A State-Level Policy Change That Would Revitalize the Electoral College
Abstract
The academic literature is rife with analyses of the US Electoral College’s flaws, but proposals to improve the system often rely upon old ideas. For example, the idea of replacing the Electoral College with a nationwide vote originated in 1816, and the derivative concept underlying the National Popular Vote Interstate Compact dates to 1976. Similarly, numerous methods for retaining the College but modifying the manner in which individual states select electors were proposed during the nineteenth and twentieth centuries, but the only one that gained significant traction – the congressional-district system currently used by Maine and Nebraska – was initially described in the 1950s by Senator Karl Mundt and Representative Frederick Coudert. This article describes the County-Elector Plan, a new approach that maintains the Electoral College but allocates a state’s electoral votes to each county’s plurality winner, in an amount proportional to the county’s voter turnout. A candidate’s statewide electoral vote total is then the rounded sum of the electoral votes the candidate receives in each county. The County-Elector Plan would seismically transform presidential elections by shifting an election’s focus from a handful of battleground states to hundreds of battleground counties spread across both current battleground and spectator states. Retrospective application of the plan to the 2016 Trump-Clinton contest shows that each candidate would have received electoral votes from 41 states, and that Clinton would have won the election by 26 electoral votes. The County-Elector Plan could be implemented on a state-by-state basis, without requiring a constitutional amendment. The plan is gerrymandering-resistant and provides all voters in a state with equal voting power.
1 Introduction
Although the Electoral College has successfully guided the US through 60 presidential cycles, its occasionally inverted results have produced a cacophony of calls for the system’s elimination, circumvention, or modification. Calls for the College’s elimination date back to 1816, when Senator Abner Lacock of Pennsylvania suggested the College be replaced by a nationwide popular vote (Spilerman and Dickens 1974, p. 444). The most direct way of eliminating the College would be via a constitutional amendment. However, embarking on such a journey in today’s political climate would be a fool’s errand, given that the dual hurdles specified in Article V of the Constitution would have to be surmounted. That is, absent a constitutional convention, a contemplated amendment must be proposed by Congress with a two-thirds vote in both houses and then ratified by three-quarters of the states before becoming part of the Constitution (US Const. art. V) “[O]ver 700 proposals have been introduced in Congress in the last 200 years to reform or eliminate the Electoral College system. Indeed, there have been more proposals for constitutional amendments to alter or abolish the Electoral College than on any other subject” (Strömberg 2008, p. 769). Elimination of the Electoral College would also strikingly shift the political paradigm, quite possibly for the worse. In speaking before the US Senate Judiciary Committee in 1970, Theodore White, political journalist and historian, saw a downside of the electoral system being replaced by a nationwide vote: “If States are abolished as voting units, TV becomes absolutely dominant. Campaign strategy changes from delicately assembling a winning coalition of States and becomes a media effort to capture the largest share of the national ‘vote market’” (Shelley 2002, pp. 82–3). Although White’s reference to a TV-based electioneering effort should now be broadened to encompass social media – everything from Facebook to TikTok to X and beyond – his point remains valid. A nationwide vote would further blur geographical and political boundaries, forever removing from American politics the concept of presidential candidates forming multi-state coalitions.
Following the inverted Bush–Gore contest of 2000, Robert Bennett, a law professor at Northwestern University, developed a plan to circumvent, not eliminate, the Electoral College. Building upon an idea that had been initially suggested by Read (1976), Bennett proposed that individual states agree to pledge their electoral votes to the winner of the nationwide popular vote. “If states with just 270 electoral votes adopted such an approach, the popular vote winner would perforce win the presidency” (Bennett 2001, p. 244). In 2006, John Koza, a Stanford computer scientist, began to promote the idea in the form of the National Popular Vote (NPV) plan, which is now more commonly called the National Popular Vote Interstate Compact (NPVIC) (Koza et al. 2013). As of February 2025, NPVIC legislation had been passed by 17 states and the District of Columbia, which collectively control 209 electoral votes. However, despite its apparent progress, the NPVIC’s future is highly uncertain, in large part because it may fall victim to the constraints imposed by the Constitution’s Compact Clause (US Const. art. I, § 10, cl. 3) (Bourgault and Schultz 2024). “If the NPVIC took effect, no noncompacting state could ever determine the outcome in presidential elections because only the compacting states’ electoral votes – cast together as a majority – would decide the winner. The noncompacting states’ sovereignty would severely suffer” (LeRoy 2024, p. 1490).
Paralleling the calls for the College’s abolition or circumvention has been a 200-year-long wave of proposals for its modification. Such proposals bifurcate based on whether a modification would require a constitutional amendment or whether it could be implemented at the state level. Constitutional changes, as noted above, are subject to Article V’s arduous path and, on that basis alone, are unlikely to be implemented in the foreseeable future, regardless of how well intentioned, popular, or necessary they may be. However, state-level changes to the manner in which electors are selected are manifestly viable and may provide a path to assuaging concerns that the College is outdated or unrepresentative of the wishes of the American people. Indeed, the Constitution leaves no doubt that the Founders wanted each state legislature to determine how the state chooses its electors. “Each State shall appoint, in such Manner as the Legislature thereof may direct, a Number of Electors, equal to the whole Number of Senators and Representatives to which the State may be entitled in the Congress[.]” (US Const. art. II, § 1, cl. 2). The first 14 presidential elections (1789–1840) saw the states experiment with ways to select electors. In the 1789 election, for example, six states (Connecticut, Delaware, Georgia, Massachusetts, New Jersey, and South Carolina) opted to have their legislatures select electors; three (Maryland, New Hampshire, and Pennsylvania) chose to select electors via a statewide popular vote; and one (Virginia) decided to divide the state into districts and hold a popular vote to select an elector from each district[1] (Martin 1958, pp. 1185–7; McPherson, v.Blacker 1892, pp. 29–32). These “experiments,” however, were not high-minded scientific endeavors; instead, they were bare-knuckled political maneuvers intended to enhance political power. Nowhere were these political machinations more volatile than Massachusetts, which switched selection methods eight times in the nine elections from 1792 to 1824. However, the states’ experiments eventually wound down, and the selection of electors by popular vote – either statewide or by district – has been the unbroken law-of-the-land since 1880.[2] (Martin 1958, pp. 1185–7; McPherson, v.Blacker 1892, pp. 30–1; Shane 2001, pp. 545–6).
Accompanying the states’ trend toward selecting electors by popular vote was the move toward “winner-take-all” (or “unit”) voting, in which all of a state’s electors are awarded to the candidate winning the plurality of the popular vote. This trend originated during the election of 1800, and, by 1836, all of the then 25 states had adopted the method. Since then, the unit system has been the vastly dominant method nationwide, with the only significant exceptions being that Maine and Nebraska switched to a district system in 1972 and 1992, respectively (Herbst 2012, pp. 226–30). However, the unit method is beset by well-chronicled problems. For example, by delivering all of a state’s electoral votes to the plurality winner in an election, the unit system completely disregards all votes cast for losing candidates. This facilitates de facto disenfranchisement of a significant portion of the electorate due to the focus of elections on a small set of battleground (or swing) states.[3],[4] That is, although battleground states are a statistical certitude of the Electoral College, their importance in an election is amplified by the unit system, which delivers a big boost to the winning candidate in a state, regardless of the margin of victory. By necessity, then, presidential campaigns focus resources on battleground states, while paying only fleeting attention to the remaining states and the District of Columbia – the so-called “spectator” states – where the election results are largely pre-ordained (Gringer 2008, pp. 221–3). This situation effectively disengages voters in spectator states from the presidential election and consequently dampens turnout, which, in turn, can affect the outcome of down-ballot races, as well as local and state propositions and measures that are on the ballot.
In sum, the electoral power held by a small number of battleground states adversely distorts the freedom and fairness that should characterize presidential elections. This article proposes a solution: the County-Elector Plan, a novel, district-based approach to selecting electors that creates competitive electoral districts in more than 40 states, is gerrymandering-resistant, and provides all voters in a state with equal voting power. The County-Elector Plan can be implemented on a state-by-state basis, thus avoiding the nationwide morass of political, legal, and societal challenges that have historically prevented modification of the Electoral College via a constitutional amendment and that would most certainly again arise in today’s highly fractured political environment. A direct derivative of the plan would be the large-scale revamping of political marketing efforts, whereby funds and staff would be drawn away from a relative handful of statewide campaigns and refocused on a multiplicity of campaigns involving single counties or contiguous groups of politically similar counties.
The next section reviews the literature on the two leading candidates for replacing the unit system: the district plan that Maine and Nebraska use and the Lawrence Proportional Plan, which was first proposed in 1848 but that has never been implemented.[5] Then, the three succeeding sections respectively describe the proposed County-Elector Plan, discuss its potential challenges, and analyze its retrospective application within the context of the 2016 Trump-Clinton contest.
2 Literature Review
Academics and politicians have long looked toward district systems and proportional systems to displace the state-level unit rule. However, both district and proportional systems are limited in the ways they can be implemented. District systems, for example, can be viably based on just a few types of districts, such as congressional districts (as currently used in Maine and Nebraska), special electoral districts (as used in Virginia in 1789, Massachusetts in 1792, and Michigan in 1892), and counties (as suggested in this article). Beyond that, however, lies a diverse collection of geographical divisions that are either inherently flawed (e.g., city or town districts would leave large rural portions of the country unrepresented) or seemingly absurd (e.g., ZIP Codes). Proportional systems are likewise limited in their possibilities: they can vary by distributing electors to all candidates or just to the top-performing candidates, or they can vary in the way they impose rounding rules. Beyond those parameters, few variations are reasonable.
2.1 Maine-Nebraska (District-Popular) Plan
District plans trace their origin to the country’s first presidential election, when Virginia “was divided into 12 separate districts and an elector elected in each district[.]” (McPherson, v.Blacker 1892, p. 29). Virginia and other states used districts in subsequent elections, but because the method can split a state’s electoral votes among multiple candidates, the practice was gradually “abandoned due to the lack of influence the states had in the election compared to those who employed the winner-take-all method”[6] (Canady 2014–15, p. 87).
Massachusetts developed a hybrid district plan – the district-legislature plan – and used it in the elections of 1796, 1812, and 1820. Under the plan, one electoral vote was awarded to the popular-vote winner in each congressional district and the Massachusetts state legislature awarded two electoral votes. The plan flowed directly from the constitutional equation used to calculate the number of a state’s electoral votes: one electoral vote for each congressional seat and two additional (“Senate bonus”) votes (Martin 1958, p. 1186; McPherson, v.Blacker 1892, pp. 31–2). In the late 1950s, Senator Karl Mundt (R–South Dakota) and Representative Frederick Coudert (R–New York) modified the district-legislature approach by having the two Senate-bonus votes go to the winner of the statewide vote, thus creating a district-popular plan. Although the Mundt-Coudert model was debated in Congress, it failed to garner significant congressional support (Bugh 2016, pp. 14–6). Its sponsors then reformulated the plan to something akin to Virginia’s original plan, in which states were divided into special districts, with one elector selected by the popular vote in each district. That plan, too, faltered and failed (Kefauver 1962, pp. 196–9). However, Maine, acting on its own, implemented the equivalent of the initial Mundt-Coudert plan in 1972, and Nebraska followed suit starting with the 1992 election. Those states still use the method, which is now widely known as the Maine-Nebraska Plan, but the other 48 states and the District of Columbia currently use the unit rule. (Professor Robert Turner (2005) provides an insightful analysis of the Maine-Nebraska Plan.)
The literature contains several proposals to modify the Maine-Nebraska Plan. For example, in a 2012 article, law professor Craig Herbst introduced a district-popular hybrid, in which the “district” component allots one electoral vote to each congressional district (as in the Maine–Nebraska Plan), while the “popular” component allocates one electoral vote to the winner of the statewide popular vote and one electoral vote to the winner of the nationwide popular vote (Herbst 2012, p. 239). To date, however, no state has seriously considered implementing Herbst’s concept.
2.2 Lawrence Proportional Plan
At their fundamental level, proportional plans are simple and straightforward: they award a state’s electoral votes in a manner that is proportional to each candidate’s statewide vote total. Proportional systems were introduced into American politics by Congressman William T. Lawrence (Whig–NY), who proposed an amendment for nationwide proportional voting in 1848. Although Lawrence’s amendment failed to solicit significant attention, pure proportional systems often carry his name, at least among American academics and informed politicians (Ames 1897, p. 95).
Law professor Gardner (2000, p. 91) has observed that during first half of the twentieth century proportional representation was used at the local government level in “nearly two dozen American cities,” but that by 2000, it was “used only in city council and school committee elections in Cambridge, Massachusetts, and in school board elections in New York City.” Despite its aura of apparent fairness, proportional representation has grown only modestly in local American elections since Gardner’s observation, although the system has gained traction in several other parts of the world, most notably Tasmania, where it has been used in statewide elections since 1909[7],[8] (Green 2006).
2.3 Challenges of District and Proportional Plans
All district and proportional plans, regardless of their implementation-specific details, target the unit rule, seeking to replace the rule’s inherent inequity with a system that produces electoral results that accommodate the political choices of a broader spectrum of voters. As discussed below, however, both district and proportional plans suffer from defects that make their ultimate implementations imperfect.
District systems are susceptible to abuse via gerrymandering, a political tactic that originated in Pennsylvania in 1705, more than 100 years before the portmanteau was coined to describe Governor Elbridge Gerry’s remapping of Massachusetts state senate districts.[9] (Griffith 1907, pp. 26–8; Wang 2016, p. 1263) Gerrymandering as a potential problem was also acknowledged within the context of the Mundt-Coudert district-popular plan when that plan was first proposed in the late 1950s (Kefauver 1962, p. 199). However, the problems of district systems extend beyond gerrymandering, with attorney Hoffman (1996, p. 1013) observing, for example, that “even the fairest possible district system results in significant inequalities among voters. Congressional districts, for example, must be roughly equal in total population. The number of eligible voters in each district may vary significantly, and the proportion of those voters who actual vote in any given election may fluctuate wildly.” As is described later, the County-Elector Plan, despite being a district plan, is not plagued by these problems.
By definition, proportional plans, absent rounding algorithms and vote thresholds, exactly reflect the popular vote. Even with reasonable rounding rules, proportional plans closely approximate the popular vote. Vote thresholds, however, can significantly disturb the paradigm by removing minor-party candidates from contention, despite the fact that minor-party candidates are often the ones who stand to gain from proportional systems. “Perhaps the greatest benefit of the proportional system, according to its adherents, is its ability to represent minorities as well as majorities.” (Wagner 2006, pp. 586–7). The fact that proportional systems are so infrequently used in the US is a sign that their initial appeal may mask underlying disadvantageous operational outcomes.
In sum, both district and proportional systems suffer from significant defects and are consequently far-from-perfect alternatives to the unit rule. As is detailed in the remainder of this article, the County-Elector Plan is an alternative that averts the primary defects of both district and proportional systems and is better able to represent both majority and minority candidates.
3 A Novel State-Level Method of Assigning Electors: The County-Elector Plan
The County-Elector Plan works within the existing national-level Electoral College framework, and it concerns only how individual states allocate their electoral votes to presidential candidates. Under the County-Elector Plan, the plurality winner in each county receives the same portion of the state’s electoral votes as the portion of the state’s presidential votes that were cast in the county. For example, if 12 % of Florida’s votes were cast in Broward County, then that county’s winner would receive 12 % of Florida’s 30 electoral votes (3.6 electoral votes). This total is despite the fact that Broward County has only 8.9 % of Florida’s population and 8.6 % of its registered voters (as of 2022). That is, the number of electoral votes a county receives is a function of the number of the county’s voters who actually voted for a presidential candidate. As such, the method produces a fairer result than an assignment based on the total population or on the number of registered voters.[10]
The County-Elector Plan is to be contrasted with other district models, which largely assign electoral units to districts prior to an election. In the Maine-Nebraska model, for example, each district has one electoral vote, and that vote is awarded to the winning candidate regardless of voter turnout. The County-Elector Plan is novel. In it, counties are assigned electoral votes only after a state’s votes are fully tallied. Hence, higher voter turnout in a given county, relative to other counties, translates directly into additional electoral votes for that county. A beneficial aspect of the plan is that each voter in the state has equal voting power, unlike other district systems, where voting power varies from one district to another, due to differences in population (or in voter turnout, if that measure of voting power is used).
To illustrate the County-Elector Plan, consider the straightforward example of Delaware – a state that has three electoral votes and, coincidentally, three counties. Under the plan, if one of Delaware’s counties accounted for 50 % of the statewide votes in a presidential election, then the candidate who won the plurality of that county’s votes would receive 50 % of the state’s three electoral votes (=1.5 electoral votes). The losing candidate(s) would receive nothing from the county.
Table 1 shows the votes tallied in each of Delaware’s three counties during the 2016 presidential election, which pitted Democrat Hillary Rodham Clinton against Republican Donald John Trump. New Castle County, for example, accounted for 59.25 % of the popular votes cast statewide.[11] Under the proposed plan, the candidate who won the plurality of New Castle’s votes (Clinton) would have received 59.25 % of Delaware’s three electoral votes – a total of 1.778 electoral votes – and Trump would have received nothing. Similarly, Trump, who captured pluralities in Delaware’s other two counties – Kent and Sussex – would have received 0.505 and 0.717 electoral votes from those counties, respectively, and Clinton would have received nothing. Summing across all three counties, Clinton and Trump would have received 1.778 and 1.222 electoral votes, respectively. After rounding these totals to the nearest integer, Delaware would certify to the U.S. Congress that Clinton had won two of the state’s electoral votes and Trump had won one vote. This result is to be contrasted with Delaware’s current unit system, under which Clinton won all three of the state’s electoral votes, despite the fact that Trump received 41.7 % of the popular vote.[12]
County-by-county allocation of electoral votes under the County-Elector Plan in Delaware for the 2016 election.
| County | Popular votes | Electoral votes (County-Elector Plan) | ||||
|---|---|---|---|---|---|---|
| Clinton | Trump | Others | % of Statewide votes | Clinton | Trump | |
| New Castle | 162,919 | 85,525 | 14,535 | 59.25 | 1.778 | – |
| Kent | 33,351 | 36,991 | 4,387 | 16.84 | – | 0.505 |
| Sussex | 39,333 | 62,611 | 4,162 | 23.91 | – | 0.717 |
| Total | 235,603 | 185,127 | 23,084 | 100.00 | 1.778 | 1.222 |
The County-Elector Plan also tracks the popular vote more closely than the unit plan. Table 2 illustrates this effect within Delaware in the 2016 presidential election. Note that the “% of Popular Votes” values are much closer to the “% of Electoral Votes” calculated by the County-Elector Plan than by the state’s current unit system.
Comparison of the County-Elector Plan and the unit system in Delaware for the 2016 election.
| Candidate | % of Popular votes | % of Electoral votes | |
|---|---|---|---|
| County-Elector Plan | Unit system | ||
| Clinton | 53.1 | 66.7 | 100.0 |
| Trump | 41.7 | 33.3 | 0.0 |
| Others | 5.2 | 0.0 | 0.0 |
3.1 The County-Elector Plan Encompasses the Entire Voting Population
A fundamental requirement for instituting a district-based voting model is that the districts collectively encompass the nation’s entire geographical extent. In the County-Elector Plan, counties and county equivalents (i.e. Louisiana’s parishes and Alaska’s boroughs) collectively satisfy that requirement. In contrast, plans based upon other governmental administrative districts generally fail to do so, leaving significantly sized swaths of numerous states disenfranchised. Governmental administrative districts that do not encompass the country’s entire geography and thus are not appropriate for use as the nationwide basis of voting districts include cities, towns, hamlets, villages, consolidated city-counties (Virginia), townships, neighborhoods (Connecticut), sections of town (Connecticut), councils of governments (Connecticut), census-designated places, and special districts (e.g., tax, school, fire, police, sanitation, water, sewer, and library). Also, although “census tracts” collectively encompass the country’s entire geographical extent and could be used as the basis of the proposed plan, the public is largely unaware of their definition or even of their existence.
3.2 The County-Elector Plan is Built upon an Existing Electoral Framework
The vast majority of states use county governments to administer presidential elections. The County-Elector Plan leverages this existing electoral framework and can consequently be implemented without significant additional overhead.
3.3 The County-Elector Plan Is Resistant to Legal Challenges
As noted above, Article II of the U.S. Constitution grants state legislatures wide-ranging discretion to determine how electors are chosen, and the states’ exploitation of this power dates to the election of 1789, when the 10 states that selected electors did so using three distinctly different methods. Article II’s language is so clear – and its application so firmly established – that it is essentially immune from legal challenges. For example, courts have long deferred to Article II’s language even when faced with the obvious voter-impact imbalances inherent in the Maine–Nebraska Plan. That the County-Elector Plan is not afflicted by such imbalances further hardens it against legal attack on federal grounds.
However, since the County-Elector Plan is a state-level methodology, it may be subject to legal attack on state constitutional grounds or as a violation of state law. For example, as Banzhaf (1964, 1966) noted in his first two papers on voting power, courts interpreting New Jersey and Nebraska laws have disallowed weighted voting systems in the selection of state representatives. “The first example came from the New Jersey Senate in which a resolution was passed to implement a weighted voting plan. Similar to the Electoral College, the plan allocated a number of votes to each of its 21 members in proportion to the population of the county that each represented. The plan was never implemented because it was found to not be in compliance with New Jersey state law” (Wenschhof 2014). Although the County-Elector Plan is not a weighted voting system – and hence does not have any obvious vulnerability to an attack on equal representation grounds – it is not necessarily immune to an attack asserting a novel theory involving state-specific rights or laws.
3.4 The County-Elector Plan Is Resistant to Gerrymandering
Inherent in their definition and existence is that the country’s 435 congressional districts do not have permanent geographical boundaries. Rather, following the completion of each decennial census, the entire congressional geography is subject to reapportionment: new districts may be created, and existing districts are subject to realignment or elimination. Although the gerrymandering of congressional districts has long been a problem, the situation could be exacerbated if congressional districts were used for the additional purpose of selecting electors. Discussion of such “presidential gerrymandering” arose with the introduction of the Mundt-Coudert model and has been analyzed within the context of the district systems currently used in Maine and Nebraska (Kefauver 1962, p. 199; Read 1976, p. 328). Clayton (2007, p. 36) asserts that there “is an even greater potential for partisan gerrymandering now that the Supreme Court has stated that congressional redistricting may occur more than once a decade.”
In vivid contrast to the short-lived geographical boundaries of many congressional districts, the boundaries of US counties are near-permanent fixtures of American government. Although counties are theoretically subject to creation, realignment, and elimination, county governments in the vast majority of states are large, complex bureaucracies that conduct a significant amount of governmental business, and, resultingly, are hardened against gerrymandering. Indeed, the country’s five most populous states (California, Texas, Florida, New York, and Pennsylvania), which collectively have 37 % of the country’s population, have not undergone a significant change to their county boundaries since 1925. Hence, the County-Elector Plan’s use of counties as the basic electoral district reduces gerrymandering as a threat to the long-term integrity of the electoral system.[13]
3.5 A County’s Electoral Votes are a Function of Voter Turnout
Consider a hypothetical state that is divided into districts for the purpose of awarding electoral votes. If the districts are congressional districts, each of which awards one electoral vote to the plurality winner of the district’s popular vote, then a candidate in a two-person race would be indifferent to winning 51 % or 100 % of any district’s vote. Thus, there would be no incentive to expend campaign resources in excess of that required to have an excellent chance of winning a simple majority of the district’s vote. On the other hand, under the proposed County-Elector Plan, higher turnout in a given district, relative to other districts, translates directly into an increase in that district’s allocated electoral votes, all of which would go to the candidate receiving a plurality of the district’s popular vote. For example, in a district that leans very heavily to one party and is a lock for that party’s presidential candidate, the candidate would still be incentivized to devote additional campaign resources trying to increase voter turnout.[14]
3.6 The County-Elector Plan Makes More States Competitive
The County-Elector Plan has the potential to reinvigorate the Electoral College. Turner (2005, p. 116) notes that the “ultimate desirability” of a “district system lies in how it would change the conduct of presidential campaigns, not the counting of Electoral College votes. Under the district system, presidential campaigns would shift their priorities from battleground states to battleground districts.”[15] This effect was evident in Nebraska’s Second Congressional District during the 2020 election, where highly concentrated marketeering secured the district’s electoral vote for Biden, reversing the outcome of the 2016 election, when Trump won the district.
However, not all types of districts are created equal; the inherent nature of some types of districts makes them well suited to changing the campaign paradigm from battleground states to battleground districts, while other types of districts are ill suited for doing so. For example, despite their apparent successes in Maine and Nebraska, congressional districts are largely unappealing candidates to become battleground districts. Congressional districts exist for the extremely limited purposes of selecting representatives and, only in Maine and Nebraska, selecting electors. This limited functionality, coupled with the instability arising from the census-to-census threat to their existence, provides no significant basis for such districts to generate any consequential level of voter allegiance. That is, although voters may have some manner of allegiance to their congressional representative, they have minimal, if any, allegiance to their congressional district, itself. This situation is to be contrasted with counties, which are long-standing legal frameworks that serve multiple administrative functions and frequently engender citizen familiarity, dependence, and some level of vicarious identity. To citizens, counties are more tangible and less amorphous than congressional districts and, as such, provide a solid basis that candidates could use to exploit voter loyalty and enthusiasm to create localized election battlegrounds. Katz (1997, p. 175) emphasizes the benefits to be derived from local political districts. “[A] sense of community is fostered by the flourishing of a series of institutions all serving the same people. In the narrowly political sphere, people participating in the politics of the same town council district should also be in the same regional district and the same national parliamentary district.” The use of counties as voting districts would exploit regional loci of solidarity that simply could not arise from congressional districts or other purely political divisions. As noted above, counties also have the advantage of being the administrative centers for federal and state elections in most states.
Since the County-Elector Plan directly translates a county’s increased voter turnout into an increase in that county’s electoral votes, presidential campaigns would be forced to shift the focus of their efforts from battleground states to a broad set of battleground counties, including counties in states that have historically not been competitive. For example, a Democratic presidential candidate might target Harris County, Texas, which has leaned toward the Democratic Party in the last five presidential elections. If the county held to its historical average of 16 % of the state’s voters, a Democrat winning a plurality there would garner 6.4 electoral votes (equal to 16 % of Texas’s 40 electoral votes), and the Republican candidate would receive none. In contrast, under the state’s current unit system, Harris County is unlikely to attract significant attention from a Democratic candidate because any increase in votes in the county would be insufficient to overcome the Republican surplus generated in other parts of the state.
Similarly, a Republican candidate might place some focus on Suffolk County, New York, which had 8 % of New York State’s voters and gave Trump 51 % of the county’s votes in 2016. In a future election with similar turnout, ensuring that the Republican candidate won the plurality in the county would result in about 2.24 electoral votes (equal to about 8 % of the state’s 28 electoral votes) for the Republican candidate and none for the Democrat. In comparison, under New York State’s current unit system, Suffolk County would not attract any attention from the Republican candidate, given the state’s overwhelming preference for Democratic presidential candidates.
The key to the County-Elector Plan is that the plurality winner gets all of a given county’s electoral votes. Hence, a solid strategy for a party would be to target counties where it has a small advantage and where it believes it could maintain or increase that advantage. The strategy would increase the number of politically competitive states. Turner (2005, p. 134) asserts that district systems also increase voter participation. “Under the district system, presidential candidates would have to campaign in more states and build a broader geographical electoral coalition that under unit systems. The change in campaign strategy would presumably increase citizen participation and voting in presidential elections.” It is this shift from battleground states to battleground counties under the County-Elector Plan that could transform campaign paradigms and reinvigorate the Electoral College.
3.7 The County-Elector Plan Provides Equal Voting Power to All Voters in a State
Analysis of the voting power of individual states within the context of the Electoral College dates back at least to the pioneering game-theoretic research by RAND Corporation mathematicians Irwin Mann and Lloyd Shapley in the 1950s and early 1960s (Mann and Shapley 1960, 1962). Several years later, John Banzhaf expanded upon this early work by analyzing the “voting power” of individual citizens, where voting power is the probability of a single voter being pivotal. (Banzhaf 1964, 1966, 1968) Banzhaf’s work was initially performed within the context of the Mundt-Coudert plan, and it was criticized “because it ignores interdependencies in the way district and state electoral votes may be cast – in particular, while individuals are casting statistically independent votes, the fact that each is casting a vote that counts in two different tiers induces a correlation between popular votes at different levels” (Miller 2014, p. 201). Despite its limitations, Banzhaf’s work has formed the foundation of numerous subsequent mathematical analyses of voting power in US presidential elections, the UN Security Council, and other forums (Davis et al. 2009; Dehez 2024; Dubey and Shapley 1979; Heilman 2021; Kaniovshi 2008; Lucas 1983; Margolis 1983; Pal 2021).
When voting-power methodologies are applied to state-level district systems, voting power is defined as the influence wielded in the selection of electors (Banzhaf 1968, pp. 320–1). In congressional district systems, districts are analogous to states. Since modern-day, intra-state congressional districts are nearly equal in population (Streb 2008, p. 100), these models show that all of a state’s voters have nearly equal voting power, with differences attributable to small variations in district populations. However, since voting-power models are inherently complex, analyses of voting power are often simplified by incorporating two assumptions that are known to be false: that all residents of a given district are eligible to vote and that those residents will vote. In reality, many residents are ineligible to vote because of age, criminal history, or citizenship status, while a significant portion of eligible voters decline to vote. Consequently, district-to-district differences in turnout introduce additional factors into the comparison of voting power between residents of congressional districts within the same state. These differences are often significant. In the 2016 presidential election, for example, Maine’s two congressional districts differed by 12 % in turnout, although their populations differed by only 1 % (Bangor Daily News 2011; Leip 2022). Hence, individual voters in Maine’s lower-turnout district had a greater impact on selecting that district’s elector than individual voters in the other district had in selecting their elector. Under the County-Elector Plan, all voters within a given state have equal voting power, regardless of either district-to-district variations in the number of residents who are eligible to vote or the number of residents who do vote.
Under the County-Elector Plan, each county’s (fractional) number of electoral votes is determined only after the state’s election results are tabulated. At that time – which could be just seconds after the tabulation is completed – the state’s electoral votes are distributed to each county in an amount proportional to the county’s turnout, and each county’s votes are awarded to the county’s plurality winner. Consequently, under the County-Elector Plan, all voters in the state have identical voting power. This result simply cannot be achieved under the Maine-Nebraska system or other district-popular plans that assign electoral votes to a district prior to an election.[16],[17] Again, it must be noted that voting-power equality under the County-Elector Plan applies only on a state-by-state basis; that is, all voters in a given state have equal voting power, but that voting power is not necessarily the same as the voting power of voters in other states. Mathematically, the voting power of a voter in a given state is proportional to the number of that state’s electoral votes and inversely proportional to the number of votes cast in that state. Since these two factors are only weakly correlated, voting power varies from state to state, as it does under the Maine-Nebraska Plan and other district-based systems.
4 Challenges of the County-Elector Plan
4.1 Campaigns Will Have to Allocate Resources across Hundreds of Counties
At its core, a campaign is “the process of acquiring and using political resources that can secure votes” (Leuthold 1968, p. 1). Starting in the 1970s, computers were used to execute relatively simplistic mathematical algorithms to determine where and how campaign resources should be allocated (Bartels 1985; Brams and Davis 1974; Colantoni, Levesque, and Ordeshook 1975). Now, in the first decades of the 21st century, presidential campaigns rely on “big data” to drive massive number-crunching operations that permit campaign strategists, pollsters, statisticians, and computer scientists to develop resource allocation algorithms for campaign budgeting, scheduling, fund raising, advertising, and voter mobilization (Davis et al. 2009; Hill and McKee 2005; Shaw 1999). Although mindful that every vote counts, campaign managers concentrate these activities on battleground states, where every vote counts more. The County-Elector Plan, by making nearly all states competitive at some level, would introduce appreciably greater marketing complexity across every aspect of presidential campaigns, particularly in resource allocation. Although this additional complexity would certainly drive campaign costs upward, a candidate who was able to technologically harness the electoral potential of the County-Elector Plan would have a tremendous advantage over those who were less able to do so.
4.2 Would Some Voters Be Disinclined to Vote?
At first blush, it appears that in counties where one presidential candidate is particularly strong, voters who favor a weaker candidate might be disinclined to vote at all, knowing that their vote would increase voter turnout and permit their county to obtain a higher share of the state’s electoral votes, all of which would go to the winning candidate (i.e. the one they opposed). This dilemma is eliminated if affected voters decline to vote for any presidential candidate but vote as normal for all other portions of the ballot.
4.3 Are the Ballots Too Complex?
One argument against districts that are not identical to congressional districts is that voters would be faced with a more complex, multi-district ballot. “There seems to be general agreement that although two sets of electoral districts are possible, they would not be desirable. Voters at the same polling place could be voting for the same congressmen but for different district presidential electors[.]” (Kefauver 1962, p. 199). Although such complexity may have been an obstacle in the 1960s, modern-day voters typically deal with significantly more complex situations. For example, in addition to voting for President and US Senator in 2016, voters in Los Angeles County voted for candidates in their US congressional district, state senate district, state assembly district, county (supervisors, judges), city (council members), health care district, recreation and park district, elementary and high school district, community college district, and water district. County voters also had to deal with 17 state propositions, two county propositions, and some of the 27 city and 29 school district propositions that were on the ballot. Although Los Angeles County’s ballot may be more intricate than the average ballot, typical election-year ballots have grown in length over the past decades, and any additional complexity the County-Elector Plan could possibly contribute to the voting process would be de minimis.
4.4 Does the County-Elector Plan Favor Democrats?
An argument could be made that the County-Elector Plan benefits Democrats because large numbers of electoral votes are clustered in urban counties that generally lean toward that party, while Republican candidates currently perform best in rural counties that have small populations (Scala and Johnson 2017, p. 181). Such a dichotomy requires that Republicans campaign in far more areas than Democrats are required to. For example, Los Angeles County, the country’s most populous county, has a population equal to the combined population of the country’s 1,200 least populated counties. That does not mean, of course, that Republican candidates would have to campaign in tiny counties – such as Loving County, Texas (pop. 51), or Kalawao County, Hawaii (pop. 82) – but it would require that Republicans judiciously allocate campaign resources, quite possibly by increasing reliance on county-level party officials to generate voter enthusiasm. This enthusiasm could be driven by the knowledge that increased voter turnout (relative to other counties) directly translates to increased electoral votes allocated to a given county. Assisting Republicans in this approach is the principle from statistical geometry that dictates that low-population counties are necessarily clustered together, meaning that large swaths of such counties can be treated as a single “super-county” for campaign purposes. For example, the 26 counties comprising the Texas Panhandle are, for all intents and purposes, a single voting bloc that voted in favor of Trump by a 4.7–1 ratio in 2016. Further, this Republican-rich area of Texas lies in the middle of a 350-mile-wide corridor of counties stretching southward into central Texas and northward through Oklahoma, Kansas, Nebraska, South Dakota, and North Dakota – more than 500 counties in total, of which only 16 voted for Clinton (see Figure 1[18]). Although it would be imprudent for a campaign to treat these counties as a single bloc, it would be reasonable for a campaign to group contiguous counties into super-counties, each with electoral votes equivalent to that of an urban county with a population of several hundred thousand. Political marketing activities could then be designed and implemented to maximize voter turnout within each super-county. This approach would bring a dynamic perspective to the entire electioneering process.
County-level winners in Texas, Oklahoma, Kansas, Nebraska, South Dakota, and North Dakota in the 2016 election.
4.5 Would the County-Elector Plan Empower Minority Voters?
Lani Guinier, former Harvard law professor and long-time proponent of proportional voting, asserted that minorities are frequently disenfranchised by winner-take-all voting because those systems “give[] those who get the most votes all the power. Thus, votes cast for ‘losers’ are wasted. This submerges the voting strength of all voters who do not support the winning candidate” (Guinier 1993, p. 1137).
The County-Elector Plan would provide concentrated minority groups – such as those in large urban areas – an opportunity to make a tangible impact on statewide voting results. The plan’s potential is illustrated by Wayne County, which held 18 % of Michigan’s voting population in 2016 and would have represented 2.88 of the state’s then-16 electoral votes under the County-Elector Plan. Given that Blacks constituted about 38 % of the county’s residents in 2016, their impact would have been very significant in the awarding of those 2.88 electoral votes to the county’s plurality winner. This approach is in stark contrast to the statewide winner-take-all system, in which Blacks have a markedly lesser role, given that they constituted only 15 % of Michigan’s population in 2016. Recognizing their potential to affect the outcome of a presidential election might also increase the turnout of urban minority voters, given that “the prospects of changing the outcome of an election and the utility that voters get from the act of voting” are the two primary reasons citizens vote[19] (Oberholzer-Gee and Waldfogel 2001, p. 6).
4.6 Is the County-Elector Plan Appropriate for Every State?
It clearly lies within each state’s Constitutionally defined domain to implement the type of elector-selection system that best suits its citizens. Maine and Nebraska currently use the district-popular method, and other states have previously implemented other methods. The potential success of the County-Elector Plan does not depend upon its implementation in all 50 states and Washington, D.C. Some states may find the County-Elector Plan appropriate, while others may not. For example, states such as Massachusetts, Oklahoma, and Vermont that have a high degree of political uniformity across all counties would likely see no reason to alter the status quo, since the probability of the state splitting electoral votes is small. However, states in which citizens are nearly balanced between both major parties might find the plan more appealing, given that it better represents the intentions of the voters. Similarly, states that lean toward one party but have a sizable contingent of citizens favoring the opposing party (e.g., California, Nevada, and Utah) may also believe the County-Elector Plan is a viable alternative to the unit system.
The initial consideration regarding alternative electoral plans is not which new plan to consider but whether to consider any deviation from the unit system. The argument at the core of this discussion harkens back to 1800, when Thomas Jefferson wrote a letter to James Madison expressing Jefferson’s opinion that it was “worse than folly” for Virginia to have a district system if other states had unit systems (Jefferson 1905, pp. 90–1).
All agree that an election by districts would be best, if it could be general; but while 10 states chuse either by their legislatures or by a general ticket, it is folly & worse than folly for the other 6 [] not to do it. In these 10 [] states the minority is entirely unrepresented; & their majorities not only have the weight of their whole state in their scale, but have the benefit of so much of our minorities as can succeed at a district election.
Two points are critical to understanding this excerpt. First, Jefferson’s statement must be properly placed within the context of his political bent. That is, as a co-founder of the Democratic-Republican party and a staunch anti-Federalist fighting to guide the country toward his vision of the proper balance of state and federal powers, Jefferson was clearly not interested in diminishing the sway that a specific state might hold in a presidential election. From that perspective, there was no advantage to a system that might split a state’s electoral votes. Second, Jefferson’s letter was written on January 12, 1800, just months before the start of voting for the election that would produce his presidency. Although Jefferson was a native Virginian and could count on winning the majority of that state’s popular vote (and hence all the state’s electors under a unit system), his support was not uniform across the state. It was thus uncertain if he would be able to capture each of the 21 special electoral districts Virginia had used during the preceding presidential election and that the state was scheduled to use in 1800. The concern was warranted, since in the mid-term elections held a year prior to Jefferson’s letter, Federalists had won 8 of Virginia’s 19 congressional districts (Ferling 2004, p. 4). Thus, in anticipation of his candidacy, it was in Jefferson’s best interest to express his strong disdain for district systems.[20]
The fact that 48 states and Washington, D.C., have retained the unit system for so long does not necessarily signal their satisfaction with it; instead, it likely represents some combination of political inertia and the void of appealing alternatives. The County-Elector Plan provides a model the states may deem preferable to the unit system and the Maine-Nebraska Plan.
5 Retrospective Analysis of the 2016 Election under the County-Elector Plan
This section expands upon the simple Delaware example by retrospectively applying the County-Elector Plan to nationwide, county-level voting results of the 2016 election and comparing the results to the actual nationwide outcome. The primary point of the analysis is to illustrate that the County-Elector Plan is not a revolutionary departure from the status quo, but that it could extend the current paradigm’s focus on battleground states to a system in which numerous other states are relevant to the outcome of a presidential election. Of course, had the County-Elector Plan been in place nationwide for the 2016 presidential election, all parties would have modified their campaign strategies in ways that cannot now be stated with certainty. At the end of the following analysis, we hypothesize what some of those changes might have been.
The 2016 presidential election saw Trump defeat Clinton, 304 electoral votes to 227, despite the fact that Clinton received 2.9 million more popular votes than Trump.[21] That election’s nationwide, county-by-county vote tallies were retrospectively analyzed using the County-Elector Plan.[22] Table 3 presents the state-level results of that analysis, with Clinton defeating Trump 282 electoral votes to 256. Key to this result is that the model introduced a deep competitiveness into the race, with 41 states splitting their electoral votes between the candidates. In comparison, only Maine split its electoral vote in the actual 2016 election.
Nationwide comparison of electoral votes under the County-Elector Plan and the current system for the 2016 election.
| State | Electoral votes | State | Electoral votes | ||||||
|---|---|---|---|---|---|---|---|---|---|
| County-Elector Plan | Current system | County-Elector Plan | Current system | ||||||
| Clinton | Trump | Clinton | Trump | Clinton | Trump | Clinton | Trump | ||
| Alabama | 2 [2.12] | 7 [6.88] | 0 | 9 | Montana | 1 [0.86] | 2 [2.14] | 0 | 3 |
| Alaska | 1 [0.86] | 2 [2.14] | 0 | 3 | Nebraska | 2 [2.23] | 3 [2.77] | 0 | 5 |
| Arizona | 2 [2.22] | 9 [8.78] | 0 | 11 | Nevada | 5 [5.21] | 1 [0.79] | 6 | 0 |
| Arkansas | 1 [1.21] | 5 [4.79] | 0 | 6 | New Hampshire | 1 [1.32] | 3 [2.68] | 4 | 0 |
| California | 51 [50.60] | 4 [4.40] | 55 | 0 | New Jersey | 9 [9.47] | 5 [4.53] | 14 | 0 |
| Colorado | 6 [5.56] | 3 [3.44] | 9 | 0 | New Mexico | 4 [3.66] | 1 [1.34] | 5 | 0 |
| Connecticut | 6 [6.37] | 1 [0.63] | 7 | 0 | New York | 20 [19.80] | 9 [9.20] | 29 | 0 |
| Delaware | 2 [1.78] | 1 [1.22] | 3 | 0 | North Carolina | 7 [7.29] | 8 [7.71] | 0 | 15 |
| DC | 3 [3.00] | 0 [0.00] | 3 | 0 | North Dakota | 0 [0.04] | 3 [2.96] | 0 | 3 |
| Florida | 12 [12.49] | 17 [16.51] | 0 | 29 | Ohio | 8 [7.64] | 10 [10.36] | 0 | 18 |
| Georgia | 9 [8.69] | 7 [7.31] | 0 | 16 | Oklahoma | 0 [0.00] | 7 [7.00] | 0 | 7 |
| Hawaii | 4 [4.00] | 0 [0.00] | 4 | 0 | Oregon | 4 [4.12] | 3 [2.88] | 7 | 0 |
| Idaho | 0 [0.17] | 4 [3.83] | 0 | 4 | Pennsylvania | 11 [10.60] | 9 [9.40] | 0 | 20 |
| Illinois | 14 [14.17] | 6 [5.83] | 20 | 0 | Rhode Island | 3 [3.29] | 1 [0.71] | 4 | 0 |
| Indiana | 3 [2.95] | 8 [8.05] | 0 | 11 | South Carolina | 2 [2.33] | 7 [6.67] | 0 | 9 |
| Iowa | 2 [2.40] | 4 [3.60] | 0 | 6 | South Dakota | 0 [0.10] | 3 [2.90] | 0 | 3 |
| Kansas | 0 [0.50] | 6 [5.50] | 0 | 6 | Tennessee | 3 [2.60] | 8 [8.40] | 0 | 11 |
| Kentucky | 2 [2.03] | 6 [5.97] | 0 | 8 | Texas | 17 [17.07] | 21 [20.93] | 0 | 38 |
| Louisiana | 2 [2.10] | 6 [5.90] | 0 | 8 | Utah | 2 [2.33] | 4 [3.67] | 0 | 6 |
| Maine | 2 [2.20] | 2 [1.80] | 3 | 1 | Vermont | 3 [2.97] | 0 [0.03] | 3 | 0 |
| Maryland | 7 [7.23] | 3 [2.77] | 10 | 0 | Virginia | 7 [6.75] | 6 [6.25] | 13 | 0 |
| Massachusetts | 11 [11.00] | 0 [0.00] | 11 | 0 | Washington | 9 [8.97] | 3 [3.03] | 12 | 0 |
| Michigan | 7 [7.33] | 9 [8.67] | 0 | 16 | West Virginia | 0 [0.00] | 5 [5.00] | 0 | 5 |
| Minnesota | 5 [5.26] | 5 [4.74] | 10 | 0 | Wisconsin | 4 [3.55] | 6 [6.45] | 0 | 10 |
| Mississippi | 2 [1.57] | 4 [4.43] | 0 | 6 | Wyoming | 0 [0.15] | 3 [2.85] | 0 | 3 |
| Missouri | 4 [3.68] | 6 [6.32] | 0 | 10 | |||||
| Total | 282 | 256 | 232 | 306 | |||||
-
The County-Elector Plan’s state-level results are shown in brackets to two decimal points. The bold values to the left of the brackets are the rounded values that states would certify and submit to the U.S. Congress. Also, the votes of faithless electors are credited back to the candidates.
Of the country’s 3,155 vote-reporting counties or county equivalents for the election of 2016, 2,627 voted for Trump, 506 voted for Clinton, and 22 voted for a third-party candidate. Trump generally did well in both small and mid-sized rural counties, while Clinton performed well in large (mostly urban) counties. It was previously discussed that the County-Elector Plan might favor Democratic presidential candidates because most large counties lean toward that party, as they did for Clinton. That argument is supported by the data in Table 4 showing how Clinton and Trump would have fared under the County-Elector Plan in the country’s 10 most populous counties in 2016. Then, for illustrative purposes, Table 5 shows how those candidates would have fared in the country’s 20 least populous counties.
Electoral votes awarded under the County-Elector Plan in the 10 most populous US counties.
| County | Electoral votes | |
|---|---|---|
| Clinton | Trump | |
| Los Angeles County, CA | 13.27 | – |
| Cook County, IL | 7.79 | – |
| Harris County, TX | 5.54 | – |
| Maricopa County, AZ | – | 6.62 |
| San Diego County, CA | 5.05 | – |
| Orange County, CA | 4.63 | – |
| Miami–Dade County, FL | 3.01 | – |
| Kings County, NY | 3.03 | – |
| Dallas County, TX | 3.24 | – |
| Riverside County, CA | 2.90 | – |
Electoral votes awarded under the County-Elector Plan in the 20 least populous US counties.
| County | Electoral votes | |
|---|---|---|
| Clinton | Trump | |
| Kalawao, HI | – | 0.0003 |
| Loving, TX | – | 0.0003 |
| King, TX | – | 0.0007 |
| Kenedy, TX | 0.0008 | – |
| Arthur, NE | – | 0.0016 |
| Blaine, NE | – | 0.0019 |
| Petroleum, MT | – | 0.0019 |
| McPherson, NE | – | 0.0017 |
| Loup, NE | – | 0.0023 |
| Grant, NE | – | 0.0024 |
| Borden, TX | – | 0.0015 |
| Yakutat, AK | – | 0.0032 |
| San Juan, CO | – | 0.0016 |
| Harding, NM | – | 0.0033 |
| Thomas, NE | – | 0.0023 |
| Treasure, MT | – | 0.0026 |
| Mineral, CO | – | 0.0021 |
| Hooker, NE | – | 0.0025 |
| Banner, NE | – | 0.0024 |
| Wheeler, NE | – | 0.0028 |
Table 6 then shows that the County-Elector Plan more closely approximates the 2016 national popular vote than does the current system. Indeed, the electoral votes delivered under the plan deviated from the nationwide popular vote by less that 2 % and would have delivered the presidency to Clinton, thus preventing the fifth inverted election in US history. The plan is a hybrid method for allocating a state’s electoral votes that produces state- and national-level results that closely match the popular voting results.
Nationwide comparison of the County-Elector Plan and the current system for the 2016 election.
| Candidate | % of Popular votes | % of Electoral votes | |
|---|---|---|---|
| County-Elector Plan | Actual results | ||
| Trump | 48.9 | 47.6 | 57.3 |
| Clinton | 51.1 | 52.4 | 42.7 |
-
Calculations exclude third-party candidates. The votes of faithless electors are not credited back to the candidates.
In the months preceding the 2016 election, conventional wisdom held that there were 11 battleground states, and that the other states and Washington, D.C., were not in serious contention (Mahtesian 2016). Had the County-Elector Plan been in effect nationwide for that election, the situation would have been vastly different: there would have been at least 40 competitive states (see Table 3). Granted, not all regions in these states would have been contested, but at least portions of the states would have been worthy of a campaign’s attention. The type of attention is impossible to retrospectively ascertain, but visits by presidential and vice-presidential candidates, national surrogates, and local political leaders have historically been a highly effective campaign tool and likely would have increased in importance in a campaign with so many competitive regions. For example, following their nomination on July 22, 2016, Trump and Pence appeared at 248 campaign events prior to the election, of which 219 (88 %) were in battleground states. Correspondingly, in the period following their nomination on July 29, 2016, Clinton and Kaine held 151 campaign events, of which 146 (97 %) were in battleground states (National Popular Vote 2025). These numbers correspond to the Trump/Pence and Clinton/Kaine tickets holding an average of 2.0 and 1.4 events in battleground states per day, respectively, which are daunting schedules by any measure. Given that the County-Elector Plan would have involved competitive counties in 41 states – not simply 11 battleground states – it is clear that the candidates would have been physically incapable of personally addressing the competitive states at the level they did in 2016. Consequently, it is reasonable to hypothesize that candidate surrogates – such as governors, mayors, senators, representatives, and other party faithful – would have been tasked with handling in-person appearances at the majority of county-level events. Since they are able to link presidential candidates to local politics, such actors generally play important roles in campaigns, but their roles likely would have been more important under the County-Elector Plan. For such a strategy to be successful, the linkage between national party staff and county operatives would had to have been strengthened and maintained throughout the election cycle. Once established, though, the linkage could be maintained during inter-election periods and would provide a strong basis for subsequent elections.
The County-Elector Plan would have produced a broadly competitive 2016 election, whose outcome would have depended greatly on each party’s ability to develop and efficiently utilize hundreds of local political organizations to an extent not previously seen in presidential elections. An important derivative question then is, “In which counties should the parties focus their efforts in a future election?” Looking beyond the above-described “large county, small county” dichotomy, one strategy would be to focus on counties where one party’s margin of victory is sufficiently small so as to warrant the other party’s dedication of resources in an attempt to flip it – i.e., the battleground counties. Figure 2 presents the distribution of county-level winning margins in the 2016 election. Each party’s initial future interest would likely be drawn to counties with the smallest margins, such as the 166 counties in which the winning margin was less than 5 %. Figure 3 shows that these 166 counties collectively contain approximately 53 electoral votes, basically 10 % of the votes at stake in the entire election. The map in Figure 4 then shows that those 166 counties are not only distributed across 42 states, but that they are also geographically dispersed within those states. This profound decentralization brings renewed focus on the importance of utilizing candidate surrogates to an extent not previously seen, and the ultimate success of the presidential campaign relies heavily upon the proper management of surrogate activities. One party’s effort, however, will not be conducted in a vacuum, since the opposing party will certainly launch counter offensives in the battleground counties in which it is vulnerable to attack, and the resulting multi-front campaigns in large numbers of battleground counties will consume financial, personnel, and technological resources on an unprecedented scale.
Winning margins in all 3,155 counties and county-equivalents in the 2016 election.
Electoral votes in 166 battleground counties and county-equivalents in the 2016 election.
Battleground counties and county-equivalents in the 2016 election.
6 Conclusions
The Electoral College has successfully guided the US through 60 presidential elections, despite having been subjected to a near-constant litany of complaints about its fairness and complexity. Still, the country’s dissatisfaction with the Electoral College has never reached the level required to replace or bypass it, although many attempts to do so have been initiated. What has changed over the past two centuries, however, are the methods by which individual states select electors. More than a dozen different methods to do so have been tried, but 48 states and the District of Columbia continue to award their electoral votes to the plurality winner of the statewide popular vote, thereby contributing to the hyper-focus presidential campaigns place on battleground states and the consequential marginalization of voters living in spectator states.
The County-Elector Plan retains the Electoral College’s federal-level structure but provides a novel method for states to select electors. By focusing on counties, the plan leverages the fact that local representation is a long-standing principle of the American political system. The plan has six attractive features: (1) it leverages the county-level administration that most states use to manage presidential elections; (2) it permits campaigns to shift their marketing focus from a small set of battleground states to a multitude of battleground counties; (3) it is gerrymandering-resistant; (4) it provides all voters in a state with equal voting power; (5) it virtually eliminates the possibility of an inverted election, and (6) it has the potential to increase voter turnout.
Whether the County-Elector Plan will ever be implemented is currently unknowable. Indeed, as Katz (2005, p. 74) asks, “[W]hy do reforms happen, especially given the assumption that parties will not want to change the rules of the game they are winning?” Maine broke from the winner-take-all pack and shifted to a novel but unproven mechanism more than 50 years ago, but Nebraska was both slow and alone to follow. Still, all it would take to adopt the County-Elector Plan is for politicians in but one state to follow Justice Brandeis’s vision that “a single courageous State may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country” (New State Ice Co. v. Liebmann 1932, p. 311). Absent such political altruism, it might be necessary for that state’s politicians to return to the bare-knuckled political brawls that changed the Electoral College’s course in the Republic’s earliest days.
References
Ames, H. V. 1897. “The Proposed Amendments to the Constitution of the United States during the First Century of Its History.” In Annual Report of the American Historical Society for the Year 1896, Vol. 2. Washington: Government Printing Office.Search in Google Scholar
BanzhafIIIJ. F. 1964. “Weighted Voting Doesn’t Work: A Mathematical Analysis.” Rutgers Law Review 19: 317–43.Search in Google Scholar
BanzhafIIIJ. F. 1966. “Multi-Member Electoral Districts – Do They Violate the ‘One Man, One Vote’ Principle.” The Yale Law Journal 75: 1309–88. https://doi.org/10.2307/795047.Search in Google Scholar
BanzhafIIIJ. F. 1968. “One Man, 3.312 Votes: A Mathematical Analysis of the Electoral College.” Villanova Law Review 13 (2): 304–32.Search in Google Scholar
Bartels, L. M. 1985. “Resource Allocation in a Presidential Campaign.” The Journal of Politics 47 (3): 928–36. https://doi.org/10.2307/2131218.Search in Google Scholar
Bennett, R. W. 2001. “Popular Election of the President without a Constitutional Amendment.” Green Bag 4: 241–6.Search in Google Scholar
Bourgault, J., and D. Schultz. 2024. “Institutionalized Disenfranchisement for Voters: The Need for Electoral College Reform or Abolishment.” Cooley Law Review 39 (2): 1–29.Search in Google Scholar
Brams, S. J., and M. D. Davis. 1974. “The 3/2’s Rule in Presidential Campaigning.” American Political Science Review 68: 113–34. https://doi.org/10.2307/1959746.Search in Google Scholar
Bugh, G. 2016. “Representation in Congressional Efforts to Amend the Presidential Election System.” In Electoral College Reform: Challenges and Possibilities, edited by G. Bugh, 5–18. New York: Routledge.10.4324/9781315579023Search in Google Scholar
Canady, M. 2014–15. “Redistribution of Electoral Votes: Following the Constitution, Nebraska, and Maine to More Equal Representation in Presidential Elections.” Adelphia Law Journal 19: 63–94.10.2139/ssrn.2324612Search in Google Scholar
Clayton, D. M. 2007. “The Electoral College: An Idea Whose Time Has Come and Gone.” The Black Scholar 37 (3): 28–41. https://doi.org/10.1080/00064246.2007.11413406.Search in Google Scholar
Colantoni, C. S., T. J. Levesque, and P. C. Ordeshook. 1975. “Campaign Resource Allocations under the Electoral College.” American Political Science Review 69: 141–61. https://doi.org/10.2307/1957891.Search in Google Scholar
Connecticut Secretary of State. 2025. “Regional Councils of Governments.” https://portal.ct.gov/sots/register-manual/section-vii/regional-councils-of-governments (accessed February 22, 2025).Search in Google Scholar
Davis, R., J. C. Baumgartner, P. L. Francia, and J. S. Morris. 2009. “The Internet in U.S. Election Campaigns.” In Routledge Handbook of Internet Politics, edited by A. Chadwick, and P. N. Howard, 13–24. New York: Routledge.Search in Google Scholar
Dehez, P. 2024. “Voting and Power.” In Game Theory for the Social Sciences: Conflict, Bargaining, Cooperation and Power. Heidelberg: Springer Verlag.10.1007/978-3-031-58241-7Search in Google Scholar
Dubey, P., and L. S. Shapley. 1979. “Mathematical Properties of the Banzhaf Power Index.” Mathematics of Operations Research 4 (2): 99–131. https://doi.org/10.1287/moor.4.2.99.Search in Google Scholar
Ferling, J. 2004. “Thomas Jefferson, Aaron Burr and the Election of 1800.” Smithsonian Magazine.10.1093/oso/9780195167719.003.0001Search in Google Scholar
Fon, V. 2008. “Integral Proportional System: Aligning Electoral Votes More Closely with State Popular Votes.” Supreme Court Economic Review 16 (1): 127–57. https://doi.org/10.1086/655882.Search in Google Scholar
Gardner, J. A. 2000. “Madison’s Hope: Virtue, Self-Interest, and the Design of Electoral Systems.” Iowa Law Review 86: 87–172.Search in Google Scholar
Green, A. 2006. Tasmania’s Hare-Clark Electoral System. ABC News.Search in Google Scholar
Griffith, E. C. 1907. The Rise and Development of the Gerrymander. Chicago: Scott, Foresman.Search in Google Scholar
Gringer, D. 2008. “Why the National Popular Vote Plan Is the Wrong Way to Abolish the Electoral College.” Columbia Law Review 108 (1): 182–230.Search in Google Scholar
Guinier, L. 1993. “The Representation of Minority Interests: The Question of Single-Member Districts.” Cardozo Law Review 14 (5): 1135–74.Search in Google Scholar
Heilman, S. 2021. “Designing Stable Elections.” Notices of the American Mathematical Society 68 (4): 516–27. https://doi.org/10.1090/noti2251.Search in Google Scholar
Herbst, C. J. 2012. “Redrawing the Electoral Map: Reforming the Electoral College with the District-Popular Plan.” Hofstra Law Review 41 (1): 217–65.Search in Google Scholar
Hill, D., and S. C. McKee. 2005. “The Electoral College, Mobilization, and Turnout in the 2000 Presidential Election.” American Politics Research 33 (5): 700–25. https://doi.org/10.1177/1532673x04271902.Search in Google Scholar
Hoffman, M. M. 1996. “The Illegitimate President: Minority Vote Dilution and the Electoral College.” The Yale Law Journal 105 (4): 935–1021. https://doi.org/10.2307/797244.Search in Google Scholar
Hunter, T. R. 2011. “The First Gerrymander? Patrick Henry, James Madison, James Monroe, and Virginia’s 1788 Congressional Districting.” Early American Studies 9 (3): 781–820. https://doi.org/10.1353/eam.2011.0023.Search in Google Scholar
Jefferson, T. 1905. The Works of Thomas Jefferson in Twelve Volumes, Vol. 9, edited by P. L. Ford. New York: G.P. Putnam’s Sons.Search in Google Scholar
Kaniovski, S. 2008. “The Exact Bias of the Banzhaf Measure of Power when Votes Are Neither Equiprobable Nor Independent.” Social Choice and Welfare 31 (2): 281–300. https://doi.org/10.1007/s00355-007-0280-x.Search in Google Scholar
Karr, J. 2015. “Proportional Union or Paper Confederacy Note.” Connecticut Law Review 48 (2): 595–636.Search in Google Scholar
Katz, R. S. 1997. Democracy and Elections. New York: Oxford University Press.10.1093/acprof:oso/9780195044294.001.0001Search in Google Scholar
Katz, R. S. 2005. “Why Are There So Many (Or So Few) Electoral Reforms?” In The Politics of Electoral Systems, edited by M. Gallagher, and P. Mitchell, 57–76. Oxford: Oxford University Press.10.1093/0199257566.003.0003Search in Google Scholar
Kefauver, E. 1962. “The Electoral College: Old Reforms Take on a New Look.” Law and Contemporary Problems 27 (2): 188–212. https://doi.org/10.2307/1190543.Search in Google Scholar
Koza, J. R., B. Fadem, M. Grueskin, M. S. Mandell, R. Richie, and J. S. Zimmerman. 2013. Every Vote Equal: A State-Based Plan for Electing the President by National Popular Vote, 4th ed. Los Altos: National Popular Vote Press.Search in Google Scholar
Lawsuit Aims to Speed Maine Redistricting. 2011. Bangor Daily News.Search in Google Scholar
Leip, D. 2022. Dave Leip’s Atlas of U.S. Presidential Elections. https://uselectionatlas.org/.Search in Google Scholar
LeRoy, S. P. 2024. “Mutually Assured Democracy.” University of Chicago Law Review 91 (5): 1451–500.Search in Google Scholar
Leuthold, D. A. 1968. Electioneering in a Democracy: Campaigns for Congress. New York: Wiley.Search in Google Scholar
Lucas, W. F. 1983. “Measuring Power in Weighted Voting Systems.” In Political and Related Models. Modules in Applied Mathematics, edited by S. J. Brams, W. F. Lucas, and P. D. Straffin. New York: Springer.10.1007/978-1-4612-5430-0_9Search in Google Scholar
Mahtesian, C. 2016. “What Are the Swing States in 2016?” https://www.politico.com/blogs/swing-states-2016-election/2016/06/what-are-the-swing-states-in-2016-list-224327 (accessed February 22, 2025).Search in Google Scholar
Mann, I., and L. S. Shapley. 1960. Values of Large Games, IV: Evaluating the Electoral College by Montecarlo Techniques. Memo No. RM-2651. Santa Monica: The RAND Corporation.Search in Google Scholar
Mann, I., and L. S. Shapley. 1962. Values of Large Games, VI: Evaluating the Electoral College Exactly. Memo No. RM-3158-PR. Santa Monica: The RAND Corporation.Search in Google Scholar
Margolis, H. 1983. “The Banzhaf Fallacy.” American Journal of Political Science 27: 321–6. https://doi.org/10.2307/2111020.Search in Google Scholar
Martin, W. L. 1958. “Presidential Electors: Let the State Legislatures Choose Them.” American Bar Association Journal 44: 1182–7.Search in Google Scholar
McPherson, v.Blacker . 1892. 146 U.S. 1.Search in Google Scholar
Melcher, J. P. 2004. “Electing to Reform: Maine and the District Plan for Selection of Presidential Electors.” In Proceedings from Annual Meeting of the New England Political Science Association. North Easton, MA: The New England Political Science Association.Search in Google Scholar
Miller, N. R. 2014. “A Priori Voting Power when One Vote Counts in Two Ways, with Application to Two Variants of the US Electoral College.” In Voting Power and Procedures: Essays in Honour of Dan Felsenthal and Moshé Machover, edited by R. Fara, D. Leech, and M. Salles. Germany: Springer International Publishing.10.1007/978-3-319-05158-1_11Search in Google Scholar
Mississippi Secretary of State. 1945. “Vote for Presidential Electors, General Election, November 7, 1944.” In Mississippi Blue Book, 228–31. Jackson: Secretary of State, State of Mississippi.Search in Google Scholar
National Popular Vote. 2025. 2016, 94% of the Presidential Campaign Was in Just 12 Closely Divided States. https://www.nationalpopularvote.com/campaign-events-2016 (accessed February 22, 2025).Search in Google Scholar
New State Ice Co. V. Liebmann. 1932. 285 U.S 262.Search in Google Scholar
Oberholzer-Gee, F., and J. Waldfogel. 2001. “Electoral Acceleration: The Effect of Minority Population on Minority Voter Turnout.” In NBER Working Paper 8252. Cambridge: National Bureau of Economic Research. (NBER).10.3386/w8252Search in Google Scholar
Pal, D. 2021. “Does Everyone Have Equal Voting Power?” Indian Economic Review 56 (2): 515–25. https://doi.org/10.1007/s41775-021-00117-6.Search in Google Scholar
Read, Jr., D. 1976. “Direct Election of the President without a Constitutional Amendment: A Call for State Action.” Washington Law Review 51 (2): 321–49.Search in Google Scholar
Scala, D. J., and K. M. Johnson. 2017. “Political Polarization along the Rural-Urban Continuum? The Geography of the Presidential Vote, 2000–2016.” Annals of the American Academy of Political and Social Science 672 (1): 162–84. https://doi.org/10.1177/0002716217712696.Search in Google Scholar
Shane, P. M. 2001. “Disappearing Democracy: How Bush V. Gore Undermined the Federal Right to Vote for Presidential Electors.” Florida State University Law Review 29 (2): 535–85.10.2139/ssrn.278283Search in Google Scholar
Shaw, D. R. 1999. “The Methods behind the Madness: Presidential Electoral College Strategies, 1988–1996.” Journal of Politics 61 (4): 893–913. https://doi.org/10.2307/2647547.Search in Google Scholar
Shelley, F. M. 2002. “The Electoral College and the Election of 2000.” Political Geography 21: 79–83. https://doi.org/10.1016/s0962-6298(01)00062-2.Search in Google Scholar
Silva, R. C. 1950. “The Lodge-Gossett Resolution: A Critical Analysis.” American Political Science Review 44: 86–99. https://doi.org/10.2307/1950349.Search in Google Scholar
Spilerman, S., and D. Dickens. 1974. “Who Will Gain and Who Will Lose Influence under Different Electoral Rules.” American Journal of Sociology 80 (2): 443–77. https://doi.org/10.1086/225808.Search in Google Scholar
Streb, M. J. 2008. Rethinking American Electoral Democracy. New York: Routledge.10.4324/9780203931080Search in Google Scholar
Strömberg, D. 2008. “How the Electoral College Influences Campaigns and Policy: The Probability of Being Florida.” American Economic Review 98 (3): 769–807. https://doi.org/10.1257/aer.98.3.769.Search in Google Scholar
Tollar, E. T., and S. H. Kimball. 2020. “A More Perfect Electoral College: Challenging Winner-takes-all Provisions under the Twelfth Amendment.” Legislation and Policy Brief 9 (1): 4–36.Search in Google Scholar
Turner, R. C. 2005. “The Contemporary Presidency: Do Nebraska and Maine Have the Right Idea? the Political and Partisan Implications of the District System.” Presidential Studies Quarterly 35 (1): 116–37. https://doi.org/10.1111/j.1741-5705.2004.00238.x.Search in Google Scholar
Vardi, S. 2018. “Stepwise Presidential Election Reform: The Biased Proportional Plan and its Implementation.” https://doi.org/10.2139/ssrn.3184012 (accessed December 22, 2024).Search in Google Scholar
Wagner, D. S. 2006. “The Forgotten Avenue of Reform: The Role of States in Electoral College Reform and the Use of Ballot Initiatives to Effect that Change.” Review of Litigation 25 (3): 575–602.Search in Google Scholar
Wang, S. S. H. 2016. “Three Tests for Practical Evaluation of Partisan Gerrymandering.” Stanford Law Review 68: 1263–321.Search in Google Scholar
Wenschhof, L. 2014. The Geography of Voter Power in the US Electoral College from 1900–2012. Master’s thesis, University of Southern California.Search in Google Scholar
Wise, G. 2022. “Politically Viable US Electoral College Reform.” Statistics, Politics and Policy 13: 119–43. https://doi.org/10.1515/spp-2021-0029.Search in Google Scholar
© 2025 the author(s), published by De Gruyter, Berlin/Boston
This work is licensed under the Creative Commons Attribution 4.0 International License.
Articles in the same Issue
- Frontmatter
- Editorial
- Editors’ Note
- Articles
- A State-Level Policy Change That Would Revitalize the Electoral College
- Pandemic Intensity Estimation using Dynamic Factor Modeling
- Spatio-Temporal Modeling of Violent Conflict and Fatality in Nigeria: A Point Process Modeling with SPDE Approach
- A Comprehensive Meta-Analysis of Determinants Influencing Military Expenditure: New Methodological Insights and Implications for Defence Budget Allocation
Articles in the same Issue
- Frontmatter
- Editorial
- Editors’ Note
- Articles
- A State-Level Policy Change That Would Revitalize the Electoral College
- Pandemic Intensity Estimation using Dynamic Factor Modeling
- Spatio-Temporal Modeling of Violent Conflict and Fatality in Nigeria: A Point Process Modeling with SPDE Approach
- A Comprehensive Meta-Analysis of Determinants Influencing Military Expenditure: New Methodological Insights and Implications for Defence Budget Allocation
