On January 20, 2009, Hillary Clinton took the oath of office as the first female President of the United States. It was a glorious day, almost as festive an inauguration day as that in 1953, when Robert Taft was sworn in, or 1913, when Teddy Roosevelt took the oath of office, becoming the first President to serve three terms. Wait, you say, none of those individuals became President in those years? Well, you are absolutely right about that. In fact, none of them was even their political party’s presidential candidate those years, but here’s the thing: all of them could plausibly claim (and did at times) that they had garnered more votes in the nominating process than the candidate who received their party’s nomination. In 1912, Roosevelt was clearly favored by more rank-and-file Republicans than the party’s nominee, William Howard Taft; in 1952, Robert Taft received more primary votes than did Dwight Eisenhower; and in 2008, some supporters of Hillary Clinton argued that she in fact had won more popular votes nationwide in the Democratic primaries and caucuses than Barack Obama had.Footnote ¹ How is it that Roosevelt was not the Republican nominee in 1912 or that Clinton was not the Democratic nominee in 2008 then?

The short answer is that, under the rules governing the presidential nomination process in both major parties, the nominee is not selected simply by totaling up the popular vote in the primary elections and party caucuses held across the nation and figuring out which candidate got the most votes nationwide; rather, both parties use a state-by-state process in which party voters in each state select a certain number of delegates, each of whom is pledged to support one of the candidates and who attend the party’s national convention. Whichever candidate receives a majority of those national convention delegates becomes the nominee, but, as past nomination contests reveal, the candidate who wins a majority of the delegates need not be the candidate who won the most popular votes nationwide. The presidential nomination process is not ineluctably majoritarian, not even close.

For party insiders and aspiring candidates, that is hardly news, but, for most Americans, that should be at least somewhat surprising, and for scholars of the presidential election process in particular, that should be at least somewhat problematic. There are books and journal articles aplenty lamenting the Electoral College, decrying how a candidate can win the White House with fewer popular votes nationwide than their opponent. In 2000, George W. Bush received fewer popular votes than Al Gore, and in 2016, Donald Trump received fewer popular votes than Hillary Clinton. These “misfires” produced considerable academic commentary, almost all of it critical.Footnote ² They even spawned a well-intentioned but ultimately misguided reform effort, the proposed National Popular Vote Compact (NPVC).Footnote ³ Yet, for all the academic hand-wringing about the 2000 and 2016 presidential elections, there has been no similar outcry about nomination misfires. In fact, unlike the general election process, the nomination process in both parties did not include rank-and-file voters in most states until the 1970s. How’s that for democracy?

Despite the evident antidemocratic and anti-majoritarian features of the nomination process, academic scholars have paid the nomination process less attention than the Electoral College general election process. That is especially true of the history of the process. In the wake of the 1968 contest, both national parties reformed the process to open it to rank-and-file voters. The vast bulk of the academic literature about the process takes as its starting point those post-1968 reforms, giving just passing attention to the world that came before it.Footnote ⁴ At some level, that is understandable – the modern nomination process bears more similarities to that used in 1972 than 1872, but, if we want to understand the modern process, we need to look back further in time. Even as early as 1808, there were discussions about how the process used at that time could produce a nominee favored by only a minority of the party. Reform of the presidential nomination process has been almost continually ongoing since then (and is still ongoing today, a half-century after the post-1968 reforms). More importantly, some of the reforms adopted in the past are precisely the features that contribute to the antidemocratic or anti-majoritarian aspects of the process today.

In short, the system that we have today is the product of the selective accretion of nomination rules and processes over time, some of which trace their origin to debates at the Constitutional Convention in 1787. As such, to truly understand today’s nomination process requires understanding the nomination processes and rules of the past. That is the task of this book – to demystify the current presidential nomination process by exposing how we ended up where we have.

Making sense of the current system is not an easy task. The presidential nomination process today is one of the most complex and confusing electoral processes in the world. To begin with, there are over 100 regulatory bodies governing the process in each party. Congress has regulated the campaign financing of nomination contests; each national party has its own set of nomination rules regarding how national convention delegates are selected; each state has its own set of statutes governing the primary or caucus process in the state; and each state party has its own rules and customs. No other candidate for public office in the United States is selected through a process in which there are so many chefs in the regulatory kitchen.

The federal nature of the process has also made the process much more difficult to comprehend. Within the ambit allowed by each national party’s set of rules, each state (meaning both the state government via statute and the state party via rule or custom) sets the rules for its voters in the state. Predictably, different states have chosen different paths or processes: The date on which voters cast their ballot varies from state to state; whether and how an ordinary voter participates in the nomination contest varies from state to state; and how vote totals in each state are then aggregated and translated into national convention delegates for each candidate varies from state to state. Even the two parties within the same state often use different rules, meaning that a Republican and a Democrat in the same state regularly confront a different process and set of rules.

Moreover, the rules of each regulatory body can be incredibly complex. Take the national party rules, which are the most important of the bunch and form the basis of much of this book. In 1832, the Democrats felt content with just four, simple rules; today, they have twenty-two devoted just to the delegate selection process alone, and each of those has multiple subparts. In 1856, the Republicans were satisfied with just three rules; today, they have forty-two, many of which also have multiple subparts.Footnote ⁵ And, of course, the sheer number of rules is only one measure of the increasing complexity of the legal framework governing the process. The content of the rules themselves often defy easy explanation.

It is truly shocking that the nomination of a candidate who seeks the one nationwide office that is elected by and ostensibly represents all Americans owes their selection to such a byzantine amalgam of national and state rules, both statutory and party. If we were designing a presidential nomination system from scratch today, we would not adopt anything like it.Footnote ⁶ Nevertheless, make sense of it we must, for the selection of the two candidates vying for the most important governmental office on the planet is governed by these rules. And make sense of it we can, for most of the rules can be traced to some earlier nomination event that exposed a vulnerability or weaknesses in the process, which in turn prompted the party to adopt a new rule or modify an existing one. Once the history of the process is understood, the modern rules governing it become much more understandable – the past event or nomination fiasco that prompted adoption of a given rule gives color to the rule’s purpose and meaning. That is not to say that the current process is perfectly coherent or rooted in some internally consistent vision of party democracy, only that our historical inquiry will pay dividends in terms of helping us make sense of why we have the rules that we do and thereby understand why Roosevelt was not the Republican nominee in 1912 and why Barack Obama was the Democrats’ nominee in 2008.

* * * *

In one respect, this is a book about the presidential nomination process – specifically, about how the two parties have structured that process, both historically and today. Yet, on a broader level, it is also a book about the development of democracy in America. Who chooses the presidential nominees? And, equally importantly, how have the parties gone about answering that question?

It is tempting to answer the former by saying “the people,” but that was clearly not the case even arguably until the 1970s, and it is not even truly the case today either as a formal or practical matter. If you have any doubt about that, consider the fact that Kamala Harris was the Democratic nominee in 2024, even though her name did not appear on any primary election or caucus ballot in any state that year. (I discuss the 2024 Democratic contest in Chapter 9.) Rather, the presidential nomination process is an example of representative democracy or republicanism: the presidential nominee is selected by some group of individuals who themselves are accountable to ordinary voters. The people’s involvement in the selection of the nominees has always been and is still today indirect.

That does not necessarily condemn the process as antidemocratic – direct popular election is obviously one way for a democratic system to select its government officials, but it is not the sole way. All federal agencies heads, which include some of the most powerful individuals in the world, such as the chairperson of the Federal Reserve Board, are selected by someone other than the people, in this case the President, subject to approval by the U.S. Senate. The same is true for federal judges.Footnote ⁷ Indirect methods of selection abound in our system of government, but so long as the selecting entity is itself elected by or somehow accountable to the people, the demands of democracy are satisfied.Footnote ⁸

Yet, not all indirect systems of selection are the same; each will place the party voter at lesser or greater remove from the nomination choice and produce a different type of nominee. As the Framers recognized, the choice among the various forms of indirect selection matters. Viewed from a high enough perspective, the entire history of the presidential nomination process has been a struggle to identify which method of indirect selection – which representatives of the party voter – is best suited to selecting the presidential nominees. Are members of Congress best suited to the task? State legislators? National convention delegates? If the latter, who should choose them? State and local party leaders? Ordinary party voters? Some combination of both (and, if so, in what combination)? As we shall see, both parties have struggled with these questions, and even more surprisingly, they both continue to do so today.

Relatedly, how should the parties make that choice – by what criteria or metric should they compare and evaluate the various types of indirect selection? Why, for instance, should we prefer a system in which the nominee is made by popularly elected national convention delegates versus popularly elected members of Congress? What considerations are driving the preference for one over the other? At bare minimum, the process has to produce a single nominee – as discussed in Chapter 2, when the Democratic-Republican Party in the early nineteenth century entrusted the nomination to a decentralized group that nominated multiple candidates, it proved disastrous. Beyond that obviously tactical consideration – the process has to be sufficiently centralized as to yield one and only one candidate for each party – there has been little consensus within the parties regarding how to judge the various selection processes.

For instance, both parties obviously want a process that will choose a candidate who can beat the other party’s candidate and win the White House. While that was and is undoubtedly the goal of the nomination process, that has proven entirely unhelpful as a proposed criteria for choosing among the various forms of indirect selection. Each of the possible selectors – members of Congress, national convention delegates selected by state and local party leaders, and so on – have thought themselves the best at selecting a winner from among the available candidates, and none was so clearly superior on that score, at least not in any objective way, as to persuade the others of their inferiority. (As we shall see in Chapter 9, that this criteria has proven entirely unhelpful in the past has not stopped groups believing themselves to possess too little influence in the process today from claiming that they should be given greater influence in the process tomorrow so as to enable the party to do a better job picking a winner.)

The most often invoked criterion is one of democratic representation: Which group best represents the interests and preferences of the party voters? That criterion, while ultimately the most prominent one in party debates, begs difficult questions regarding who counts as a party member worthy of representation and how accurately the selection process must reflect and transmit the views of the relevant party electorate. To use just one historically salient example that convulsed the parties from the late nineteenth century to the mid twentieth century, were African-Americans party members? As we shall see in Chapter 5, at that time, neither the Republican nor Democratic state parties in the south thought so. More contemporaneously, is the relevant electorate all voters interested in participating in one of the parties’ primaries or caucuses, or just those voters who are long-standing members of the party? And last but not least in this regard, must the selection process ensure that the delegates mirror and proportionately reflect divisions within the electorate, or is it sufficient if the delegates represent only the views of a majority (or even less) of the party? As we shall see, the intraparty battles over which group of voters should participate in or drive the selection of the nominee implicates deeply contested questions of democratic representation that neither party has found easily tractable.

In piecing together the history of the nomination process, then, we are assembling a story about the development of democracy in America. Gradually, the nomination process became more open and democratic – more inclusive groups replaced more exclusive groups; the more inclusive groups themselves became more accountable to ordinary voters over time – but that progress was uneven and there were setbacks as well. Equally interestingly, the democratization of the process did not really take place until the 1970s, and even today, as I discuss in Chapter 8, the democratization of the process is incomplete in several important respects. The Electoral College preoccupies numerous legal scholars, political scientists, and newspaper editorial boards, but it is the two parties’ nomination process that provide the clearest (and perhaps most disturbing) window into the state of democracy in America.

* * * *

Finally, this is very much a book of legal history, but my goal is not knowledge of history for history’s sake. Nor is my goal simply to make sense of the present, though that is a vitally important task, and if this book does nothing more for the reader, I would feel satisfied. Rather, it is to enable all of us to think in a more sophisticated fashion about the future – about what the process should look like in 2028 and beyond.

No one today thinks that the current process is the best of all possible worlds. In recent decades, scholars of the nomination process have particularly worried that the modern voter-driven process is contributing to the ideological polarization of the nation. As political scientists have documented, the two major parties’ presidential candidates are more ideologically extreme now than they were in the past.Footnote ⁹ Recent Republican nominees have been more conservative than ones in the past; recent Democratic nominees have been more liberal. Whether that is because American voters are moving to the ideological poles or because our nomination rules favor the selection of more ideologically extreme candidates is hotly contested. Virtually all scholars of the process, however, agree something needs to be done about it – that the increasing ideological polarization of the presidential election process poses a danger to both the parties and nation as a whole.

While they agree that there is a problem, academic commentators disagree about the solution. Proposed solutions typically fall into one of two camps and involve diametrically opposite prescriptions for reform. In what I suspect will be a surprise to many readers, the prevailing view among scholars is that the nomination process has become too democratic and gives too much influence to ordinary voters.Footnote ¹⁰ In these scholars’ view, the process should be closed in whole or part, restoring influence to party elites like in the past so as (it is hoped) to moderate the ideological extremism of the party voters. On the flip side are a much smaller group of scholars (of which I count myself as one) who argue that the ideological extremism is the result not of too many voters participating in it but too few. For instance, tens of millions of voters, such as independents who comprise around 40 percent of the electorate, are excluded from the nomination process as a legal matter in most states and a practical matter in others. On this view, the nomination process has not been sufficiently democratized and needs to be opened even further.

In the Conclusion, I take up this debate and lay out my case for further democratizing reforms, but, as we will see, both sides in this debate rely to a great extent on history. Did the party-driven process in the past work well or not? Can its virtues be replicated without its flaws? And what of the people? How has the voter-driven process worked? Are the flaws in the current system the byproduct of the people’s role in it or something else? In a sense, there are few, if any, options for the presidential nomination process that have not already been tried or at least considered and debated in the past. As such, arguments about the future of the process inevitably implicate arguments about the past. It is there that we should therefore begin our inquiry.