9.31 Myths that a Nationwide Vote for President would Favor One Political Party Over the Other

If Democrats had an inherent advantage in winning the national popular vote for President, we would see some evidence of this tendency in the historical record.

The United States is, politically, an evenly divided country in which the cumulative nationwide vote for the two parties from the start of the modern political era in 1932 through 2008 (table 9.25) has been virtually tied:

• 746,260,766 total votes for the Democrats and
• 745,502,654 total votes for the Republicans.

Table 9.25 shows the national popular vote for President between 1932 and 2008. Columns 4 and 5 show the Democratic and Republican margin, respectively, in each election.

Table 9.25 The national popular vote for President 1932–2008

 

The Republican Party has fared well in terms of the national popular vote. Since the formation of the Republican Party, nine Republicans have won more than 53% of the national popular vote, namely Ulysses Grant, Theodore Roosevelt, Warren Harding, Calvin Coolidge, Herbert Hoover, Dwight Eisenhower, Richard Nixon, Ronald Reagan, and George H.W. Bush, whereas only two Democrats have done so (Franklin Roosevelt and Lyndon Johnson).

Based on past performance, there is nothing to indicate the Republican Party is either advantaged or disadvantaged if presidential elections are decided on the basis of the national popular vote.

The candidate who is best aligned with the views and values of the country’s voters generally wins the national popular vote.

Former Congressman and presidential candidate Tom Tancredo (R–Colorado) said in an article entitled “Should Every Vote Count?”

“There is another reason why I have come to support the concept of the National Popular Vote Initiative. I believe, as do many of my readers, we are a center-right nation.”[555]

Those who believe that the United States is inherently a center-right country should expect center-right results from a national popular vote for President. Those who believe that there is no bias in the national popular vote—including the authors of this book—should prefer a level playing field that eliminates the gaming of the system inherent in presidential campaigns that concentrate on only a handful of closely divided battleground states.

Footnotes

Republican voters support the idea of a national popular vote for President by an average of 66% in 35 states where state-level polls are available.

Table 9.26 shows the results, by party, from these polls.[556]

Table 9.26 Results, by party, from state-level polls

 

Footnotes

The myth that the small states (i.e., those with three or four electoral votes) confer a partisan advantage on the Republican Party is prevalent because it was once true. However, this statement is not true today, and it has not been true for two decades.

In the 1960s and 1970s, most of the 13 smallest states usually voted Republican in most presidential elections. During that period, Rhode Island, Hawaii, and the District of Columbia were usually the only small jurisdictions that voted Democratic.

However, in the six presidential elections in the two-decade period between 1992 and 2012, seven of the 13 smallest states have gone Democratic (with only one exception in 2000[557]), namely

• Delaware,
• the District of Columbia,
• Hawaii,
• Maine,
• New Hampshire,
• Rhode Island, and
• Vermont.

During the same two-decade period, six of the 13 smallest states have gone Republican (with only one exception in 1992[558]), namely

• Alaska,
• Idaho,
• Montana,
• North Dakota,
• South Dakota, and
• Wyoming.

Only one of the 13 smallest states (New Hampshire) has been a closely divided battleground state during this two-decade period. Although it has been hotly contested, New Hampshire has ended up supporting the Democratic nominee in five of the six elections between 1992 and 2012.

Curiously, the Democratic presidential candidate has sometimes enjoyed a distinct political advantage among the small states because of the state-by-state winner-take-all system.

In 2004, Senator John Kerry won more electoral votes than President George W. Bush in the 13 smallest states (25 for Kerry to 19 for Bush), despite the fact that Kerry received only about two-thirds as many popular votes as Bush (453,286 for Kerry and 650,421 for Bush).

Table 9.27 shows the 2004 presidential election results in the six reliably Republican small states. The table shows that George W. Bush’s 650,421-vote lead in the six reliably Republican small states yielded him 19 electoral votes.

Table 9.27 Bush’s 650,421-vote lead in the six reliably Republican small states yielded 19 electoral votes.

 

Table 9.28 shows the 2004 presidential election results in the seven usually-Democratic small states. The table shows that John Kerry’s 453,286-vote lead yielded him 25 electoral votes. In other words, Kerry won more electoral votes than Bush with considerably fewer popular votes.

Table 9.28 Kerry’s 453,286 vote lead in the seven usually Democratic small states yielded 25 electoral votes.

 

The reason for this outcome under the current winner-take-all system is that the small red states are redder than the small blue states are blue.

Specifically, the popular-vote percentages in the reliably Republican six small states in 2004 were uniformly overwhelming:

• Alaska–64%,
• Idaho–69%,
• Montana–61%,
• North Dakota–64%,
• South Dakota–61%, and
• Wyoming–70%.

In contrast, the Democrats won three of their small states (Delaware, Hawaii, and Maine) with just 54% of the vote.[559] In addition, the Democrats carried two of their small states (Vermont and Rhode Island) with only 60% of the vote—a percentage smaller than the percentage by which the Republicans carried any of their six small states. The District of Columbia (with three electoral votes) is the only small jurisdiction where the Democrats won by an overwhelming margin. The Democrats won the battleground state of New Hampshire by a 2% margin in 2004.

Overall, an enormous number of Republican votes in the small states were wasted because of the overwhelming victory margins in the six reliably Republican small states, compared to the Democrat’s modest margins of victory in their states. This can be seen by pairing each of the six Republican states with one of the Democratic states.

• Wyoming’s 96,509-vote Republican margin exceeded Vermont’s 62,911-vote Democratic margin.
• Alaska’s 65,812-vote Republican margin exceeded Delaware’s 28,356-vote Democratic margin.
• North Dakota’s 85,336-vote Republican margin exceeded Hawaii’s 37,209-vote Democratic margin.
• Montana’s 92,110-vote Republican margin exceeded Rhode Island’s 85,753-vote Democratic margin.
• South Dakota’s 83,320-vote Republican margin exceeded Maine’s 65,017-vote Democratic margin.
• Idaho’s 227,334-vote Republican margin exceeded the District of Columbia’s 164,869-vote Democratic margin.

To place the magnitude of these wasted Republican votes into perspective, consider the fact that George W. Bush’s margin of 227,334 votes in 2004 in Idaho alone was almost twice his margin of 118,599 votes in the crucial and decisive state of Ohio. Presidential candidates of both parties vigorously solicited votes in Ohio on the basis of Ohio issues and values because Ohio voters were important, while they ignored Idaho issues and values.

Even if one expands the discussion from the nation’s 13 smallest states (i.e., those with three or four electoral votes) to the 25 smallest states (i.e., those with seven or fewer electoral votes), the Republican Party receives no partisan advantage under the state-by-state winner-take-all system.

In the 2008 election, the 25 smallest states

• divided 12–13 by party,
• divided 57–58 in electoral votes, and
• the Democrats led with 9,965,724 votes (compared to the Republicans’ 9,821,558 votes).

Table 9.29 shows that the 25 smallest states divided almost equally in 2008 in terms of number of states won, electoral votes, and the popular vote. Column 1 shows each state’s number of electoral votes (EV). Columns 3 and 4 show the number of popular votes won by the Democrats (D) and the Republicans (R), respectively. Columns 5 and 6 show the number of electoral votes won by the Democrats and the Republicans, respectively.[560] Columns 7 and 8 show the Democratic and Republican margins, respectively, for each state that the party carried. Column 9 shows the number of campaign events (a total of 43) out of 300 post-convention events in these states in 2008.

Table 9.29 The 25 smallest states divided almost equally in 2008.

 

In the 2012 election, the 25 smallest states

• divided 12–13 by party (exactly the same states and numbers),
• divided 60–56 in electoral votes, and
• the Republicans led with 10,098,119 votes (compared to the Democrats’ 9,221,230 votes).

Table 9.30 shows that the 25 smallest states divided almost equally in 2012 in terms of number of states won, electoral votes, and the popular vote. Column 9 shows the number of campaign events (a total of 53) out of 253 post-convention events in these states in 2012.

Table 9.30 The 25 smallest states divided almost equally in 2012.

 

Appendices CC, DD, and EE show the popular vote for President for 2000, 2004, and 2008, respectively. Appendix HH shows the 2012 results.

Former Congressman and presidential candidate Tom Tancredo (R–Colorado) wrote the following in an article entitled “Should Every Vote Count?”

“Today the chase for electoral votes is a force for corruption and special-interest payoffs. I will never forget the torture of sitting in the House and watching as our ‘leadership’ went about threatening, bribing and breaking arms of my colleagues until they got the requisite number of votes to pass Bush’s trillion-dollar Medicare prescription drug plan. A bigger piece of garbage I have never seen—especially one being pushed by the Republican Party.

“One could rationally ask why, in heaven’s name, the party of smaller government would push so hard for what was, at the time, the biggest increase in government since the creation of Medicare. Alas the reason was crystal clear: Bush needed Florida for his re-election.

“I wish I could say that was the only time something like that happened, but, of course, it’s not. It is part of the routine practice of buying electoral votes. I am sick of it. Whether it's buying Pennsylvania’s electoral votes with steel tariffs or Ohio’s with ‘No Child Left Behind,’ it all stinks to high heaven.…

“Some argue that the present system protects the interests of small states, especially those that hold conservative values. However, today 12 of the 13 smallest states are ignored after party conventions and are derisively referred to as ‘flyover’ country.…

“Under the [National Popular Vote] plan, an evangelical voter in rural Wyoming would count the same as the union steward in Cleveland.”[561] [Emphasis added]

Footnotes

Hans von Spakovsky has stated:

“National Popular Vote Inc. is one of California’s lesser-known advocacy organizations. Its chairman, John Koza, is best known as the co-founder of Scientific Games Inc., the company that invented the instant lottery ticket.

“Now Mr. Koza and his fellow liberal activists want to ‘scratch off’ the Electoral College.”[562] [Emphasis added]

The facts are that over 90% of the contributions supporting the National Popular Vote effort have come—in about equal total amounts—from

• Tom Golisano (a pro-life, anti-Buffett-rule, registered Republican businessman residing in Florida) and
• John R. Koza (a pro-choice, pro-Buffett-rule, registered Democratic businessman residing in California).

John R. Koza’s contributions have largely been spent by National Popular Vote, a 501(c)4 non-profit corporation.

Tom Golisano’s contributions have largely been spent by Support Popular Vote, a 501(c)4 non-profit corporation (originally called “National Popular Vote Initiative”).

Support for a nationwide popular vote for President has been bipartisan for some time. Appendix S shows, state by state, members of Congress who have sponsored proposed constitutional amendments for nationwide popular election of the President in recent years or who voted in favor of constitutional amendments in the 338–70 roll call in the House of Representatives in 1969 or the 1979 roll call in the Senate. As shown in appendix S, there has been at least one supporter in Congress from each of the 50 states. As of 2012, over 250 Republican state legislators have either sponsored or cast a recorded vote in favor of the National Popular Vote bill. See section 9.31.2 for recent state-level polling results showing that Republican voters support a nationwide vote for President.

Footnotes

As a result of each recent census, Republican-leaning states have gained population (and hence electoral votes) at the expense of Democratic-leaning states. Some have argued that this fact should be interpreted as a long-term trend favoring the Republican Party in the Electoral College. In fact, this trend is not necessarily helpful to the Republican cause.

Consider the 2010 census. The Republican Party would have received 12 more electoral votes in the 2008 presidential election if the allocation of electoral votes based on the 2010 census had been in effect for the 2008 election. Five states that voted Republican in the 2008 presidential election gained electoral votes as a result of the 2010 census, namely Arizona (+1), Georgia (+1), South Carolina (+1), Utah (+1), and Texas (+4), but only two states that voted Republican in 2008 lost electoral votes, namely Louisiana (–1) and Missouri (–1). In addition, eight states that voted Democratic in the 2008 presidential election lost electoral votes as a result of the 2010 census, namely Illinois (–1), Iowa (–1), Massachusetts (–1), Michigan (–1), New Jersey (–1), New York (–2), Ohio (–2), and Pennsylvania (–1), but only three states that voted Democratic in 2008 gained electoral votes, namely Florida (+2), Nevada (+1), and Washington state (+1).[563]

The above facts about the census do not, however, constitute a long-term trend favoring the Republicans in the Electoral College because population growth does not necessarily reinforce a state’s pre-existing political complexion. In fact, population growth frequently upsets a state’s political complexion.

Population growth occurs as the result of a net difference in the number of newcomers versus the number of leavers.

There is usually a considerable difference in the political outlook of

• newcomers to a state,
• leavers, and
• those staying in a state.

People come to a state, leave a state, and stay in a state because of numerous economic, demographic, and psychological factors. As a result, population growth is not necessarily advantageous to the currently dominant political party in a given state.

For example, Florida, Virginia, Colorado, Nevada, and North Carolina were reliably Republican for decades in presidential elections until recently. Virginia, Colorado, Nevada, and North Carolina were not even considered battleground states as recently as 2004. Rapid population growth converted Florida into a battleground state in 1996 (when Clinton carried the state after several decades of Republican victories at the presidential level). However, population growth upset the political equilibrium of these states with the result that Obama swept all of these states in 2008. Population growth not only contributed to the Republican’s loss of all these states in 2008, but also increased the electoral-vote prize when the Democratic Party won them.

Arizona’s recent rapid population growth (largely due to an influx of Hispanics and, to a lesser extent, newcomers from California) appears to be transforming it from a reliably Republican state in presidential elections into a battleground state (perhaps as early as 2016).[564]

Rapid population growth (largely due to Hispanics) in Texas (with 38 electoral votes) creates the possibility of destabilizing Republican control of the nation’s second largest state (perhaps as early as 2020). As Charles Mahtesian wrote in a Politico article entitled “Obama’s Texas Battleground Prediction”:

“When Barack Obama asserted Tuesday that Texas will be a battleground state ‘soon,’ he was echoing the belief, commonly held among Democrats, that the state’s changing demographics make the transition from red to blue inevitable.”[565],[566] [Emphasis added]

Meanwhile, there does not appear to be any Democratic-leaning big state (even among the numerous Democratic-leaning states that lost electoral votes as a result of the 2010 census) moving in the Republican direction to counter-balance possible future changes in the political environment in states such as Arizona and Texas.

Footnotes

The 12 biggest states contain more than half the population of the United States and possess 53% of the electoral votes (283 of 538). In fact, the 11 biggest states contain a bare majority of the electoral votes (270 of 538).

Critics of a nationwide popular vote for President sometimes argue that only the 12 biggest states would matter under such a system.

Under the critics’ hypothetical scenario, candidates would win the White House by winning 100% of the popular vote in the 12 biggest states and 0% in the 39 remaining jurisdictions (i.e., 38 states and the District of Columbia).

Referring to the National Popular Vote compact, Hans A. von Spakovsky stated in 2011:

“This would give the most populous states a controlling majority of the Electoral College, letting the voters of as few as 11 states control the outcome of presidential elections.”[567] [Emphasis added]

Senator Mitch McConnell said in 2011:

“This would mean that from now on, just 12 states could decide our presidential elections. A few of the most populous and most liberal states determine who actually wins.”[568] [Emphasis added]

Ed Gillespie stated in 2011:

“With 11 of the most populous states accounting for 56 percent[569] of the population, the presidential election will essentially become a race for a dozen states with big cities.”[570] [Emphasis added]

A 2011 letter signed by House Speaker John Boehner (R–Ohio), Senator Mitch McConnell (R–Kentucky), and Governor Rick Perry (R–Texas) stated:

“The goal of this effort is clear: to put the fate of every presidential election in the hands of the voters in as few as 11 states and thus to give a handful of populous states a controlling majority of the Electoral College.”[571] [Emphasis added]

None of the above quotations about 11 or 12 states “controlling” the national popular vote reflects political reality.

It is the current state-by-state winner-take-all system—not the national popular vote approach—that would theoretically permit the 11 most populous states to control the outcome of presidential elections.

Under the current winner-take-all system, a candidate could win the Presidency by winning only 50.01% of the popular vote in the 11 biggest states. That is, under the current system, a President could be elected with about a quarter of the nationwide popular vote.

Opponents of a nationwide vote for President complain that if 100% of the voters of the 11 biggest states were to vote for one candidate, they could alone elect a President—while ignoring the fact that 50.01% of the voters of these same 11 states could elect a President today under the current state-by-state winner-take-all system.

That is, 26% of the nation’s voters could elect a President under the current system.

Moreover, getting 50.01% in 11 states is a far more likely scenario than getting 100% of the vote from these 11 states.

Curiously, the current system permits even fewer than 26% of the voters to elect a President. According to calculations (shown in table 9.31) made by MIT Professor Alexander S. Belenky using actual voter turnout data, an Electoral-College majority theoretically could have been won, under the current winner-take-all system, with between 16% and 22% of the national popular vote in the 15 elections between 1948 and 2004.[572]

Table 9.31 Smallest percentage of voters who theoretically could have elected a President under the current system

 

The implausibility of the hypothetical scenario in which one candidate receives 100% of the popular vote from the 12 biggest states is demonstrated by the fact that no big state delivered more than 63% of its popular vote (that is, five out of eight votes) to any candidate in the 2000, 2004, 2008, or 2012 presidential elections.

Table 9.32 shows the percentage of the popular vote won by the winner of the 12 biggest states between 2000 and 2012.

Table 9.32 Popular-vote percentage won by the winner of the 12 biggest states 2000–2012

 

In fact, many of the winning percentages in table 9.32 are near 50% because many of the 12 biggest states (e.g., Ohio, Florida, Virginia, Pennsylvania, Michigan, and North Carolina) were battleground states in one or more elections shown in the table.

The 12 biggest states are not, of course, all Democratic bastions. In both 2000 and 2004, for example, the 12 biggest states divided 6–6 between the political parties (with Texas, Florida, Ohio, North Carolina, Georgia, and Virginia voting for George W. Bush in both years).[573]

The popular vote in the 12 biggest states split 54%–46% in 2012 and split 50.2%–49.8% in 2004.

In short, no candidate could win 100% of the popular vote in the 12 biggest states (or, indeed, any percentage close to 100%).

The relatively close campaigns of 2004 and 2012 convey a far more realistic picture of presidential politics than any contrived scenario.

The winner’s two-party popular-vote percentage was almost identical in these two re-election campaigns:

• 51.2% for Bush in 2004,[574] and
• 51.96% for Obama in 2012.[575]

The two elections were mirror images of one another in terms of the popular-vote margin generated by the 12 biggest states and the 39 smallest jurisdictions:

• In 2004, Bush fought Kerry to a near-tie in the popular vote in the 12 biggest states (50.2% to 49.8%), and Bush’s margin from the 39 smallest jurisdictions was roughly equal to his nationwide margin (3,012,171 votes).
• In 2012, Obama fought Romney to a near-tie in the popular vote in the 39 smallest jurisdictions (51% to 49%), and Obama’s margin from the 12 biggest states was roughly equal to his nationwide margin (4,966,945).

In 2004, the voters in the 39 smallest jurisdictions did not “control the outcome of the presidential election” in terms of the national popular vote. Every vote from every state—not just those 39 states—contributed to producing Bush’s nationwide popular vote total. The voters in the 39 smallest jurisdictions were not any more important or “controlling” than the voters of the 12 biggest states.

Similarly, in 2012, the voters in the 12 biggest states did not “control the outcome of the presidential election” in terms of the national popular vote. Every vote from every state contributed to producing Obama’s nationwide popular vote total. The voters in the 12 biggest states were not any more important or “controlling” than the voters of the 39 smallest jurisdictions.

2004—Bush Ties in the 12 Biggest States

In 2004, the 69,323,699 votes cast in the 12 biggest states divided almost equally:

• 34,784,178 votes were for Kerry, and
• 34,539,521 votes were for Bush.

Kerry’s slender 244,657-vote margin of victory in the 12 biggest states was about one-third of one percent of the 69,323,699 votes cast in those states (and about one-fifth of one percent of the votes cast nationwide).

Kerry received 50.2% of the popular vote from the 12 biggest states, and Bush received 49.8%.

Having fought Kerry to a near-tie in the 12 biggest states, Bush then won the 39 smallest jurisdictions by a margin of 3,256,828 votes (out of 51,745,350 votes cast in those states), thereby ending up with a margin of victory of 3,012,171 in the national popular vote.

Table 9.33 shows the popular vote for Senator John Kerry and President George W. Bush in the 2004 election in the 12 biggest states. Column 4 shows Bush’s percentage of the two-party vote. Columns 5 and 6 show the Republican and Democratic margins, respectively, for each state. Columns 7 and 8 show the Republican and Democratic electoral votes, respectively, for each state.

Table 9.33 Results of the 2004 election in the 12 biggest states

 

Table 9.34 shows the popular vote for President George W. Bush and Senator John Kerry in the 2004 election in the 39 smallest jurisdictions.

Table 9.34 Results of the 2004 election in the 39 smallest jurisdictions

 

2012—Obama Ties in the 39 Smallest Jurisdictions

In 2012, the 54,209,884 votes cast in the 39 smallest jurisdictions divided almost equally:

• 26,578,682 votes were for Obama, and
• 27,631,202 were for Romney.

Romney’s 1,052,520-vote margin in the 39 smallest jurisdictions give him a slender 51%–49% win from the 54,209,884 votes cast in those states.

Having fought Romney to a near-tie in the 39 smallest jurisdictions, Obama then won the 12 biggest states by a margin of 6,019,465 votes (out of 72,618,625 votes cast in those states), thereby ending up with a margin of victory of 4,966,945 in the national popular vote.

Table 9.35 shows the popular vote for Governor Mitt Romney and President Barack Obama in the 2012 election in the 39 smallest jurisdictions.

Table 9.35 Results of the 2012 election in the 39 smallest jurisdictions.

 

Table 9.36 shows the popular vote for Governor Mitt Romney and President Barack Obama in the 2012 election in the 12 biggest states.

Table 9.36 Results of the 2012 election in the 12 biggest states

 

Appendix HH presents the 2012 two-party presidential vote for all 50 states and the District of Columbia in alphabetical order.[576] See table 9.45 for the presidential vote for Barack Obama (D), Mitt Romney (R), Gary Johnson (Libertarian), Jill Stein (Green), and the other 22 minor-party and independent candidates who were on the ballot in 2012 in at least one state.

Erroneous Statements about Big States May Possibly Be the Result of Misunderstanding the Way that the National Popular Vote Compact Operates

The four statements quoted at the beginning of this section are so far removed from what actually happens in the real world that we should mention the following possibility. It is possible that all four statements quoted at the beginning of this section are based on a total misunderstanding of how the National Popular Vote compact would operate.

The National Popular Vote compact would take effect when enacted by states possessing a majority of the electoral votes (270 of 538).

The assertion that the National Popular Vote compact would

“give a handful of populous states a controlling majority of the Electoral College”

could conceivably be true if the National Popular Vote compact were written so that it counted only the popular votes of the states belonging to the compact. If that were the case (and it is not) and if one makes the additional implausible assumption that the compact consisted only of the 12 biggest states, the four statements would be true. However, the National Popular Vote compact would not operate that way even if only the 12 biggest states belonged to the compact.

The National Popular Vote compact would add up the votes cast in all 50 states and the District of Columbia to determine the national popular-vote winner regardless of whether a state is a member of the compact. Under the National Popular Vote compact, every vote in all 50 states would be counted in arriving at the national popular vote total for each candidate. Under the National Popular Vote compact, there would be nothing special about a vote cast in the member states (or in the 12 biggest states) in comparison to votes cast anywhere else. Every vote would be equal throughout the United States under the National Popular Vote compact.

Note also that the National Popular Vote compact has not been enacted primarily by big states. As of 2012, the compact has been enacted by nine jurisdictions, including three small jurisdictions (Hawaii, Vermont, and the District of Columbia), three medium-sized states (Maryland, Massachusetts, and Washington state), and three of the 12 biggest states (California, Illinois, and New Jersey).

Role of Big Cities

Many of the critics of a nationwide popular vote for President who argue that the 12 biggest states would control a nationwide election for President also claim that big cities, such as Los Angeles, would control a nationwide election.

Big cities, such as Los Angeles, do not even control California elections, as evidenced by the historical fact that Republicans Ronald Reagan, George Deukmejian, Pete Wilson, and Arnold Schwarzenegger were all elected Governor without ever carrying Los Angeles (or San Francisco, San Jose, Oakland, or most of the other big cities in the state). If Los Angeles cannot control statewide elections in its own state, it can hardly control a nationwide election.

While is certainly true that most of the biggest cities in the country vote Democratic, smaller cities and towns, exurbs, rural areas, and many suburbs usually vote Republican.

If big cities controlled the outcome of elections, every Governor and every U.S. Senator would be a Democrat in every state with a significant city. There are, of course, examples from every state with a significant city, of Republicans winning races for Governor and U.S. Senator without ever carrying the state’s biggest city.

The origins of the myth about big cities may stem from the incorrect belief that big cities are bigger than they actually are, and that big cities account for a greater fraction of the nation’s population than they actually do.

A look at our country’s actual demographics contradicts these misconceptions concerning big cities.

Table 9.37 shows the population of the nation’s 50 biggest cities according to the 2010 census.

Table 9.37 Population of the 50 biggest U.S. cities

 

The combined population of the nation’s five biggest cities (New York, Los Angeles, Chicago, Houston, and Philadelphia) constitutes only 6% of the nation’s population of 308,745,538 (based on the 2010 census).

The combined population of the 20 biggest cities constitutes only 10% of the nation’s population. To put this group of 20 cities in perspective, Memphis is the nation’s 20^th biggest city. Memphis had a population of 646,889 in 2010.

The combined population of the 50 biggest cities constitutes only 15% of the nation’s population. To put this group of 50 cities in perspective, Arlington, Texas, is the nation’s 50^th biggest city (and had a population of 365,438 in 2010).

To put it another way, 85% of the population of the United States lives in places with a population of less than 365,000 (the population of Arlington, Texas).

Moreover, the population of the nation’s 50 biggest cities is declining. In 2000, the 50 biggest cities together accounted for 19% of the nation’s population (compared to 15% in 2010).

Even if one makes the far-fetched assumption that a candidate could win 100% of the votes in the nation’s 50 biggest cities, that candidate would win only 15% of the national popular vote.

In a nationwide vote for President, a vote cast in a big city would be no more (or less) valuable or controlling than a vote cast in a suburb, an exurb, a small town, or a rural area.

When every vote is equal and the winner is the candidate who receives the most popular votes, candidates know that they need to solicit voters throughout their entire constituency in order to win.

Perhaps the most convincing evidence for the fact that big cities do not control elections comes from looking at the way that presidential races are actually run today.

Inside a battleground state in a presidential election today, every vote is equal, and the winner is the candidate who receives the most popular votes.

When presidential candidates campaign to win the electoral votes of a closely divided battleground state, they campaign throughout the state. The big cities do not receive all the attention—much less control the outcome. Cleveland and Miami have certainly not received all the attention when presidential candidates have campaigned in the closely divided battleground states of Ohio and Florida. Moreover, Cleveland and Miami manifestly do not control the statewide outcomes in Ohio and Florida, as evidenced by the outcome of the 2000 and 2004 presidential elections in those states. The Democrats carried both Cleveland and Miami in 2000 and 2004, but the Republicans carried both states. In fact, Senator John Kerry won the five biggest cities in Ohio in 2004, but he did not win the state.

In summary, under the National Popular Vote compact, every vote would be equal throughout the United States. Votes cast in all 50 states and the District of Columbia would be added together to determine the national popular vote winner. A vote cast a big city or big state would be no more, or less, valuable or “controlling” than a vote cast anywhere else.

Footnotes

John Samples of the Cato Institute has stated:

“NPV will encourage presidential campaigns to focus their efforts in dense media markets where costs per vote are lowest.…

“In general, because of the relative costs of attracting votes, the NPV proposal seems likely at the margin to attract candidate attention to populous states.”[577] [Emphasis added]

Claremont College Professor Michael Uhlmann stated in a January 20, 2012, debate at the Sutherland Institute in Salt Lake City:

“Under the National Popular Vote system, necessarily, there’s going to be tilting toward where the greater masses of votes are contained—in the larger cities and the immediate suburbs. That’s where the votes are. That’s where they can be reached the most cheaply. That’s where the maximum bang for the media buck gets paid. I think that’s the likely tendency.”[578] [Emphasis added]

The arguments made by both Samples and Uhlmann are contrary to the facts.

Television advertising (by far the costliest component of presidential campaigns) is generally higher on a per-impression basis in the larger media markets than in smaller markets.

Based on 488 quotations from television stations in media markets of various sizes for 30-second prime-time television ads for the weeks of October 15 and 22, 2012, compiled by Ainsley-Shea (a Minneapolis public relations firm) in July 2012, the average cost per impression was:

• 4.235 cents for the 1^st–5^th markets,
• 4.099 cents for the 26^th–30^th markets, and
• 3.892 cents for the 101^st–105^th markets.

The details of television advertising costs in the 1^st, 26^th, and 101^st largest media markets further illustrate the conclusion that television advertising is generally more expensive in the larger media markets than in smaller markets.

Table 9.38 shows the cost of a 30-second prime-time television slot in New York City—the nation’s No. 1 media market. Columns 1, 2, and 3 show the station, the time of day (all P.M.), and the program name, respectively. Columns 4, 5, and 6 show the rating,[579] share, and gross rating points (GRP), respectively, for adults age 18 and older. Column 7 shows the cost of the slot. Column 8 shows the cost per 1,000 impressions (that is, the cost in column 7 divided by the media market’s population of 15,334,000). The average cost for New York City was $51.90 per 1,000 impressions—5.190 cents per impression.

Table 9.38 Television ads in New York City—the nation’s No. 1 media market—averaged 5.190 cents per impression.

 

The similarly computed cost of a 30-second prime-time television slot in Los Angeles—the nation’s No. 2 media market—averaged $56.53 per 1,000 impressions—5.653 cents per impression.

Table 9.39 shows the cost of a 30-second prime-time television slot in Indianapolis—the nation’s No. 26 media market. Column 8 shows the cost per 1,000 impressions (that is, the cost in column 7 divided by the market’s population of 2,094,000). The average cost for Indianapolis was $39.80 per 1,000 impressions—3.980 cents per impression.

Table 9.39 Television ads in Indianapolis—the nation’s No. 26 media market—averaged 3.980 cents per impression.

 

Table 9.40 shows the cost of a 30-second prime-time television slot in the nation’s No. 101 media market—Fort Smith, Fayetteville, Springdale, and Rogers, Arkansas. Column 8 shows the cost per 1,000 impressions (that is, the cost in column 7 divided by the market’s population of 573,000). The average cost for this market is $30.84 per 1,000 impressions—3.084 cents per impression.

Table 9.40 Television ads in the Fort Smith, Fayetteville, Springdale, and Rogers, Arkansas market—the nation’s No. 101 media market—averaged 3.084 cents per impression.

 

An NPR story entitled “Ads Slice Up Swing States With Growing Precision” reported on presidential campaigning in Colorado’s small media markets:

“Republicans outnumber Democrats in El Paso County more than 2 to 1. Barack Obama lost this part of Colorado to John McCain by 19 points in 2008.

“‘It's not a matter of just winning; it's winning by how much,’ says Rich Beeson, a fifth-generation Coloradan and political director for the Romney campaign.

“Presidential campaigns know exactly the margin of victory or defeat that they have to hit in each town in order to carry an entire state. Democratic media strategist Tad Devine says campaigns set extremely specific goals based on hard data.…

“Although no one suggests that President Obama will win Colorado Springs, whether he loses it by 15 or 25 points could determine whether he carries Colorado.

“Beeson of the Romney campaign says smaller cities are vital to this chess game, especially since they're cheaper to advertise in.

“‘A lot of secondary markets are very key to the overall map, whether it's a Charlottesville in Virginia or a Colorado Springs in Colorado,’ he says. ‘You can’t ever cede the ground to anyone.’”[580] [Emphasis added]

Soliciting every available vote is a strategic necessity when the winner of an election is the candidate who receives the most popular votes.

Footnotes

Under federal law, non-citizens cannot vote in presidential elections. Nonetheless, non-citizens significantly impact presidential elections because they affect the allocation of electoral votes among the states.

As Professor George C. Edwards III has pointed out:

“Representation in the House is based on the decennial census, which counts all residents—whether citizens or not.… States … where non-citizens compose a larger percentage of the population receive more electoral votes than they would if electoral votes were allocated on the basis of the number of a state’s citizens.”[581]

The U.S. Constitution requires that the census be used to determine each state’s number of seats in the U.S. House of Representatives. Each state receives a number of electoral votes equal to the state’s number of U.S. Representatives plus two (representing the state’s two U.S. Senators).

The Constitution specifies that the census count all “persons,” thereby including non-citizens living in the United States in the count:

“Representatives … shall be apportioned among the several States which may be included within this Union, according to their respective Numbers, which shall be determined by adding to the whole Number of free Persons, including those bound to Service for a Term of Years, and excluding Indians not taxed, three fifths of all other Persons.”[582],[583] [Emphasis added]

The Census Bureau uses a mathematical formula (specified by a federal statute adopted in 1941) known as the “method of equal proportions” to apportion seats in the U.S. House of Representatives automatically among the states.[584]

A state having a disproportionally large number of non-citizens (relative to other states) acquires additional U.S. House seats and, hence, additional electoral votes.

Because of the winner-take-all rule, legal voters in a state that acquired additional electoral votes by virtue of the disproportionate presence of non-citizens deliver an enlarged bloc of electoral votes to the candidate receiving the most popular votes in their state. That is, the influence of the legal voters is increased because of the presence of non-citizens.

Similarly, legal voters in a state that lost electoral votes deliver a diminished bloc of electoral votes.

The apportionment of the U.S. House and Electoral College resulting from the 2010 census governs the 2012, 2016, and 2020 elections.

Professor Leonard Steinhorn of American University has computed the effect of non-citizens on presidential elections. He plugged American Community Survey data on the number of citizens and non-citizens in each state in 2010 into the statutory formula to apportion U.S. House seat among the states.[585]

In an article entitled “Without Voting, Noncitizens Could Swing the Election for Obama,” Steinhorn found that non-citizens affected the number of electoral votes possessed by 15 states.

Five states gained between one and five electoral votes, and 10 states each lost one electoral vote because of non-citizens.

Overall, the Democrats have a net 10 electoral-vote advantage in the 2012, 2016, and 2020 elections from the 15 states whose representation was affected by the counting of non-citizens in allocating electoral votes among the states.

• Democratic non-battleground states gained 7 electoral votes:
• +5 for California
• +1 for New York
• +1 for Washington state.
• Republican non-battleground states lost 3 electoral votes:
• +2 for Texas.
• –1 for Indiana
• –1 for Missouri
• –1 for Louisiana
• –1 for Montana
• –1 for Oklahoma.
• Six battleground states were affected:
• +1 Florida
• –1 for Iowa
• –1 for Michigan
• –1 for North Carolina
• –1 for Ohio
• –1 for Pennsylvania.

Battleground states can, by definition, go either way, and therefore do not constitute a built-in advantage to either party.

Excluding non-citizens from the calculation used to apportion seats in the U.S. House of Representatives would require a federal constitutional amendment.

The National Popular Vote compact would eliminate the distortion in presidential elections caused by the disproportionate presence of non-citizens in certain states. Nationwide voting for President would equalize the vote of every legal voter in the country by guaranteeing the Presidency to the candidate who receives the most popular votes in all 50 states and the District of Columbia.

Footnotes

An argument became prevalent during the 1980s that the Republican Party had a permanent “lock” on the Electoral College because a large number of states had repeatedly voted Republican for President between 1968 and 1988.

The notion of a “lock” arose from the fact that Republicans won five of the six presidential elections during this period, and that Republicans won landslide victories in 1972 and 1984.

In fact, neither party has a lock on the Electoral College because the United States is, politically, an evenly divided country in which the cumulative nationwide vote for the two parties from the start of the modern political era in 1932 through 2008 (table 9.25) has been virtually tied:

• 746,260,766 total votes for the Democrats and
• 745,502,654 total votes for the Republicans.

The Republican Party won five of the six presidential elections between 1972 and 1984. The reason for this result was that more voters (often in landslide numbers) voted for the Republican nominee during that period—not because of the mechanics of the Electoral College.

In any event, the Republican Party does not have any such “electoral lock” today.

To the extent that this kind of “electoral lock” argument has a small element of validity, if the Electoral College map of 2012 were to persist, the electoral map would, if anything, be slightly be unfavorable to the Republican Party.

Table 9.41 shows that 32 states that voted for the same party in all six presidential elections between 1992 and 2012. These 32 states possess about two-thirds (64%) of the 538 votes in the Electoral College. Of these 32 states, 19 states (possessing 242 electoral votes after the 2010 census) voted Democratic in all six presidential elections between 1992 and 2012, and 13 states (possessing 102 electoral votes after the 2010 census) voted Republican in the six elections. The table is organized in terms of number of elections (from zero to six) in which a state voted Democratic. The number of electoral votes shown in the table are those applicable to the 2012 election.

Table 9.41 The 32 states that voted for the same party in the six presidential elections between 1992 and 2012

 

Table 9.41 reflects one aspect of the current polarization of American politics. One possible cause of this polarization may be the tendency, discussed in Bill Bishop’s book The Big Sort, of like-minded Americans to cluster together geographically.[586]

Regardless of the causes behind the behavior shown in table 9.41, if the 2016 presidential election is conducted under the state-by-state winner-take-all rule and is reasonably close, it is likely that most of the 32 states that have voted consistently for the same party in the past six presidential elections would continue to support that same party.[587]

In any event, table 9.41 certainly does not support the notion of the existence today of an “electoral lock” in favor of the Republican Party.

Table 9.42 shows a simulation of the 2012 presidential election produced by applying a tie-producing uniform shift to actual election returns (as shown in table 9.35, table 9.36, table 9.45, and appendix HH). In 2012, Governor Romney received 48.0418657% of the two-party national popular vote—a shortfall of 1.9581343%. Column 2 shows the simulated figures for Romney obtained by applying a uniform upward adjustment of 1.9581343% to Romney’s actual vote in each state (and a corresponding downward adjustment to Obama’s actual vote in each state), thereby producing a virtual tie in the national popular vote (63,414,254 to 63,414,255). Column 4 shows Romney’s percentage of the two-party vote using this method of simulation. Columns 5 and 6 show the Republican and Democratic margins, respectively, for each state using this method of simulation. Columns 7 and 8 show the Republican and Democratic electoral votes, respectively, for each state using this method of simulation. The table is sorted according to the simulated Republican percentage in column 4.

Table 9.42 Simulated tie-producing uniform shift of 2012 election data

 

The result of the tie-producing uniform shift shown in table 9.42 is that President Obama loses Florida (29 electoral votes) and Ohio (18 electoral votes), but still ends up with a 285–253 lead in the Electoral College. Thus, even if Romney had received enough additional voter support to create a tie in the national popular vote (preserving each candidate’s relative profile in each state), Obama would still have ended up with a lead of 28 electoral votes using this method of simulation.

Table 9.42 also shows that Obama’s lead in Virginia (13 electoral votes) shrinks to an eminently recountable 701 votes (1,897,522 to 1,896,820) using this method of simulation. Even if Romney had won Virginia, Obama would still have had a 272–266 lead in the Electoral College.

In a second simulation (shown in table 9.43), Romney’s actual results are adjusted uniformly upward by 2.732% in each state (with Obama’s vote receiving a corresponding downward adjustment in each state). This adjustment would give Romney a lead of 1,962,965 votes nationwide (64,395,737 to 62,432,772). This adjustment is just sufficient to move both Virginia and Pennsylvania (by 8 votes) into Romney’s column, thus giving Romney a winning 286–252 margin in the Electoral College. The table is sorted according to the simulated Republican percentage in column 4.

Table 9.43 Simulated uniform shift producing a 1,962,965-vote nationwide lead for Romney

 

In other words, it takes a national popular vote lead of almost two million votes to yield a simulated win for Romney in the Electoral College using this method of simulation.

If Romney’s simulated lead were to be increased slightly beyond the 1,962,965-vote nationwide lead shown in table 9.43, Colorado (nine electoral votes), New Hampshire (four electoral votes), Iowa (six electoral votes), and Nevada (six electoral votes) would move into the Republican column.

Of course, no future election will exactly replicate the state-by-state percentage contour of the two major parties in 2012. President Obama cannot run for another term, and Governor Romney will almost certainly not be a candidate in 2016. Candidates with different personalities and records will compete on the basis of different issues in a political environment consisting of a different history of immediate past events and changed demographics.

Nonetheless, the simulations in table 9.42 and table 9.43 certainly do not support the notion of the existence today of an “electoral lock” in favor of the Republican Party.

Footnotes