Any Smith-Hoover Counties?

[TheReckoning]

If so, what caused the shift?

[Calthrina950]

There are none. According to Albert J. Menendez's Geography of Presidential Elections in the United States, not a single county shifted from Smith in 1928 to Hoover in 1932, while Roosevelt gained 1,791 counties which Hoover had carried in 1928. This is probably the only presidential election in which the losing party did not gain any counties. In 1964, Goldwater flipped over 500 counties in the South that either Kennedy or the States' Rights candidates (Orval Faubus and Harry F. Byrd) had won in 1960. In 1972, McGovern picked up several of Nixon's 1968 counties, such as Athens County, Ohio and Pitkin County, Colorado.

In 1984, Mondale won several Reagan 1980 counties, including Marin County, California and Denver County, Colorado. Parker in 1904, Cox in 1920, Davis in 1924, Landon in 1936, and Stevenson in 1952/56 all gained counties despite losing in landslides. 1932 had the greatest swing between parties of any presidential election. Hoover did gain several points in heavily Catholic Elk County, which Smith had flipped in 1928 (after it had been won in landslides by Harding and Coolidge).

[Sir Mohamed]

There are none. According to Albert J. Menendez's Geography of Presidential Elections in the United States, not a single county shifted from Smith in 1928 to Hoover in 1932, while Roosevelt gained 1,791 counties which Hoover had carried in 1928. This is probably the only presidential election in which the losing party did not gain any counties. In 1964, Goldwater flipped over 500 counties in the South that either Kennedy or the States' Rights candidates (Orval Faubus and Harry F. Byrd) had won in 1960. In 1972, McGovern picked up several of Nixon's 1968 counties, such as Athens County, Ohio and Pitkin County, Colorado.

In 1984, Mondale won several Reagan 1980 counties, including Marin County, California and Denver County, Colorado. Parker in 1904, Cox in 1920, Davis in 1924, Landon in 1936, and Stevenson in 1952/56 all gained counties despite losing in landslides. 1932 had the greatest swing between parties of any presidential election. Hoover did gain several points in heavily Catholic Elk County, which Smith had flipped in 1928 (after it had been won in landslides by Harding and Coolidge).

Very interesting. That means there are Roosevelt 1932/Landon 1936 counties, too?

[𝕭𝖆𝖕𝖙𝖎𝖘𝖙𝖆 𝕸𝖎𝖓𝖔𝖑𝖆]

As Calthrina notes, no county flipped from Smith to Hoover in 1932. Even finding counties which swung Republican is incredibly difficult. The only places I am aware of outside of Elk County, which he mentioned and had a pretty massive swing (double digits percentage points) for some obscure reason, are Berkshire County in Western Massachusetts (Smith won by 0.5% and Roosevelt won by 0.1%) and some places in notoriously free and fair South Carolina (or Mississippi) - for example according to Wikipedia Smith won Bamberg County, SC 99.5-0.5 but Roosevelt "only" 99.1-0.9 don't ask me why.

[Calthrina950]

There are none. According to Albert J. Menendez's Geography of Presidential Elections in the United States, not a single county shifted from Smith in 1928 to Hoover in 1932, while Roosevelt gained 1,791 counties which Hoover had carried in 1928. This is probably the only presidential election in which the losing party did not gain any counties. In 1964, Goldwater flipped over 500 counties in the South that either Kennedy or the States' Rights candidates (Orval Faubus and Harry F. Byrd) had won in 1960. In 1972, McGovern picked up several of Nixon's 1968 counties, such as Athens County, Ohio and Pitkin County, Colorado.

In 1984, Mondale won several Reagan 1980 counties, including Marin County, California and Denver County, Colorado. Parker in 1904, Cox in 1920, Davis in 1924, Landon in 1936, and Stevenson in 1952/56 all gained counties despite losing in landslides. 1932 had the greatest swing between parties of any presidential election. Hoover did gain several points in heavily Catholic Elk County, which Smith had flipped in 1928 (after it had been won in landslides by Harding and Coolidge).

Very interesting. That means there are Roosevelt 1932/Landon 1936 counties, too?

Yes. Major County, Oklahoma and Winston County, Alabama-two ancestrally Republican counties that Roosevelt managed to win in 1932-went back to Landon in 1936. Landon also flipped a substantial number of counties in Missouri, Kentucky, Iowa, the Dakotas, and Nebraska which Roosevelt had carried in 1932, and regained a number of counties in the Texas Hill Country (Texas was an all-county sweep for Roosevelt in 1932). Overall, Landon carried 203 counties which Roosevelt had won four years previously.

[TheElectoralBoobyPrize]

There are none. According to Albert J. Menendez's Geography of Presidential Elections in the United States, not a single county shifted from Smith in 1928 to Hoover in 1932, while Roosevelt gained 1,791 counties which Hoover had carried in 1928. This is probably the only presidential election in which the losing party did not gain any counties. In 1964, Goldwater flipped over 500 counties in the South that either Kennedy or the States' Rights candidates (Orval Faubus and Harry F. Byrd) had won in 1960. In 1972, McGovern picked up several of Nixon's 1968 counties, such as Athens County, Ohio and Pitkin County, Colorado.

In 1984, Mondale won several Reagan 1980 counties, including Marin County, California and Denver County, Colorado. Parker in 1904, Cox in 1920, Davis in 1924, Landon in 1936, and Stevenson in 1952/56 all gained counties despite losing in landslides. 1932 had the greatest swing between parties of any presidential election. Hoover did gain several points in heavily Catholic Elk County, which Smith had flipped in 1928 (after it had been won in landslides by Harding and Coolidge).

Very interesting. That means there are Roosevelt 1932/Landon 1936 counties, too?

Yes. Major County, Oklahoma and Winston County, Alabama-two ancestrally Republican counties that Roosevelt managed to win in 1932-went back to Landon in 1936. Landon also flipped a substantial number of counties in Missouri, Kentucky, Iowa, the Dakotas, and Nebraska which Roosevelt had carried in 1932, and regained a number of counties in the Texas Hill Country (Texas was an all-county sweep for Roosevelt in 1932). Overall, Landon carried 203 counties which Roosevelt had won four years previously.

I always wondered why despite losing by worse than Hoover in both the popular vote and electoral vote, losing four states that Hoover had won, and failing to win any states that Hoover lost (not even Landon's own state), nonetheless there were 15 states where Landon did better than Hoover. In most of them, the shift was significant. There obviously were a lot of Roosevelt '32/Landon '36 voters.

[E-Dawg]

There are none. According to Albert J. Menendez's Geography of Presidential Elections in the United States, not a single county shifted from Smith in 1928 to Hoover in 1932, while Roosevelt gained 1,791 counties which Hoover had carried in 1928. This is probably the only presidential election in which the losing party did not gain any counties. In 1964, Goldwater flipped over 500 counties in the South that either Kennedy or the States' Rights candidates (Orval Faubus and Harry F. Byrd) had won in 1960. In 1972, McGovern picked up several of Nixon's 1968 counties, such as Athens County, Ohio and Pitkin County, Colorado.

In 1984, Mondale won several Reagan 1980 counties, including Marin County, California and Denver County, Colorado. Parker in 1904, Cox in 1920, Davis in 1924, Landon in 1936, and Stevenson in 1952/56 all gained counties despite losing in landslides. 1932 had the greatest swing between parties of any presidential election. Hoover did gain several points in heavily Catholic Elk County, which Smith had flipped in 1928 (after it had been won in landslides by Harding and Coolidge).

Very interesting. That means there are Roosevelt 1932/Landon 1936 counties, too?

Yes. Major County, Oklahoma and Winston County, Alabama-two ancestrally Republican counties that Roosevelt managed to win in 1932-went back to Landon in 1936. Landon also flipped a substantial number of counties in Missouri, Kentucky, Iowa, the Dakotas, and Nebraska which Roosevelt had carried in 1932, and regained a number of counties in the Texas Hill Country (Texas was an all-county sweep for Roosevelt in 1932). Overall, Landon carried 203 counties which Roosevelt had won four years previously.

Why did Roosevelt win a number of these Southern ancestrally R counties in 1932 only to lose them in 1936 when he did better nationwide? (Winston AL, Newton and Searcy AR, Gillespie and Kendall TX, Major OK) Some of these swings were insane too, he won the Texas ones in 1932 by around 60 points only to lose them in 1936 by around 20. What happened here?

[Calthrina950]

There are none. According to Albert J. Menendez's Geography of Presidential Elections in the United States, not a single county shifted from Smith in 1928 to Hoover in 1932, while Roosevelt gained 1,791 counties which Hoover had carried in 1928. This is probably the only presidential election in which the losing party did not gain any counties. In 1964, Goldwater flipped over 500 counties in the South that either Kennedy or the States' Rights candidates (Orval Faubus and Harry F. Byrd) had won in 1960. In 1972, McGovern picked up several of Nixon's 1968 counties, such as Athens County, Ohio and Pitkin County, Colorado.

In 1984, Mondale won several Reagan 1980 counties, including Marin County, California and Denver County, Colorado. Parker in 1904, Cox in 1920, Davis in 1924, Landon in 1936, and Stevenson in 1952/56 all gained counties despite losing in landslides. 1932 had the greatest swing between parties of any presidential election. Hoover did gain several points in heavily Catholic Elk County, which Smith had flipped in 1928 (after it had been won in landslides by Harding and Coolidge).

Very interesting. That means there are Roosevelt 1932/Landon 1936 counties, too?

Yes. Major County, Oklahoma and Winston County, Alabama-two ancestrally Republican counties that Roosevelt managed to win in 1932-went back to Landon in 1936. Landon also flipped a substantial number of counties in Missouri, Kentucky, Iowa, the Dakotas, and Nebraska which Roosevelt had carried in 1932, and regained a number of counties in the Texas Hill Country (Texas was an all-county sweep for Roosevelt in 1932). Overall, Landon carried 203 counties which Roosevelt had won four years previously.

I always wondered why despite losing by worse than Hoover in both the popular vote and electoral vote, losing four states that Hoover had won, and failing to win any states that Hoover lost (not even Landon's own state), nonetheless there were 15 states where Landon did better than Hoover. In most of them, the shift was significant. There obviously were a lot of Roosevelt '32/Landon '36 voters.

Part of it may have been a regional or "native-son" effect; Landon seems to have had considerable appeal to Appalachian voters, particularly voters in ancestrally Republican areas. Part of it was emerging disenchantment with Roosevelt's (and the Democratic Party's) foreign policies, which were to manifest themselves more fully in 1940 and 1944. Congressman William Lemke of North Dakota ran as the nominee of the right-wing Union Party that year, and he did well in counties with a heavy German and Scandinavian influence. I also think that the Dust Bowl may have had an impact, in reducing Roosevelt's margins on the Great Plains.

[Calthrina950]

There are none. According to Albert J. Menendez's Geography of Presidential Elections in the United States, not a single county shifted from Smith in 1928 to Hoover in 1932, while Roosevelt gained 1,791 counties which Hoover had carried in 1928. This is probably the only presidential election in which the losing party did not gain any counties. In 1964, Goldwater flipped over 500 counties in the South that either Kennedy or the States' Rights candidates (Orval Faubus and Harry F. Byrd) had won in 1960. In 1972, McGovern picked up several of Nixon's 1968 counties, such as Athens County, Ohio and Pitkin County, Colorado.

In 1984, Mondale won several Reagan 1980 counties, including Marin County, California and Denver County, Colorado. Parker in 1904, Cox in 1920, Davis in 1924, Landon in 1936, and Stevenson in 1952/56 all gained counties despite losing in landslides. 1932 had the greatest swing between parties of any presidential election. Hoover did gain several points in heavily Catholic Elk County, which Smith had flipped in 1928 (after it had been won in landslides by Harding and Coolidge).

Very interesting. That means there are Roosevelt 1932/Landon 1936 counties, too?

Yes. Major County, Oklahoma and Winston County, Alabama-two ancestrally Republican counties that Roosevelt managed to win in 1932-went back to Landon in 1936. Landon also flipped a substantial number of counties in Missouri, Kentucky, Iowa, the Dakotas, and Nebraska which Roosevelt had carried in 1932, and regained a number of counties in the Texas Hill Country (Texas was an all-county sweep for Roosevelt in 1932). Overall, Landon carried 203 counties which Roosevelt had won four years previously.

Why did Roosevelt win a number of these Southern ancestrally R counties in 1932 only to lose them in 1936 when he did better nationwide? (Winston AL, Newton and Searcy AR, Gillespie and Kendall TX, Major OK) Some of these swings were insane too, he won the Texas ones in 1932 by around 60 points only to lose them in 1936 by around 20. What happened here?

I don't know, other to say that normally Republican voters, disenchanted by Hoover (who was extremely unpopular in 1932), may have been turned off by aspects of the New Deal and returned to their ancestral party in the relatively more settled year of 1936.

[E-Dawg]

According to the Atlas Statistics on the 1932 election (https://uselectionatlas.org/RESULTS/stats.php?year=1932&f=1&off=0&elect=0) these were the 10 counties with the highest Republican swings between 1928-1932, with the amount GOP they swung.

Iberia, LA:  21.95%   
Iberville, LA: 20.34%   
St. John the Baptist, LA: 16.28%
Elk, PA: 13.68%
St. James, LA: 8.69%
St. Bernard, LA: 6.68%
West Baton Rouge, LA: 5.13%   
Zapata, TX: 4.21%   
Jackson, LA: 3.82%   
Lafourche, LA: 2.67%

(Note that these were all still Smith-Roosevelt counties despite the R swings). The Louisiana ones were all Catholic areas where Smith likely turned out many Catholic voters that Roosevelt failed to. There was also probably a group of ancestrally GOP Catholics who voted Smith due to perceived religious prejudice of the Hoovercrats but returned to their party in 1932 when religion wasn't an issue in the campaigns. Elk, PA is also a heavily Catholic area that probably shifted for the same two reasons.

The only one left is the now notorious Zapata, TX. Knowing that these South Texas counties were heavily machine-controlled at the time, the Democratic machine probably just was more effective in 1928 than in 1932. I have no idea how to otherwise explain the shift there. Considering that the county went from 89-10 GOP in 1916, to 66-34 GOP in 1920, to 60-40 DEM in 1924, to 94-6 DEM in 1928, to 92-8 DEM in 1932, this area was probably completely controlled by competing corrupt machines. None of these insane anti-bellwether swings seem natural or make clear sense otherwise.

Does anyone have access to a full list of counties that swung GOP in 1932? There can't be too many more than these 10 I imagine.

[If my soul was made of stone]

Part of it may have been a regional or "native-son" effect; Landon seems to have had considerable appeal to Appalachian voters, particularly voters in ancestrally Republican areas. Part of it was emerging disenchantment with Roosevelt's (and the Democratic Party's) foreign policies, which were to manifest themselves more fully in 1940 and 1944. Congressman William Lemke of North Dakota ran as the nominee of the right-wing Union Party that year, and he did well in counties with a heavy German and Scandinavian influence. I also think that the Dust Bowl may have had an impact, in reducing Roosevelt's margins on the Great Plains.

It's a bit disingenuous to label the Union Party with any particular ideology. Lemke was an old-school Progressive and an ally of Huey Long who broke with Roosevelt over his refusal to allow the Frazier–Lemke Farm Bankruptcy Act to be re-passed against a Supreme Court ruling against it; the Union Party consisted of those who were left holding the infrastructural bag after Huey Long's assassination prematurely ended his presidential ambitions. Some in the Union Party, such as Charles Coughlin, held far-right, anti-Semitic sympathies, and attempted to steer them in that direction, but the party was too fragmented and caught up in quarreling with itself to have much ideological consistency beyond vague economic populism and isolationism, being much of its appeal to Germans and Scandinavians in the Upper Midwest and Catholics in New England.

[Calthrina950]

Part of it may have been a regional or "native-son" effect; Landon seems to have had considerable appeal to Appalachian voters, particularly voters in ancestrally Republican areas. Part of it was emerging disenchantment with Roosevelt's (and the Democratic Party's) foreign policies, which were to manifest themselves more fully in 1940 and 1944. Congressman William Lemke of North Dakota ran as the nominee of the right-wing Union Party that year, and he did well in counties with a heavy German and Scandinavian influence. I also think that the Dust Bowl may have had an impact, in reducing Roosevelt's margins on the Great Plains.

It's a bit disingenuous to label the Union Party with any particular ideology. Lemke was an old-school Progressive and an ally of Huey Long who broke with Roosevelt over his refusal to allow the Frazier–Lemke Farm Bankruptcy Act to be re-passed against a Supreme Court ruling against it; the Union Party consisted of those who were left holding the infrastructural bag after Huey Long's assassination prematurely ended his presidential ambitions. Some in the Union Party, such as Charles Coughlin, held far-right, anti-Semitic sympathies, and attempted to steer them in that direction, but the party was too fragmented and caught up in quarreling with itself to have much ideological consistency beyond vague economic populism and isolationism, being much of its appeal to Germans and Scandinavians in the Upper Midwest and Catholics in New England.

I was using the label which was provided in the Geography of Presidential Elections, but I appreciate the explanation, and I'm certainly aware of Father Coughlin.

[Sailor Haumea]

There are none. According to Albert J. Menendez's Geography of Presidential Elections in the United States, not a single county shifted from Smith in 1928 to Hoover in 1932, while Roosevelt gained 1,791 counties which Hoover had carried in 1928. This is probably the only presidential election in which the losing party did not gain any counties. In 1964, Goldwater flipped over 500 counties in the South that either Kennedy or the States' Rights candidates (Orval Faubus and Harry F. Byrd) had won in 1960. In 1972, McGovern picked up several of Nixon's 1968 counties, such as Athens County, Ohio and Pitkin County, Colorado.

In 1984, Mondale won several Reagan 1980 counties, including Marin County, California and Denver County, Colorado. Parker in 1904, Cox in 1920, Davis in 1924, Landon in 1936, and Stevenson in 1952/56 all gained counties despite losing in landslides. 1932 had the greatest swing between parties of any presidential election. Hoover did gain several points in heavily Catholic Elk County, which Smith had flipped in 1928 (after it had been won in landslides by Harding and Coolidge).

Very interesting. That means there are Roosevelt 1932/Landon 1936 counties, too?

Yes. Major County, Oklahoma and Winston County, Alabama-two ancestrally Republican counties that Roosevelt managed to win in 1932-went back to Landon in 1936. Landon also flipped a substantial number of counties in Missouri, Kentucky, Iowa, the Dakotas, and Nebraska which Roosevelt had carried in 1932, and regained a number of counties in the Texas Hill Country (Texas was an all-county sweep for Roosevelt in 1932). Overall, Landon carried 203 counties which Roosevelt had won four years previously.

Why did Roosevelt win a number of these Southern ancestrally R counties in 1932 only to lose them in 1936 when he did better nationwide? (Winston AL, Newton and Searcy AR, Gillespie and Kendall TX, Major OK) Some of these swings were insane too, he won the Texas ones in 1932 by around 60 points only to lose them in 1936 by around 20. What happened here?

Not sure about the others, but Major County, OK is part of a block of counties in northwestern Oklahoma, anchored by the city of Enid, that were BEDROCK Republican ever since statehood. They were colonized by mostly Republican settlers from Kansas and the rest of the Great Plains, and were primarily economically dependent on wheat farming, etc. Contrast this with other political regions (the Panhandle, the southwest, the southeast, and the northeast), all of which favored Democrats until at least 1952, when piece by piece different regions detected to the GOP over time starting with the Panhandle. IIRC in the 60s and 70s Major County was the ONLY county in the state where registered Republicans outnumbered registered Democrats.

[TransfemmeGoreVidal]

There are none. According to Albert J. Menendez's Geography of Presidential Elections in the United States, not a single county shifted from Smith in 1928 to Hoover in 1932, while Roosevelt gained 1,791 counties which Hoover had carried in 1928. This is probably the only presidential election in which the losing party did not gain any counties. In 1964, Goldwater flipped over 500 counties in the South that either Kennedy or the States' Rights candidates (Orval Faubus and Harry F. Byrd) had won in 1960. In 1972, McGovern picked up several of Nixon's 1968 counties, such as Athens County, Ohio and Pitkin County, Colorado.

In 1984, Mondale won several Reagan 1980 counties, including Marin County, California and Denver County, Colorado. Parker in 1904, Cox in 1920, Davis in 1924, Landon in 1936, and Stevenson in 1952/56 all gained counties despite losing in landslides. 1932 had the greatest swing between parties of any presidential election. Hoover did gain several points in heavily Catholic Elk County, which Smith had flipped in 1928 (after it had been won in landslides by Harding and Coolidge).

Cox won counties in 1920 that Wilson didn't in 1916? That's honestly a surprise to me given that Harding's popular vote landslide was greater then FDR's.

[E-Dawg]

There are none. According to Albert J. Menendez's Geography of Presidential Elections in the United States, not a single county shifted from Smith in 1928 to Hoover in 1932, while Roosevelt gained 1,791 counties which Hoover had carried in 1928. This is probably the only presidential election in which the losing party did not gain any counties. In 1964, Goldwater flipped over 500 counties in the South that either Kennedy or the States' Rights candidates (Orval Faubus and Harry F. Byrd) had won in 1960. In 1972, McGovern picked up several of Nixon's 1968 counties, such as Athens County, Ohio and Pitkin County, Colorado.

In 1984, Mondale won several Reagan 1980 counties, including Marin County, California and Denver County, Colorado. Parker in 1904, Cox in 1920, Davis in 1924, Landon in 1936, and Stevenson in 1952/56 all gained counties despite losing in landslides. 1932 had the greatest swing between parties of any presidential election. Hoover did gain several points in heavily Catholic Elk County, which Smith had flipped in 1928 (after it had been won in landslides by Harding and Coolidge).

Cox won counties in 1920 that Wilson didn't in 1916? That's honestly a surprise to me given that Harding's popular vote landslide was greater then FDR's.

Yes, Polk County, North Carolina was a Hughes-Cox county, though it was very close in both elections. North Carolina was a whole barely shifted between these elections and there was a massive turnout increase, so I don't find this too odd. There are some sources claiming that Manistee County Michigan was a Hughes-Cox county too, but they are most likely false for several reasons. This issue was discussed in a preivous topic I made here: https://talkelections.org/FORUM/index.php?topic=415039.0

[Calthrina950]

There are none. According to Albert J. Menendez's Geography of Presidential Elections in the United States, not a single county shifted from Smith in 1928 to Hoover in 1932, while Roosevelt gained 1,791 counties which Hoover had carried in 1928. This is probably the only presidential election in which the losing party did not gain any counties. In 1964, Goldwater flipped over 500 counties in the South that either Kennedy or the States' Rights candidates (Orval Faubus and Harry F. Byrd) had won in 1960. In 1972, McGovern picked up several of Nixon's 1968 counties, such as Athens County, Ohio and Pitkin County, Colorado.

In 1984, Mondale won several Reagan 1980 counties, including Marin County, California and Denver County, Colorado. Parker in 1904, Cox in 1920, Davis in 1924, Landon in 1936, and Stevenson in 1952/56 all gained counties despite losing in landslides. 1932 had the greatest swing between parties of any presidential election. Hoover did gain several points in heavily Catholic Elk County, which Smith had flipped in 1928 (after it had been won in landslides by Harding and Coolidge).

Cox won counties in 1920 that Wilson didn't in 1916? That's honestly a surprise to me given that Harding's popular vote landslide was greater then FDR's.

Yes. He won Polk County, North Carolina, and Menendez's Geography of Presidential Elections states that he also won Manistee County, Michigan-both of which had gone to Hughes in 1916. There has been some dispute about Manistee County, however, with sources from the time period indicating that it actually went to Harding. 

[TransfemmeGoreVidal]

There are none. According to Albert J. Menendez's Geography of Presidential Elections in the United States, not a single county shifted from Smith in 1928 to Hoover in 1932, while Roosevelt gained 1,791 counties which Hoover had carried in 1928. This is probably the only presidential election in which the losing party did not gain any counties. In 1964, Goldwater flipped over 500 counties in the South that either Kennedy or the States' Rights candidates (Orval Faubus and Harry F. Byrd) had won in 1960. In 1972, McGovern picked up several of Nixon's 1968 counties, such as Athens County, Ohio and Pitkin County, Colorado.

In 1984, Mondale won several Reagan 1980 counties, including Marin County, California and Denver County, Colorado. Parker in 1904, Cox in 1920, Davis in 1924, Landon in 1936, and Stevenson in 1952/56 all gained counties despite losing in landslides. 1932 had the greatest swing between parties of any presidential election. Hoover did gain several points in heavily Catholic Elk County, which Smith had flipped in 1928 (after it had been won in landslides by Harding and Coolidge).

Cox won counties in 1920 that Wilson didn't in 1916? That's honestly a surprise to me given that Harding's popular vote landslide was greater then FDR's.

Yes. He won Polk County, North Carolina, and Menendez's Geography of Presidential Elections states that he also won Manistee County, Michigan-both of which had gone to Hughes in 1916. There has been some dispute about Manistee County, however, with sources from the time period indicating that it actually went to Harding. 

Maybe in Polk County, NC new woman voters ended up favoring Cox.

[TDAS04]

What about 1952?  Obviously there were Thurmond/Stevenson Counties.  Any Dewey/Stevenson counties?  I know Philadelphia was close, as Dewey almost carried it.

[E-Dawg]

What about 1952?  Obviously there were Thurmond/Stevenson Counties.  Any Dewey/Stevenson counties?  I know Philadelphia was close, as Dewey almost carried it.

Dawson County, Georgia was the only one in the country.

[Tartarus Sauce]

There are none. According to Albert J. Menendez's Geography of Presidential Elections in the United States, not a single county shifted from Smith in 1928 to Hoover in 1932, while Roosevelt gained 1,791 counties which Hoover had carried in 1928. This is probably the only presidential election in which the losing party did not gain any counties. In 1964, Goldwater flipped over 500 counties in the South that either Kennedy or the States' Rights candidates (Orval Faubus and Harry F. Byrd) had won in 1960. In 1972, McGovern picked up several of Nixon's 1968 counties, such as Athens County, Ohio and Pitkin County, Colorado.

In 1984, Mondale won several Reagan 1980 counties, including Marin County, California and Denver County, Colorado. Parker in 1904, Cox in 1920, Davis in 1924, Landon in 1936, and Stevenson in 1952/56 all gained counties despite losing in landslides. 1932 had the greatest swing between parties of any presidential election. Hoover did gain several points in heavily Catholic Elk County, which Smith had flipped in 1928 (after it had been won in landslides by Harding and Coolidge).

Very interesting. That means there are Roosevelt 1932/Landon 1936 counties, too?

Yes. Major County, Oklahoma and Winston County, Alabama-two ancestrally Republican counties that Roosevelt managed to win in 1932-went back to Landon in 1936. Landon also flipped a substantial number of counties in Missouri, Kentucky, Iowa, the Dakotas, and Nebraska which Roosevelt had carried in 1932, and regained a number of counties in the Texas Hill Country (Texas was an all-county sweep for Roosevelt in 1932). Overall, Landon carried 203 counties which Roosevelt had won four years previously.

Don't forget the bizarre, anomalous, and inexplicable Roosevelt-Landon flip of Assumption Parish, LA, which went right back to voting like normal in 1940.

[𝕭𝖆𝖕𝖙𝖎𝖘𝖙𝖆 𝕸𝖎𝖓𝖔𝖑𝖆]

There are none. According to Albert J. Menendez's Geography of Presidential Elections in the United States, not a single county shifted from Smith in 1928 to Hoover in 1932, while Roosevelt gained 1,791 counties which Hoover had carried in 1928. This is probably the only presidential election in which the losing party did not gain any counties. In 1964, Goldwater flipped over 500 counties in the South that either Kennedy or the States' Rights candidates (Orval Faubus and Harry F. Byrd) had won in 1960. In 1972, McGovern picked up several of Nixon's 1968 counties, such as Athens County, Ohio and Pitkin County, Colorado.

In 1984, Mondale won several Reagan 1980 counties, including Marin County, California and Denver County, Colorado. Parker in 1904, Cox in 1920, Davis in 1924, Landon in 1936, and Stevenson in 1952/56 all gained counties despite losing in landslides. 1932 had the greatest swing between parties of any presidential election. Hoover did gain several points in heavily Catholic Elk County, which Smith had flipped in 1928 (after it had been won in landslides by Harding and Coolidge).

Very interesting. That means there are Roosevelt 1932/Landon 1936 counties, too?

Yes. Major County, Oklahoma and Winston County, Alabama-two ancestrally Republican counties that Roosevelt managed to win in 1932-went back to Landon in 1936. Landon also flipped a substantial number of counties in Missouri, Kentucky, Iowa, the Dakotas, and Nebraska which Roosevelt had carried in 1932, and regained a number of counties in the Texas Hill Country (Texas was an all-county sweep for Roosevelt in 1932). Overall, Landon carried 203 counties which Roosevelt had won four years previously.

Don't forget the bizarre, anomalous, and inexplicable Roosevelt-Landon flip of Assumption Parish, LA, which went right back to voting like normal in 1940.

I think Assumption Parish's completely anomalous and very drastic flips (compare also 1920 and 1924) had mostly to do with disgruntlements about the state of the sugar trade, although I know zero about the topic so I can't give specifics. Its economy seems to be based overwhelmingly on growing sugar cane.