An Experiment Regarding Senate Elections 1990-present

[Vosem]

While perusing this site's page for the 2012 Senate elections, I noticed that while the popular vote was 53D-44R-2I-1L, the actual percentages among which way the seats went was 70D-24R-6I. And I'm not complaining -- I'm a fan of FPTP, and I accept that this is how the US system works.

But what if the percentage of how Senate seats always 'went' always as closely as possible reflected the popular vote? So, for instance, in 2012, of 33 seats, Democrats would have won 17; Republicans would have won 15; and independents would have won exactly 1 seat. (The real numbers were 23-8-2). Thus, in the experiment, I would shift the most marginal independent seat (Maine) to the Republicans, and the 6 most marginal Democratic seats (North Dakota, Montana, Virginia, Ohio, New Mexico, Wisconsin) to the Republicans to create this number.

But, of course, simply showing the 2012 results wouldn't be very interesting. Atlas records for non-paying members begin in 1990, where by coincidence the results actually do directly reflect the popular vote (Democrats won, the popular vote, 51D-47R, and the number of seats 18-17). So I will begin in 1992. Rather than merely parroting the results, however, I would like to see how this would change the balance in the US Senate; for instance, people who might have left politics in real life would seek reelection in this timeline (and vice versa). Also, such a system will probably occasionally bring in third-party candidates as well.

I will begin, in my next post, with 1992.

[Barnes]

What an interesting idea!  Looking forward to it.

[Vosem]

What an interesting idea!  Looking forward to it.

Thank you very much

In 1992, the popular vote was 51% Democratic, 45% Republican, 1% Libertarian, 1% Independent, 1% American Independent, and 1% Peace and Freedom. If we correspond to the mathematically correct laws of rounding, this becomes 18 Democratic seats, 16 Republican seats, and 2 vacancies. However, one of the vacancies is merely due to rounding; if I move on to the hundredths place, one of the vacancies goes to the Libertarian Party. Another remains. The vague group 'independent' came closest to a seat (going out to the hundredths place), so I will give it to them.

Thus, in reality, Democrats won these Senate elections 22-14. In this timeline, however, they win them 18-16-1-1.

The best-performing Libertarian candidate in 1992 was John Perry III, of Pennsylvania, who in this timeline is given a Senate seat (replacing Arlen Specter). The best-performing independent was Evan Mecham, of Arizona, who also wins in this timeline (replacing John McCain – obviously McCain losing reelection to a second term in this timeline will have major butterflies down the road...)

This means that, to bring the numbers to 18-16-1-1, the four closest Democratic victories must be given to the Republicans. These are South Carolina (in this timeline Thomas Hartnett defeats Ernest Hollings), California's regularly scheduled election (in this timeline Bruce Herschensohn defeats Barbara Boxer to replace Alan Cranston), meaning both of California's seats switch parties in 1992; Wisconsin (where Bob Kasten is simply reelected); and Washington (where Rod Chandler defeats Patty Murray for the right to replace Democrat Brock Adams). Since there were two mutually exclusive pickups in California, which is something the Electoral College Calculator can't really express, I've elected to simply have the map declare there were no Senate elections in California in 1992.

So, here is the map:

I'm sorry that it's difficult to tell the difference between the independent pickup in Arizona and the Libertarian pickup in Pennsylvania, but it should be made clear by the content.

I think it is reasonable to assume that in the aftermath of the election, the newly rehabilitated Evan Mecham would likely choose to caucus with the Republican Party, and indeed would eventually be absorbed into it. For the purposes of counting he will be a Republican. However, James Perry elects to form a brand-new Libertarian caucus, as he knows that more Libertarians will be elected in the coming years.

Thus, the ultimate result of the 1992 elections is five Republican pickups, one Democratic pickup, and also one Libertarian pickup (from the Republicans). This would leave the Senate with 52 Democrats, led by George Mitchell, 47 Republicans, led by Bob Dole and 1 Libertarian, led by himself (James Perry). Democrats maintain control of the chamber.

[Vosem]

In 1994, the popular vote was 51% Republican, 45% Democratic, 1% Libertarian, 1% Independent, 1% Taxpayers', and 1% Peace and Freedom. If we correspond to the mathematically correct laws of rounding, this becomes 18 Republican seats, 16 Democratic seats, and 1 vacancy. (The reason there were 2 in 1992 but just 1 now is that there were 36 seats up for election in 1992 but only 35 in 1994). The party that came closest to another seat is the Libertarians, who are therefore given that seat.

Thus, in reality, Republicans won these Senate elections 21-14. In this timeline, however, they win them 18-16-1. The best-performing Libertarian in 1994 was Scott Grainger, of Arizona, who in this timeline is given a Senate seat (Jon Kyl won this seat in real life, but this is actually a pickup from the Democrats, who held this seat with Dennis deConcini prior to the '94 elections).

This means that, to bring the numbers to 18-16-1, the two closest Republican victories must be given to the Democrats. These are Pennsylvania (in this timeline Harris Wofford wins reelection against Rick Santorum) and Minnesota (which actually becomes a Democratic pickup; in real life David Durenberger retired and was replaced by Republican Rod Grams, but here he is replaced by Democrat Ann Wynia).

So, here is the map:

Keep in mind that, like in real life, there were actually two Republican pickups in Tennessee, not just one.

Thus the ultimate result of the 1994 elections 6 Republican pickups, 1 Democratic pickup, and 1 Libertarian pickup (from the Democrats). This would leave the Senate with 52 Republicans (led by Bob Dole), 46 Democrats (led, after George Mitchell's retirement, by Tom Daschle), and 2 Libertarians (led by James Perry). However, there is no reason to believe that Richard Shelby of Alabama and Ben Campbell of Colorado, who both switched parties in the aftermath of the Republican victory in 1994, wouldn't also switch in this timeline, giving Republicans an advantage of 54-44-2 over the Democrats. There is also no reason to think the special Senate race in Oregon (which was a Democratic pickup) would've gone any differently; it would've narrowed the Republican advantage back down to 53-45-2. (Also, like in real life, after Bob Dole's resignation to pursue his run for the Presidency Trent Lott would likely have taken over the Senate Republican caucus).

[Vosem]

In 1996, the popular vote was 49% Republican, 48% Democratic, 1% Libertarian, 1% Reform, and 1% Natural Law. If we correspond to the mathematically correct laws of rounding, this becomes 17 Republican seats, 16 Democratic seats, and 1 vacancy. The party that came closest to another seat was actually the Democratic party, just edging out the Libertarians. So the 1996 Senate elections would've been a tie, 17-17.

However, in real life these elections were won by the Republicans 21-13. Without any third parties in the mix, the shift compared to real life is simpler; the four closest Republican victories are given to the Democrats. These are New Hampshire (in real-life, wacky future-Constitution Party member Bob Smith narrowly beat off a challenge from hilariously named former Congressman Dick Swett; here, Swett beats Smith); Oregon (Mark Hatfield retired; in real life, Gordon Smith, who just narrowly lost a special Senate election to Ron Wyden, defeated businessman Tom Bruggere; here, Bruggere wins); Colorado (after Hank Brown's retirement, Wayne Allard beat Tom Strickland in real life; this is reversed), and Virginia (where very popular Republican John Warner faced a surprisingly virulent challenge from future-Governor and Senator Mark Warner, of no relation; here, M. Warner defeats J. Warner).

So, here is the map:

The result is appropriate for changing times; both parties gain and lose large numbers of seats but the party balances barely shift. Democrats picked up 5 seats; Republicans picked up 3. The Libertarians remained constant. The net Democratic gain of 2 seats narrowed the Senate Republican majority to 51 Republicans (led by Trent Lott), 47 Democrats (led by Tom Daschle), and 2 Libertarians (led by James Perry).

[Mehmentum]

Great thread, thanks for posting!

[Vosem]

In 1998, the popular vote was 50% Democratic, 48% Republican, 1% Libertarian, and 1% Reform. If we correspond to the mathematically correct laws of rounding, this becomes 17 Democratic seats, 16 Republican seats, and 1 vacancy. The party that came closest to another seat was the Libertarians, who are therefore given the seat.

This is the first election fought with different incumbents compared to our timeline. In the interests of accurately representing the incumbent advantage, I have added 5% to all "alternate-party incumbents" (Republicans Thomas Hartnett of South Carolina, Bruce Herschensohn of California, Bob Kasten of Wisconsin, and Rod Chandler of Washington; along with Libertarian James Perry of Pennsylvania).

Thus, in reality, Democrats won these Senate elections, 18-16. With the adjustments noted above, Wisconsin switches to the Republicans, making these 17-17. With the adjustments noted above, the best-performing Libertarian is James Perry of Pennsylvania, switching the distribution of seats to 17D-16R-1I; exactly what is needed. However, since there are three pickups (by Democrats in California, South Carolina, and Washington) that did not occur in our timeline, three "alternate incumbents" must be named. I have chosen Evan Mecham in Arizona (who joins the GOP after being elected as an independent), Al Checchi in California, Elliott Close in South Carolina, and Maria Cantwell in Washington.

So, here is the map:

Thus, the ultimate result of the 1998 elections is is six Democratic pickups and three Republican pickups, with Libertarians remaining constant. The result of the net Democratic gain of three seats was a 'hung Senate', with 50 Democrats (led by Tom Daschle), 48 Republicans (led by Trent Lott), and 2 Libertarians (led by James Perry). Since most likely despite these changes the Clinton Administration would've taken place, I judge that Al Gore would've provided the 'fifty-first vote', leading to Democratic takeover of the Senate in 1998 in this world.

In the aftermath of the elections, Republican Tom Coverdell died and was replaced by Democrat Zell Miller. This would likely still happen, formalizing the Democratic majority, to 51 Democrats, 47 Republicans, and 2 Libertarians.

[Goldwater]

Very interesting. I look forward to seeing more!

[H.E. VOLODYMYR ZELENKSYY]

See, this shows why you can't judge people by their abhorrent political views.

[Goldwater]

I hope you haven't forgotten about this...

[Goldwater]

...I see that you have.

[Vosem]

In 2000, the popular vote was 48% Democratic, 48% Republican, 1% Libertarian, and 1% Green. If we correspond to the mathematically correct laws of rounding, this becomes 16 Democratic seats, 15 Republican seats, and 3 vacancies. The parties that come closest to another seat are the Democrats, the Libertarians, and the Greens, in that order.

Thus, in reality, Democrats won these Senate elections 19-15. In this timeline, however, they win them 17-15-1-1. In the interests of accurately representing the incumbent advantage, I have added 5% to all “alternate-party incumbents” (Scott Grainger of Arizona, Harris Wofford of Pennsylvania, and Ann Wynia of Minnesota).

In spite of adding the incumbent advantage, Arizona is not the best-performing Libertarian state, and Scott Grainger loses reelection; instead, Libertarians gain Massachusetts, where Carla Howell defeats Ted Kennedy. However, the best-performing Green Party Senate candidate in 2000 was Vance Hansen of Arizona, who here takes a seat (making him the first-ever Green Senator, as well as the first third-party pickup from a third-party). Taking these adjustments into account, Democrats now lead the Senate elections 18-14-1-1; therefore, the most marginal Democratic Senate seat must be given to the Republicans. That is Washington, where Slade Gorton wins reelection. However, an “alternate pickup” is created in Pennsylvania, which went Democratic ITTL in 1994 but still votes Republican in 2000; I have chosen Mike Fisher as the new Republican Senator. (Notably, “alternate Senator” Ann Wynia, Democrat of Minnesota, is reelected).

So, here is the map:

Thus, the ultimate result of the 2000 Senate elections is four Democratic pickups, three Republican pickups, one Libertarian pickup (from the Democrats), and one Green pickup (from the Libertarians). The net result is the loss of a Republican seat to the Green Party, resulting in a Senate of 50 Democrats, 47 Republicans, 2 Libertarians, and 1 Green. While at the beginning of the Congress a Democrat-Green coalition may be necessary to keep Senate Majority Leader Tom Daschle in power, by the end of the Congress Republican Jim Jeffords of Vermont would've switched to the Democratic Party, creating a Democratic majority of 51-46.

[Vosem]

In 2002, the popular vote was 50% Republican, 46% Democratic, 2% Libertarian, 1% Independent, and 1% Reform. If we correspond to the mathematically correct laws of rounding, this becomes 17 Republicans, 16 Democrats, and 1 Libertarian. There are no vacancies.

In reality, Republicans won these elections 22-12. In this timeline, they win them 17-16-1. However, the 5% “incumbent advantage” added to four alternate Democratic Senators elected in 1996 (Dick Swett of New Hampshire, Tom Bruggere of Oregon, Tom Strickland of Colorado, and Mark Warner of Virginia) is enough to swing Colorado and New Hampshire back to the Democratic column, so the Republican result is already “just” 20-14.

The best-performing Libertarian was Michael Cloud of Massachusetts, who in this timeline is given a Senate seat (from the Democrats); the balance is now 20-13-1. That means that, to bring the results to 17-16-1, the three closest Republican victories must be given to the Democrats. These are Minnesota (where Democrat Walter Mondale returns to the Senate to replace Paul Wellstone), Missouri (where Jean Carnahan is reelected), and Georgia (where Max Cleland is reelected).

This generates two “alternate Republican pickups” in Oregon and Virginia. (In Virginia, it can perhaps be assumed Mark Warner was still elected Governor in 2001, and that it is therefore his replacement running for reelection; I'll say it's Tim Kaine). I have chosen Jack Roberts to be the new Senator from Oregon, and Jim Gilmore to be the new Senator from Virginia.

So, here is the map:

In spite of all the adjustments to Democrats, it is still Republicans who gain in 2002. The ultimate result in 2002 is two Republican pickups, one Democratic pickup, and one Libertarian pickup (from the Democrats). The resulting Senate has 49 Democrats, 47 Republicans, 3 Libertarians, and 1 Green. Since a Republican Dick Cheney, is Vice President, it is the Libertarians, led by James Perry, who control the balance of power. While at first it is thought that the Libertarians may support the Republicans, to whom they are ideologically closer, ultimately the Libertarian caucus (all 3 of whom come from Gore states) decide that their constituents would prefer an alliance with the Democrats, and a Democratic-Libertarian coalition, led by Tom Daschle but greatly constrained by the Libertarian Party, is formed, controlling the Senate with 52 members.

[Mehmentum]

So far, proportional representation seems to be favoring the Democrats.  It makes sense, though.  The Senate as is favors the Republicans because they are very strong in a bunch of low population states.  This system proportional representation cancels that out.

Anyway, its also interesting to see how the third parties turn out.  Unfortunately, it seems unlikely that the Libertarians could get more than 3 Senators, since there only seems to be one or two vacancies each election.

[Cranberry]

This is great! Good to see you picked this up once again!

[Goldwater]

Yay! It's back!