebraska, followed by six Southern states (Oklahoma, Alabama, Virginia, Florida, South Carolina, and Mississippi; after Mississippi there is something of a 'natural gap').

1996 (map)
New York (33) (1,822,685) (1.89%) (59.47%-30.61%)
California (54) (1,291,455) (1.34%) (51.10%-38.21%)
Massachusetts (12) (853,656) (0.89%) (61.47%-28.09%)
Illinois (22) (754,723) (0.78%) (54.32%-36.81%)
New Jersey (15) (549,251) (0.57%) (53.72%-35.86%)
Michigan (18) (508,441) (0.53%) (51.69%-38.48%)
Pennsylvania (23) (414,650) (0.43%) (49.17%-39.97%)
Minnesota (10) (353,962) (0.37%) (51.10%-34.96%)
Florida (25) (302,334) (0.31%) (48.02%-42.32%)
Ohio (21) (288,339) (0.30%) (47.38%-41.02%)
Maryland (10) (284,677) (0.30%) (54.25%-38.27%)
Washington (11) (282,611) (0.29%) (49.84%-37.30%)
Texas (32) (276,484) (0.29%) (48.76%-43.83%)
Connecticut (8) (252,631) (0.26%) (52.83%-34.69%)

Clinton's second win was also the second time that a Democrat had won without netting at least 0.25% of the national vote out of any Southern state (again, leaving aside Maryland). His best raw vote margin in a Southern state (excluding Maryland, but otherwise whether we use George Will's definition or not) was 215,251 (or ~0.22% of the national vote) out of Louisiana.

The most natural break in Dole's best raw vote margin states (apart from that between Texas and Kansas) is perhaps between Idaho (margin 91,152 votes) and South Carolina (margin 69,407 votes). Dole's best raw vote margin states through Idaho were: Texas, Kansas, Utah, Nebraska, Indiana, North Carolina, Alabama, Oklahoma, and Idaho.

This was only the second time that a losing Republican netted over 0.25% of the national vote out of a Southern state (the first being 1964, when Goldwater did so out of Mississippi and Alabama) (although, amongst narrowly-winning Republicans, Nixon had also done so in 1968, out of Florida).

2000 (map)
New York (33) (1,703,323) (1.62%) (60.21%-35.23%)
Texas (32) (1,365,893) (1.30%) (59.30%-37.98%)
California (54) (1,293,774) (1.23%) (53.45%-41.65%)
Massachusetts (12) (737,985) (0.70%) (59.80%-32.50%)
Illinois (22) (569,605) (0.54%) (54.60%-42.58%)
New Jersey (15) (504,677) (0.48%) (56.13%-40.29%)
North Carolina (14) (373,471) (0.35%) (56.03%-43.20%)
Indiana (12) (343,856) (0.326%) (56.65%-41.01%)
Maryland (10) (331,985) (0.31%) (56.57%-40.18%)
Utah (5) (312,043) (0.30%) (66.83%-26.34%)
Georgia (13) (303,490) (0.29%) (54.67%-42.98%)
Oklahoma (8) (270,061) (0.256%) (60.31%-38.43%)
Connecticut (8) (254,921) (0.242%) (55.91%-38.44%)

2004 (map)
Texas (34) (1,694,213) (1.39%) (61.09%-38.22%)
New York (31) (1,351,713) (1.11%) (58.37%-40.08%)
California (55) (1,235,659) (1.01%) (54.31%-44.36%)
Massachusetts (12) (732,691) (0.60%) (61.94%-36.78%)
Georgia (15) (548,105) (0.45%) (57.97%-41.37%)
Illinois (21) (545,604) (0.45%) (54.82%-44.48%)
Indiana (11) (510,427) (0.42%) (59.94%-39.26%)
Alabama (9) (482,461) (0.39%) (62.46%-36.84%)
Oklahoma (7) (455,826) (0.37%) (65.57%-34.43%)
North Carolina (15) (435,317) (0.36%) (56.02%-43.58%)
Utah (5) (422,543) (0.35%) (71.54%-26.00%)
Florida (27) (380,978) (0.31%) (52.10%-47.09%)
Kentucky (8) (356,706) (0.29%) (59.55%-39.69%)
Tennessee (11) (347,898) (0.28%) (56.80%-42.53%)
Maryland (10) (309,790) (0.253%) (54.82%-44.48%)

2008 (map)
California (55) (3,262,692) (2.48%) (61.01%-36.95%)
New York (31) (2,052,174) (1.56%) (62.88%-36.03%)
Illinois (21) (1,388,169) (1.06%) (61.92%-36.78%)
Texas (34) (950,695) (0.72%) (55.45%-43.68%)
Michigan (17) (823,940) (0.63%) (57.43%-40.96%)
Massachusetts (12) (795,243) (0.61%) (61.80%-35.99%)
Maryland (10) (669,605) (0.51%) (61.92%-36.47%)
Pennsylvania (21) (620,478) (0.47%) (54.49%-44.17%)
New Jersey (15) (602,215) (0.46%) (57.27%-41.70%)
Washington (11) (521,632) (0.40%) (57.65%-40.48%)
Oklahoma (7) (457,669) (0.35%) (65.65%-34.35%)
Alabama (9) (453,067) (0.35%) (60.32%-38.74%)
Wisconsin (10) (414,818) (0.32%) (56.22%-42.31%)
Tennessee (11) (391,741) (0.30%) (56.90%-41.83%)
Connecticut (7) (368,344) (0.28%) (60.59%-38.22%)
Louisiana (9) (365,286) (0.28%) (58.56%-39.93%)

2012 (map)
California (55) (3,014,327) (2.34%) (60.24%-37.12%)
New York (29) (1,995,310) (1.55%) (63.35%-35.17%)
Texas (38) (1,261,719) (0.98%) (57.17%-41.38%)
Illinois (20) (884,296) (0.69%) (57.60%-40.73%)
Massachusetts (11) (732,976) (0.57%) (60.65%-37.51%)
Maryland (10) (705,975) (0.55%) (61.97%-35.90%)
New Jersey (14) (647,533) (0.50%) (58.38%-40.59%)
Tennessee (11) (501,621) (0.39%) (59.48%-39.08%)
Utah (6) (488,787) (0.38%) (72.79%-24.75%)
Washington (12) (464,726) (0.36%) (56.16%-41.29%)
Alabama (9) (460,229) (0.36%) (60.55%-38.36%)
Michigan (16) (449,313) (0.35%) (54.21%-44.71%)
Oklahoma (7) (447,778) (0.35%) (66.77%-33.23%)
Kentucky (8) (407,820) (0.32%) (60.49%-37.80%)
Louisiana (8) (343,121) (0.27%) (57.78%-40.58%)

2016 (map)
California (55) (4,269,978) (3.12%) (61.73%-31.62%)
New York (29) (1,736,590) (1.27%) (59.01%-36.52%)
Illinois (20) (944,714) (0.69%) (55.83%-38.76%)
Massachusetts (11) (904,303) (0.66%) (60.01%-32.81%)
Texas (38) (807,179) (0.59%) (52.23%-43.24%)
Maryland (10) (734,759) (0.54%) (60.33%-33.91%)
Tennessee (11) (652,230) (0.48%) (60.72%-34.72%)
Alabama (9) (588,708) (0.43%) (62.08%-34.36%)
Kentucky (8) (574,177) (0.42%) (62.52%-32.68%)
New Jersey (14) (546,345) (0.40%) (55.45%-41.35%)
Oklahoma (7) (528,761) (0.39%) (65.32%-28.93%)
Indiana (11) (524,160) (0.38%) (56.82%-37.91%)
Missouri (10) (523,443) (0.38%) (56.77%-38.14%)
Washington (12) (520,971) (0.38%) (52.54%-36.83%)
Ohio (18) (446,841) (0.327%) (51.69%-43.56%)
Louisiana (8) (398,484) (0.29%) (58.09%-38.45%)

2020 (map)
1. Biden, California (5,103,821) (3.22%) (63.48%-34.32%)
2. Biden, New York (1,986,187) (1.25%) (60.86%-37.75%)
3. Biden, Massachusetts (1,215,000) (0.77%) (65.60%-32.14%)
4. Biden, Illinois (1,025,024) (0.65%) (57.54%-40.55%)
5. Biden, Maryland (1,008,609) (0.64%) (65.36%-32.15%)
6. Biden, Washington (784,961) (0.496%) (57.97%-38.77%)
7. Biden, New Jersey (725,061) (0.46%) (57.33%-41.40%)
8. Trump, Tennessee (708,764) (0.45%) (60.66%-37.45%)
9. Trump, Texas (631,221) (0.40%) (52.06%-46.48%)
10. Trump, Alabama (591,546) (0.37%) (62.03%-36.57%)
11. Trump, Kentucky (554,172) (0.35%) (62.09%-36.15%)
12. Trump, Oklahoma (516,390) (0.326%) (65.37%-32.29%)
13. Trump, Indiana (487,103) (0.31%) (57.02%-40.96%)
14. Trump, Ohio (475,669) (0.30%) (53.27%-45.24%)
16. Trump, Missouri (465,772) (0.29%) (56.80%-41.41%)
17. Biden, Virginia (451,138) (0.28%) (54.11%-44.00%)
18. Biden, Colorado (439,745) (0.28%) (55.40%-41.90%)
19. Trump, Louisiana (399,742) (0.252%) (58.46%-39.85%)

Biden's netting over 0.25% of the national vote out of Virginia made him the first Democrat (winning or otherwise) to net over 0.25% of the national vote out of a Southern state since 1980. Hence why there has been so much discussion about how Biden was able to revive the party's fortunes in the South after a generation. Except for one thing.

When the top state raw vote margins are ordered irrespective of party in 2020, we see that they sort out such that the top seven are all Biden margin states, followed by eight straight Trump margin states. They had never sorted out quite this neatly, but Biden's top seven raw vote margin states had also been Kerry's top seven in 2004, Obama's top seven in 2012 (but not in 2008), and Hillary Clinton's top seven in 2016. (The last time they were not for a losing or narrowly-winning Democrat was 2000, when Connecticut--and also Michigan, Pennsylvania, and DC--gave Gore a bigger raw vote margin than Washington State did.)

These seven states could be said to be the Democrats' new 'Solid South' (the phenomenon Nate Silver was alluding to in the above-quoted article).

Nate Silver and 'Foggy Mtn Breakdown' both imply that if one party is more regionally concentrated, it will be more prone to lose close elections. However, in the close elections of 1916, 1948, and 1976, one party was clearly more regionally concentrated than the other, and yet it was the more regionally concentrated party that won the Electoral College. 

In the 21st century, it seems that it becomes harder to say which party (if any) is more regionally concentrated. Nate Silver implies that before 2016, neither party had been particularly more regionally concentrated than the other since the Solid South days, but Foggy Mtn Breakdown clearly saw 2004 as an election in which the GOP was more regionally concentrated. 

In fact, in 2004, George W. Bush netted 5,634,474 votes out of the states giving him at least 0.25% of the national vote, and Kerry netted 4,175,457, about 74.1% as much. This was the last time that the Republican netted more vote out of his > 0.25%-net states than the Democrat did. How does the discrepancy compare with 1916, 1948, and 1976? Well, while it is clearly less discrepant than 1948 or 1976, it is comparably discrepant to (although still slightly less discrepant than) 1916, when Hughes netted about 68.3% as much vote out of his > 0.25%-net states as Wilson did out of his (675,927 vs. 990,275). 

There is another similarity between 1916 and 2004. In 1916, Hughes netted at least 0.25% of the national vote out of six states; Wilson did so out of twelve. In 2004, Kerry did so out of five states; Bush did so out of ten. And, finally, in 2004, even though Bush netted more vote out of his > 0.25%-net states than Kerry did out of his, Kerry predominated amongst the states giving the very largest raw vote margins to either nominee, as did Hughes in 1916. As Foggy Mtn Breakdown notes, the raw vote margins rendered by California, New York, and Texas in 2004 are comparable only to one another--and of those three, two went to Kerry; Kerry netted 2,587,372 votes out of California and New York combined, whereas Bush netted only 1,694,213 votes out of Texas (and the fourth-largest state raw vote margin also went to Kerry). In 1916, even though Hughes' raw vote margin out of Pennsylvania falls just below the 1% threshold, one could argue that the raw vote margins out of Pennsylvania, Texas, and Illinois are the only ones comparable to one another, and once again, two of these went to Hughes (who collectively netted 384,359 votes out of PA and IL), and only one, to Wilson (who netted just 221,515 votes out of TX). (And, once again, the fourth-largest state raw vote margin also went to Hughes.)

In more recent elections, such as 2012, 2016, and 2020, the Democrat clearly netted more vote out of his or her best raw vote margin states, than the Republican did out of his, whether one cuts these states off at 1%, 1/3 of 1%, or 0.25%. (This was also true in 2000, when Gore netted 5,141,349 votes out of his > 0.25%-net states, and Bush, only 2,968,814.)

This might lead one to conclude that the Democracy is more regionally concentrated. And that might be true, if we are willing to grant that 'urban coastal states' (+ Illinois) is a 'region' (which Foggy Mtn Breakdown might not buy into--but it is uncanny how the same seven states have provided Democrats with their seven best state raw vote margins in four of the last five elections). But--unlike in 1948 or 1976, certainly--the other party seems abnormally regionally concentrated as well. It just nets less raw vote out of its region of concentration. But, for example, in 2020, Trump netted 4,830,379 votes out of his > 0.25%-net states--3.05% of the national total. That is much less than Biden netted out of his > 0.25%-net states (12,739,546 votes, or 8.04% of the national total), or even than Biden netted out of the seven reliable top Democratic raw vote margin states (11,848,663 votes, or 7.48% of the national total). But it is comparable to the proportion of the national vote that Truman netted out of his > 0.25%-net states (1,914,411 votes, or 3.92% of the national total) and to the proportion of the national vote that Wilson netted out of his > 0.25%-net states (706,753 votes, or 3.81% of the national total). (It is greater than the proportion of the national vote Carter netted out of his > 0.25%-net states in 1976--2,042,853 votes, or 2.51% of the national total--although no-one seems to point to that election as an example of regional partisan concentration.) 

In 2016--when Trump won--Trump's collective net out of his > 0.25%-net states was even more comparable to Wilson's in 1916 or Truman's in 1948: 5,043,983 votes, or 3.69% of the national total.

Of course, Foggy Mtn Breakdown points out specifically that Bush's popular vote win in 2004 was accounted for in the South. He notes, in particular, that it could be accounted for in five Southern states (again, the data might have been incomplete as of when he was writing; it ended up that Bush's margin could be made up in four Southern states, of which there are a few different possible combinations). He also notes, more trivially, that Bush's popular vote win could be made up 'in the South', which was also true of Wilson's popular vote win in 1916 and Carter's in 1976 (both proportionally comparable to Bush's in 2004). 

In contrast, if one is not willing to grant that the reliable top seven Democratic raw vote margin states could be considered a 'region', it is impossible to make up Biden's raw vote margin in 2020 in 'a region'. His popular vote win cannot be accounted for in these states/entities, for example (which is the broadest possible definition of 'the Northeast' currently, as Virginia could be considered Northeastern now given that it is Northeastern in the only way Maryland ever was, i.e., in that it votes like the Northeast). Nor can it be accounted for in these states/entities (the broadest possible idea of what Biden could be said to have carried in 'the West'). (The latter comes exceedingly close--Biden's combined net out of what is coloured blue on the second map was 7,045,592. That said, I would not generally count districts within states in these considerations, leaving aside what one may or may not consider 'the West'.)

It is interesting that in general, the more regionally concentrated party has tended to win close elections--until recently. Nate Silver mentions the EC-PV mismatch that nearly occurred in 1948, but Truman, who piled up more wasted vote, did find a way to win the Electoral College, as did Wilson in 1916 and Carter in 1976 (and Bush in 2004). Hillary Clinton, however, did not in 2016. Perhaps it is something to do with the region of concentration being the South in 1916, 1948, 1976, and 2004--and being something else ('urban coastal states', or even simply California) in 2016.

---

Another interesting observation is that in 2008, Obama temporarily interrupted the streak of the same seven states being the Democrats' seven best raw vote margin states, as Michigan gave him his fourth-best raw vote margin that year, and Pennsylvania, his seventh. This would tend to mark 2008 out as a 'generic Democratic year', in contrast to the other elections of this century (especially given that those two states ended up voting for the next narrowly-winning Republican). 

From (at least) 1948 through 1996, every winning Democrat (except for Kennedy, but plus narrowly-losing Hubert Humphrey) netted at least 0.25% of the national total out of Minnesota. Gore and Kerry fell well short of this, only narrowly winning the state, but more surprisingly, even Obama in 2008 was unable to do so. He netted only 0.227% of the national vote out of Minnesota that year. Of course, it could be a matter of a state's changing proportion of the national population--Obama would have needed to win Minnesota by at least 11.28% to have netted 0.25% of the national vote out of it. (He actually won it by 10.24%.) Well, by comparison, Truman won it by 17.27% in 1948; Humphrey, by 12.53% in 1968; Carter, by 12.87% in 1976; and Clinton, by 11.63% in 1992 and by 16.14% in 1996.

Even though Biden won Minnesota by a margin that should be discomfitingly comfortable for Republicans in 2020, he fell well short, once again, of netting 0.25% of the national vote out of it.

On the other hand, Obama netted over 0.25% of the national vote out of Wisconsin in 2008. In fact, aside from the reliable Democratic top seven and Connecticut (which had been in Gore's top seven instead of Washington in 2000), the only states Obama netted at least 0.25% of the national vote out of in 2008 were the three Blue Wall states Trump ended up flipping in 2016.

In 2000, Gore's eighth-best raw vote margin state (with Connecticut, again, taking the place that would thereafter be normally occupied by Washington) was Michigan, which he won by 5.13%. In 2004, Kerry's was DC--but after DC, it was Michigan (which he won by 3.42%, a slightly larger margin than Trump's much-heralded 3.36% margin in Florida in 2020). In 2012, Obama's was Michigan (and he netted over 0.25% of the national vote out of it). In 2016, Hillary Clinton's was DC, but after DC, it was former top-7 member Connecticut. In 2020, as we can see, Biden's was Virginia, followed by Colorado, showing perhaps that the limitations of the Democracy in the Midwest that first evidenced themselves in 2016 are to at least some degree permanent. (On the other hand, the fact that the two Dole-Hillary Clinton states of Virginia and Colorado not only voted for Biden but gave him over 0.25% of the national total shows just how far gone those two states, considered battlegrounds as recently as 2016, are as far as the GOP is concerned.)