Response #3—that the two political parties have changed little over time and reflect their nineteenth-century origins—is the least persuasive and most problematic of the three possible responses to the student’s original question. As the rest of this post will show, this perspective is a misleading one propagated mostly by commentators who are not professional historians and which rests on a selective and out-of-context reading of facts, many of them trivial or anecdotal, that ignores quite a bit of countervailing evidence.
Is there ANY truth to Response #3?
It is rare when historical arguments are 100% false. Let's see if we can find any grain of truth in Response #3 before we dismantle it. In the presidential election of 1828, the John Quincy Adams campaigned portrayed Andrew Jackson as an “jackass,” or donkey, to describe the latter’s lack of formal education and unsuitability for the presidency. Jackson’s campaign actually embraced the donkey symbol and it has stuck to this day. The origins of the elephant imagery for the GOP go back to 1874, when famed political cartoonist Thomas Nash depicted them for Harper’s Magazine. The two parties still share these symbols but is there anything substantive beyond this surface-level similarity?
The most generous reading we can offer is that Republicans today are the party of Big Business, just as they were in the time of Lincoln and Herbert Hoover. Democrats have likewise maintained a mild suspicion of capitalism going back to the days of Jackson’s fight against the National Bank and William Jennings Bryan’s “Cross of Gold” speech. Presumably they stood for “the common man,” to take Jackson’s campaign rhetoric at face value.
Moreover, in the nineteenth century it was Democrats, not Republicans, who were more likely to embrace newly-arrived immigrants and a multi-ethnic society, which is still true today (though the immigrants were exiting different countries at the time). Many northern Whigs in the 1840s were reform-minded, anti-slavery evangelical Protestants who became Lincoln Republicans. White evangelicals are overwhelmingly Republican today, too, though the sectional orientation of this trend has shifted since the highest percentage of evangelicals today live in the South (see map below). Another interesting commonality is that Democrats today mostly accept the benefits of free trade just as their low-tariff, nineteenth-century predecessors did, though historically this path from past to present has not been linear. Think of how the Democratic Party was once more strongly associated with labor unions that tend to favor a more "closed" economy that adherents of free trade reject. It is worth adding that favoring free trade seems to be at odds with an anti-capitalist mindset.
It is rare when historical arguments are 100% false. Let's see if we can find any grain of truth in Response #3 before we dismantle it. In the presidential election of 1828, the John Quincy Adams campaigned portrayed Andrew Jackson as an “jackass,” or donkey, to describe the latter’s lack of formal education and unsuitability for the presidency. Jackson’s campaign actually embraced the donkey symbol and it has stuck to this day. The origins of the elephant imagery for the GOP go back to 1874, when famed political cartoonist Thomas Nash depicted them for Harper’s Magazine. The two parties still share these symbols but is there anything substantive beyond this surface-level similarity?
The most generous reading we can offer is that Republicans today are the party of Big Business, just as they were in the time of Lincoln and Herbert Hoover. Democrats have likewise maintained a mild suspicion of capitalism going back to the days of Jackson’s fight against the National Bank and William Jennings Bryan’s “Cross of Gold” speech. Presumably they stood for “the common man,” to take Jackson’s campaign rhetoric at face value.
Moreover, in the nineteenth century it was Democrats, not Republicans, who were more likely to embrace newly-arrived immigrants and a multi-ethnic society, which is still true today (though the immigrants were exiting different countries at the time). Many northern Whigs in the 1840s were reform-minded, anti-slavery evangelical Protestants who became Lincoln Republicans. White evangelicals are overwhelmingly Republican today, too, though the sectional orientation of this trend has shifted since the highest percentage of evangelicals today live in the South (see map below). Another interesting commonality is that Democrats today mostly accept the benefits of free trade just as their low-tariff, nineteenth-century predecessors did, though historically this path from past to present has not been linear. Think of how the Democratic Party was once more strongly associated with labor unions that tend to favor a more "closed" economy that adherents of free trade reject. It is worth adding that favoring free trade seems to be at odds with an anti-capitalist mindset.
All of these caveats cast doubt on the helpfulness of Response #3. Since nineteenth-century and modern Democrats have always had their own men of business, there is good reason to be skeptical. Most scholars working in the historiography of Jacksonian era politics and political economy have rejected Charles Sellers’s view that the Bank War was an “acid test” in the fight between democracy and capitalism. The Bank War, according to the entrepreneurial perspective put forth by the likes of Bray Hammond and Richard Hofstadter, was really about making Wall Street wealthy at the expense of Chestnut Street in Philadelphia. The manner in which Amos Kendall and Roger Taney appropriated some of the pet banks’ deposits to finance their own investments lends credence to this view. Indeed, the non-trivial amount of merchants who were against the Second Bank in my own work confirms that the Bank War showcased an anti-monopolist and anti-privilege mindset, but this was not the same as being anti-capitalist. That Trump can gain at least some political traction by arguing that Republicans stand for the “forgotten [white] man” who is “left behind” by “coastal elites” and the cozy relationship between Wall Street and Democratic bigwigs like Hillary Clinton and Chuck Schumer suggests that there is more disruption than continuity in the evolution of the Democrats’ core ideas and constituencies.
For the most part, Responses #1, #2, and #3 are mutually exclusive. Considering the totality, weight, and volume of evidence that supports Response #2, it could very well be that any commonalities between modern and nineteenth-century Democrats are purely coincidental and too broad to be meaningful. The burden of proof is on those who maintain, in spite of nearly unanimous scholarly opinion to the contrary, that there is continuity in the party systems between the antebellum era and today. Thus far the proponents of Response #3 have not met this burden and it could be that it is an impossible threshold to pass.
For the most part, Responses #1, #2, and #3 are mutually exclusive. Considering the totality, weight, and volume of evidence that supports Response #2, it could very well be that any commonalities between modern and nineteenth-century Democrats are purely coincidental and too broad to be meaningful. The burden of proof is on those who maintain, in spite of nearly unanimous scholarly opinion to the contrary, that there is continuity in the party systems between the antebellum era and today. Thus far the proponents of Response #3 have not met this burden and it could be that it is an impossible threshold to pass.
Party Support for the Civil Rights Act of 1964
Conspiracy theorist, conservative provocateur, Reagan devotee, and ex-felon Dinesh D’Souza has made a career out of propagating the dishonest claim that current-day Democrats are the party of fascists and the KKK, implying that Republicans are the true guarantors of civil rights. According to D’Souza, there was no party switch, no “Southern Strategy,” and Democrats have always been the real racists. Variations of this argument have appeared on the Twitter handles of other conservative activists like Carol Swain, Charlie Kirk, and Candace Owens.
D’Souza’s claims rest on a few cherrypicked facts. One of them is that a higher percentage of Republicans voted “yes” on the Civil Rights Act compared to Democrats. A little bit of scratching beneath the surface, courtesy of Princeton historian Kevin Kruse, will show that even these isolated facts should be placed with their appropriate historical context.
First, it is among the weakest intellectual defenses out there for an apologist of one political party to have to go back fifty-five years, during a tumultuous period in our nation’s past when the two parties were still in a state of flux and struggling to define themselves, to find the last time that this party took a praiseworthy stance on civil rights when mountains of accumulated evidence since then point to a very different historical record. But to the vote…yes, there were liberal and moderate Republicans, mostly from the North, who voted “yes” and there were conservative Democrats, mostly from the South, who voted “no.” Republican votes may have been crucial to the passage of the Civil Rights Act and Voting Rights Act, but they were junior partners in the process and soon marginalized within their party. The two landmark pieces of legislation were introduced by Democratic presidents, ushered through Congress by Democratic committee chairs and leaders, given more votes in the end by Democrats than the GOP, and then signed by a Democratic president. 221 Democrats joined 112 Republicans in voting “yes” on the VRA in the House while 49 Democrats and 30 Republicans voted “yes” in the Senate.
Opposition to the two laws was a major reason why Republicans secured significant gains in the 1966 midterm elections, upon which they built to win the presidency in 1968. Since then, the Republican Party has won the white vote in every single presidential election. Notably, when Republicans first started winning in the South, they were note like those liberal and moderate Republicans who backed the Civil Rights Act. The new Republicans behaved like Dixiecrats.
Response #3 collapses under the slightest bit of scrutiny because it relies almost exclusively on examining the partisan composition of Congress while ignoring everything else. A more holistic approach to this question would consider state legislatures, ideology, ordinary voters, governorships, party platforms, etc. Included below is a chart showing the partisan affiliation of each state legislature from 1960 to 2018. The states that are blue mean that the Democrats controlled both houses of the state legislature; the red ones indicate Republicans ran both houses; the gray color shows there was a split. As late as 1990, Republicans were limited to controlling both houses of the state legislatures in only six states: North Dakota, Wyoming, Utah, South Dakota, New Hampshire, and Colorado. None of these were southern states. Yet after the 2018 elections, Republicans controlled both houses of 30 state legislatures, including all 11 states in the former Confederacy. Two trends have added to the GOP dominance of state legislatures in the South and elsewhere. Citizens' United has given Republicans as much as a 5-point advantage in some state legislative races. The low-turnout midterm election of 2010 placed Republicans in the fortuitous situation of using the latest technology to draw state and national-level legislative districts to their advantage. In other words, they gerrymandered to lock themselves in power.
Conspiracy theorist, conservative provocateur, Reagan devotee, and ex-felon Dinesh D’Souza has made a career out of propagating the dishonest claim that current-day Democrats are the party of fascists and the KKK, implying that Republicans are the true guarantors of civil rights. According to D’Souza, there was no party switch, no “Southern Strategy,” and Democrats have always been the real racists. Variations of this argument have appeared on the Twitter handles of other conservative activists like Carol Swain, Charlie Kirk, and Candace Owens.
D’Souza’s claims rest on a few cherrypicked facts. One of them is that a higher percentage of Republicans voted “yes” on the Civil Rights Act compared to Democrats. A little bit of scratching beneath the surface, courtesy of Princeton historian Kevin Kruse, will show that even these isolated facts should be placed with their appropriate historical context.
First, it is among the weakest intellectual defenses out there for an apologist of one political party to have to go back fifty-five years, during a tumultuous period in our nation’s past when the two parties were still in a state of flux and struggling to define themselves, to find the last time that this party took a praiseworthy stance on civil rights when mountains of accumulated evidence since then point to a very different historical record. But to the vote…yes, there were liberal and moderate Republicans, mostly from the North, who voted “yes” and there were conservative Democrats, mostly from the South, who voted “no.” Republican votes may have been crucial to the passage of the Civil Rights Act and Voting Rights Act, but they were junior partners in the process and soon marginalized within their party. The two landmark pieces of legislation were introduced by Democratic presidents, ushered through Congress by Democratic committee chairs and leaders, given more votes in the end by Democrats than the GOP, and then signed by a Democratic president. 221 Democrats joined 112 Republicans in voting “yes” on the VRA in the House while 49 Democrats and 30 Republicans voted “yes” in the Senate.
Opposition to the two laws was a major reason why Republicans secured significant gains in the 1966 midterm elections, upon which they built to win the presidency in 1968. Since then, the Republican Party has won the white vote in every single presidential election. Notably, when Republicans first started winning in the South, they were note like those liberal and moderate Republicans who backed the Civil Rights Act. The new Republicans behaved like Dixiecrats.
Response #3 collapses under the slightest bit of scrutiny because it relies almost exclusively on examining the partisan composition of Congress while ignoring everything else. A more holistic approach to this question would consider state legislatures, ideology, ordinary voters, governorships, party platforms, etc. Included below is a chart showing the partisan affiliation of each state legislature from 1960 to 2018. The states that are blue mean that the Democrats controlled both houses of the state legislature; the red ones indicate Republicans ran both houses; the gray color shows there was a split. As late as 1990, Republicans were limited to controlling both houses of the state legislatures in only six states: North Dakota, Wyoming, Utah, South Dakota, New Hampshire, and Colorado. None of these were southern states. Yet after the 2018 elections, Republicans controlled both houses of 30 state legislatures, including all 11 states in the former Confederacy. Two trends have added to the GOP dominance of state legislatures in the South and elsewhere. Citizens' United has given Republicans as much as a 5-point advantage in some state legislative races. The low-turnout midterm election of 2010 placed Republicans in the fortuitous situation of using the latest technology to draw state and national-level legislative districts to their advantage. In other words, they gerrymandered to lock themselves in power.
The image above is proof of a very fundamental point that D’Souza and his acolytes have either failed to consider or ignored: the Southern strategy took a long time to pan out and it lasted far longer than Richard Nixon. The rollback of welfare benefits, deregulation of the financial sector, and reinforcement of “tough on crime” policies during the Clinton years might be plausibly viewed as a continuation of the Southern Strategy by a moderate Democrat. D’Souza posits that Richard Nixon could not have engaged in a Southern Strategy, in part, because there were plenty of southern Democrats in Congress in the 1990s. No academic historian denies the presence of these southern Democrats, and more importantly, their presence does not negate the notion that a Southern Strategy took place, especially when contemporaries like Kevin Phillips and Harry Dent wrote about it at the time.
Nor would we expect something as fundamental as a party switch to happen overnight, especially when one considers that it is common for parents to pass down party affiliation to their children. The maps included in Response #2 of this series demonstrate clearly that the increasing dominance of the GOP in the white South took place in fits and starts. Sometimes Democrats could briefly reverse this bleeding of white southerners from their party if they happened to run white southerners at the top of ticket ala Jimmy Carter and Bill Clinton. There were rumblings from white southerners who were uncomfortable with the Democratic Party as early as the 1930s. They often joined with Republicans in pushing back against FDR’s boldest reforms. Dixiecrats left the party in 1948 to support Thurmond, but returned for the next three presidential elections. Republicans made significant inroads in the electoral college in the South in 1964, based primarily on Barry Goldwater’s opposition to the Civil Rights Act, but this was the Deep South. Appalachia took a lot longer to turn red (see the map comparing Gore’s performance with Obama). An additional map created by the New York Times shows that even in the twenty-first century, the South has continued on a rightward trajectory.
Nor would we expect something as fundamental as a party switch to happen overnight, especially when one considers that it is common for parents to pass down party affiliation to their children. The maps included in Response #2 of this series demonstrate clearly that the increasing dominance of the GOP in the white South took place in fits and starts. Sometimes Democrats could briefly reverse this bleeding of white southerners from their party if they happened to run white southerners at the top of ticket ala Jimmy Carter and Bill Clinton. There were rumblings from white southerners who were uncomfortable with the Democratic Party as early as the 1930s. They often joined with Republicans in pushing back against FDR’s boldest reforms. Dixiecrats left the party in 1948 to support Thurmond, but returned for the next three presidential elections. Republicans made significant inroads in the electoral college in the South in 1964, based primarily on Barry Goldwater’s opposition to the Civil Rights Act, but this was the Deep South. Appalachia took a lot longer to turn red (see the map comparing Gore’s performance with Obama). An additional map created by the New York Times shows that even in the twenty-first century, the South has continued on a rightward trajectory.
As for Carol Swain’s emphasis on the fact that only 1 of the 21 Democratic US Senators who voted against the Civil Rights Act became a Republican (Strom Thurmond of South Carolina), Kruse has a very astute response. We would not expect members of Congress to switch political parties all that often because doing so would mean the loss of their seniority and status as a committee chairman in an era in which committee chairs exercised a lot of power. Thurmond was able to negotiate his way into keeping his seniority, but no other potential Republican could replicate the same deal. Eventually the Dixiecrats retired and their successors ran successfully as Republicans while maintaining the same opposition to federal civil rights legislation that Dixiecrats did. Such was the case with Trent Lott of Mississippi, who served as an aide to Democratic Representative William Colmer, head of the House Rules Committee. When Colmer retired in 1972, he handpicked Lott to fill the seat, but told him to run as a Republican. Lott did and won, eventually becoming Senate Majority Leader.
From Ideological Heterogeneity to Homogeneity: The Sorting of Liberals and Conservatives into Democrats and Republicans
To the extent that we can discern logical coherence in D’Souza’s statements, he seems to believe that a Democrat automatically equals a liberal, no matter what year we are examining. So if he can search through the semi-distant past to find a few examples of Democrats committing racist acts or making racist statements, he can imply that liberals have historically done these things. And if he believes that liberals have done racist things in the past, he can leap to the dubious conclusion that liberals today must be the real racists. Perhaps the most glaring flaw in this reasoning is that it misses the extent to which both Democrats and Republicans in the 1960s were ideologically fluid and heterogeneous organizations composed of liberal, moderate, and conservative members. Liberal Republicans in the North at this time could easily be more progressive on civil rights than conservative Democrats from the South. If one chooses to focus on the party affiliation of members of Congress—and again, this is only one piece of evidence among many—then the key to thing to understand is that liberals from both parties fought for civil rights at the federal level while conservatives opposed them.
To come at this from another vantage point, it may be more helpful to break down the votes on the Civil Rights Act in Congress according to North versus South, not Democrats versus Republicans. The two parties at this time were interregional coalitions with northern and southern members. There were Democrats who voted “no” on the Civil Rights Act because the party still clung to the vestiges of the “Solid South” of the Jim Crow Era. Indeed, the formal end to the Jim Crow era only came during 1964-65 with the passage of these landmark laws.
Today the two parties do not have this ideological diversity. If you’re a liberal, you’re almost assuredly a Democrat. And almost all conservatives are Republicans. The white South’s voters, a very large percentage of which are conservative, gradually transitioned from Democrats to Republicans because it became clear over time that the GOP, having banished its moderates, was the only conservative party.
This ideological sorting has been one of the most profound and consequential developments in the country’s political history in the last 50-60 years. It is related to the often-told characterization that Americans are now more politically divided than ever according to race, education, gender, ideology, religion, population density (rural, suburban, and urban), and even personality traits. It is the subject of a lengthy literature backed by innumerable graphs and polls. If D’Souza is to truly persuade rather than function as a professional troll who will be dismissed by any thoughtful person, he must demonstrate that he is aware of this literature of sorting.
When Does a New Model Successfully Replace an Old One?
History, like any academic discipline, thrives when long-cherished paradigms are continually challenged and reevaluated in good faith. Sometimes academics can make a name for themselves by overturning the previous model and by offering a new one. Media outlets and some scholarly journals, unfortunately, incentivize this type of work since an article that boldly overturns the decades of research propping up the old model is more likely to garner attention than the meticulous hard work that only confirms previous interpretations. Occasionally even reputable journals publish pieces that purport to overturn conventional wisdom even if the evidence upon which they rely is thin. This encourages us to ask, is D’Souza making a good faith attempt to overturn conventional wisdom? And how does one decide whether a new model can successfully replace the old one? Peer review, despite its many imperfections, is the answer.
I am confident that the arguments proffered by D’Souza et al (which are not made in good faith by the way) would not be published in high-ranking peer reviewed journals. “Preponderance of evidence” is a phrase we often hear in legal circles, but it may also apply to how historians make accurate statements out of complex events with conflicting pieces of evidence. If there are 100 different pieces of evidence in one direction—let’s call this the mainstream position that there was a party switch with regard to federal support for civil rights—and yet only 25 pieces that support D’Souza’s view, which we might call the contrarian view, the best we can say about the contrarian view is that the overall story is complex, but the white South switched from a Democratic to a Republican stronghold throughout the second half of the twentieth century. Saying anything less than this, or even worse, relegating it to a “myth” concocted by “left-leaning academic elites and journalists,” commits a grievous act of malpractice and mischaracterization of evidence. D’Souza et al. take the contrarian view because it places the political party they champion in the best possible light. But in doing so, they have ignored or minimized a much larger body of evidence that scholars have known about for a long time. They have failed to present enough evidence to overturn the mainstream response laid out in Response #2.
Here is the next post in this series.
From Ideological Heterogeneity to Homogeneity: The Sorting of Liberals and Conservatives into Democrats and Republicans
To the extent that we can discern logical coherence in D’Souza’s statements, he seems to believe that a Democrat automatically equals a liberal, no matter what year we are examining. So if he can search through the semi-distant past to find a few examples of Democrats committing racist acts or making racist statements, he can imply that liberals have historically done these things. And if he believes that liberals have done racist things in the past, he can leap to the dubious conclusion that liberals today must be the real racists. Perhaps the most glaring flaw in this reasoning is that it misses the extent to which both Democrats and Republicans in the 1960s were ideologically fluid and heterogeneous organizations composed of liberal, moderate, and conservative members. Liberal Republicans in the North at this time could easily be more progressive on civil rights than conservative Democrats from the South. If one chooses to focus on the party affiliation of members of Congress—and again, this is only one piece of evidence among many—then the key to thing to understand is that liberals from both parties fought for civil rights at the federal level while conservatives opposed them.
To come at this from another vantage point, it may be more helpful to break down the votes on the Civil Rights Act in Congress according to North versus South, not Democrats versus Republicans. The two parties at this time were interregional coalitions with northern and southern members. There were Democrats who voted “no” on the Civil Rights Act because the party still clung to the vestiges of the “Solid South” of the Jim Crow Era. Indeed, the formal end to the Jim Crow era only came during 1964-65 with the passage of these landmark laws.
Today the two parties do not have this ideological diversity. If you’re a liberal, you’re almost assuredly a Democrat. And almost all conservatives are Republicans. The white South’s voters, a very large percentage of which are conservative, gradually transitioned from Democrats to Republicans because it became clear over time that the GOP, having banished its moderates, was the only conservative party.
This ideological sorting has been one of the most profound and consequential developments in the country’s political history in the last 50-60 years. It is related to the often-told characterization that Americans are now more politically divided than ever according to race, education, gender, ideology, religion, population density (rural, suburban, and urban), and even personality traits. It is the subject of a lengthy literature backed by innumerable graphs and polls. If D’Souza is to truly persuade rather than function as a professional troll who will be dismissed by any thoughtful person, he must demonstrate that he is aware of this literature of sorting.
When Does a New Model Successfully Replace an Old One?
History, like any academic discipline, thrives when long-cherished paradigms are continually challenged and reevaluated in good faith. Sometimes academics can make a name for themselves by overturning the previous model and by offering a new one. Media outlets and some scholarly journals, unfortunately, incentivize this type of work since an article that boldly overturns the decades of research propping up the old model is more likely to garner attention than the meticulous hard work that only confirms previous interpretations. Occasionally even reputable journals publish pieces that purport to overturn conventional wisdom even if the evidence upon which they rely is thin. This encourages us to ask, is D’Souza making a good faith attempt to overturn conventional wisdom? And how does one decide whether a new model can successfully replace the old one? Peer review, despite its many imperfections, is the answer.
I am confident that the arguments proffered by D’Souza et al (which are not made in good faith by the way) would not be published in high-ranking peer reviewed journals. “Preponderance of evidence” is a phrase we often hear in legal circles, but it may also apply to how historians make accurate statements out of complex events with conflicting pieces of evidence. If there are 100 different pieces of evidence in one direction—let’s call this the mainstream position that there was a party switch with regard to federal support for civil rights—and yet only 25 pieces that support D’Souza’s view, which we might call the contrarian view, the best we can say about the contrarian view is that the overall story is complex, but the white South switched from a Democratic to a Republican stronghold throughout the second half of the twentieth century. Saying anything less than this, or even worse, relegating it to a “myth” concocted by “left-leaning academic elites and journalists,” commits a grievous act of malpractice and mischaracterization of evidence. D’Souza et al. take the contrarian view because it places the political party they champion in the best possible light. But in doing so, they have ignored or minimized a much larger body of evidence that scholars have known about for a long time. They have failed to present enough evidence to overturn the mainstream response laid out in Response #2.
Here is the next post in this series.
RSS Feed