Two-round system in congress?

I put it in this forum but perhaps it should be in After 1900'. I am writing an AH, where the election of the presidents and governors (Mexico) are of 2 rounds, but I would also like the elections to the Senate and the House of Representatives to be decided with an election of 2 rounds similar to the one France does.
I have two proposals but I still have not made up my mind:

Proposal 1:
*First Presidential and Legislative Round: September 1
*Presidential and Legislative Second Round: October 1

Proposal 2:
*First Presidential Round: September 1
*First legislative round: September 15
*Presidential and Legislative Second Round: October 1
*Second legislative round: September 15

Which option is better? Or do you have a proposal?
 
i like the second proposal because a close presidential election bleeds into supporters from either party trying to secure the legislature.

Might I suggest the system start as popular vote first then, if no candidate gets a majority, it goes to the chamber? This is the system that France used during the Second Republic and could be contemporary to a late 1800s reformer like Juarez.
 
I put it in this forum but perhaps it should be in After 1900'. I am writing an AH, where the election of the presidents and governors (Mexico) are of 2 rounds, but I would also like the elections to the Senate and the House of Representatives to be decided with an election of 2 rounds similar to the one France does.
I have two proposals but I still have not made up my mind:

Proposal 1:
*First Presidential and Legislative Round: September 1
*Presidential and Legislative Second Round: October 1

Proposal 2:
*First Presidential Round: September 1
*First legislative round: September 15
*Presidential and Legislative Second Round: October 1
*Second legislative round: September 15

Which option is better? Or do you have a proposal?
Do you mean a preference among these choices? Not really.

But I do think I have a more radical idea that would be easy to implement and much better.

The basic problem first past the post presents is that at the end of the day whoever gets a plurality represents everyone as though everyone voted for them, leaving nearly half (or more!) of the electorate unrepresented at all. A two round system guarantees that the choice between two evils people have is pre-defined, and a plurality equals a majority, so at least as is so often the case with US congress and senate elections, the representative is not of an actual minority! But would it not be better if every vote counts toward positively electing someone, so that a faction of people who can't expect to muster more than say 5 percent national support still do get 5 percent of the seats in the legislature? Having a second round of election as a runoff between the top two candidates is little change from OTL and solves practically nothing. Having a House and Senate with proportional representation is transformative.

I would achieve PR via a remarkably simple form that resembles most closely the German (and many other nation, such as New Zealander) MMP system. But still more simplified.

You divide the nation (or state, or city, or whatever region is being represented) into half the number of districts that the body has. Or a bit less--have an odd number for the whole body--US Congress has been holding at 435 for a century now--and subtracting one to make it even, divide that number in two. So--for the USA to adopt this system today, we would apportion 217 Congressional districts to the 50 states, instead of as at present 435.

Now voters can vote for one person for Congress. A vote for a given candidate is also counted as a vote for their political party. Whoever wins the plurality in their district is given the office of Representative, no run-offs--this is normal in most states, indeed I am not 100 percent sure a state can require a runoff for Congress if no one gets a majority, but even supposing they can, few do. Note though that this mechanism elects less than half the members of Congress!

Now, the votes for candidates are re-considered first as votes for parties--every vote for a candidate who ran as say a Libertarian would count toward the Libertarian party as a whole in the nation. From this, we compute how many of the total 435 seats each party should have proportionally. There are different approaches to doing this--for filling a legislature my favorite by far is to use Hamilton's method of greatest remainders, which is most inclusive of small parties. Other methods tend to transfer a few seats that would go to small parties to beef up the largest party a seat or two more so I have little sympathy for arguments that suggest limiting the number of parties in the body is a good or necessary thing. Anyway, of several systems that exist they all give pretty similar results, so I am going with the one I think is best.

At this point, at least some parties will have gained seats, 217 of them in fact altogether, via having the largest number of votes in their district races. First past the post tends to be hard on small parties, and also to create quite chaotic deviations of representation--some small blocs of voters can achieve disproportionally large representations while others as large or larger are denied any seats at all. This will be the case with the plurality winners.

However, we now take the total number of seats assigned by proportional representation, which is figured on the full size of the legislature, and subtract the number each party already won in the plurality seats, leaving each party a number of additional seats to claim which generally speaking would bring the total strength of each party in the legislature up to the proportional amount. Parties that do very well in gaining district seats would add few more, parties that fail to win any pluralities anywhere will be seated in proportion regardless.

A big issue in all PR systems is, who fills the seats? Shall the party make a list and fill it in published sequence, with the voters having no control? Shall voters vote within party lists to determine which party candidates will win?

In my system it is pretty simple. The candidates of each party are ranked by the total number of votes each one of them got as district candidates. I would make the right of voters to vote outside their district, to vote for a party candidate running in another district in fact, fundamental to the system--out of district votes do not count toward winning district plurality but do count toward party quota and toward the ranking of the candidate nationally. All party candidates already seated by winning a plurality are removed from the list, and the number each party has left to seat are seated according to the number of votes each candidate got. We can have procedures for the rare case that one party gets more than half the total votes, but in the general case, all members seated are candidates who ran in some district or other.

The system presumes that voters will probably stop voting for just one of two leading parties and will instead favor smaller parties that hone their platform closely to what each bloc of voters wants in detail. It is then necessary for the various parties, none of whom control a majority, to bargain among themselves to get legislation passed. Voters who are unhappy with how their representatives did are free to transfer votes to another party.

At the same time, all the advantages of personal accountability of candidates to local constituencies found in district first past the post systems are preserved. Also, MMP systems are said to have a weakness in a way their votes can be gamed which relates to the practice, common to many nations, of voters casting two votes, one for a candidate in a district and another for their party. In my system you can't game it that way because people are voting for just one choice--a candidate, who also represents a party, so voting for a party is indirect. If you support a person you support their party.

You might wonder how to apply such a system to the US Senate--trust me I have thought about how to do it if you wonder!

I also have a different approach to electing a President that is somewhat more complicated to explain but gives some interesting results when OTL election numbers are applied.
 
Let's look at the US Congress election of 2016. Assume everyone votes exactly as they did, but Congress is doubled in size plus one so that 871 Representatives are being elected. I cannot go through all the states and consolidate their OTL districts down so that we get the exact outcomes in 217 districts unfortunately, so the overall outcomes are best indicated with a double sized Congress.

Per Wikipedia, with 49.11 percent of the total vote, the Republicans won 241 or 55.4 percent of the seats in Congress, while with 48.03 percent the Democrats won all of the remaining 194 seats and 44.6 percent of them. No third parties or Independents won any seats whatsoever.

With a Hamilton Greatest Remainder proportional assignment on the other hand, of 871 seats, the totals for each party or Independent candidate would look like this:

Republican 428 seats, 49.14 percent, 8 short of majority control. All but seven Republicans who could run would be seated.
Democractic 419, 48.11 percent, 17 short of majority control. All but sixteen Democrats who could run would be seated.
Libertarian 12 seats, 1.38 percent One from WI, Four from TX, one each from Massachusetts, Kansas and Indiana, three from Arkansas and one from Alaska.
Green Party 4 seats; One from Arizona, two from Texas, one from Illinois.
One each seat for the Constitution, Legal Marijuana Now, Reform and Conservative Party
Four Independents elected to Congress:

Oregon 3 Independent David Walker 74,807
Kansas 1 Independent Alan LaPolice 67,739
Massachusetts 1 Independent Frederick O. Mayock 57,226
California 12 Independent Preston Picus 41,618

I should note that in real life, the two major parties did not run candidates in every district and so would in fact fall short of enough candidates to fill all their seats. I imagine we would have procedures allowing parties to specify how they would fill seats beyond the total they ran, such as allowing each candidate to name an alternate and letting the top vote getting candidates' alternates also be chosen. Different parties might have different procedures, it would be one of the things voters would choose between them on the basis of.
 
This is basically a system proposed elsewhere of Dual-Member proprortional isn't it?

The difference is that I am not allowing people to split their votes between candidate and party. According to discussion on Wikipedia, the power to do that--that each voter votes for a candidate in their district and also a party, separately--leads to some perverse incentives that cause the outcome to deviate from proportional. I propose that it is better if they have a hard chaoice to make of one candidate to also represent their party preference.

In a first past the post system, where third parties are as in the USA practically irrelevant, the candidate for each of the two parties that dominate must be all things to all people--or anyway, seek to capture just over 50 percent of the total vote. In a system where a person can be confident their vote will be counted toward some representation (unless they are such outliers that they can't agree with as much as one fellow citizen in a thousand!) I believe people will be able to count on representatives being very close matches to their own interests and preferences and so the choice of a person and a party will be pretty well parallel. It does not matter whether your preferred local candidate wins the local plurality or not--by supporting them you support that someone similar in values to that candidate and yourself will reach the legislature to represent yourself, and by voting for the local candidate you do all you can to see to it that the person who represents your faction is in fact your own local candidate. If you are a small minority in a district whose voters overwhelmingly disagree with you, it is not reasonable to hope that your candidate will win, but at least you can make sure that your vote assists someone from somewhere who does support your views.

I believe a number of other mixed member proportional systems achieve an approximation of proportionality via multi-member districts--people can vote for 5-9 candidates but have no influence on the outcome at a national level, that being a composite of local outcomes. The system I am proposing is fully national in sweep-as the German system is for instance.

It is an almost universal practice in PR systems to introduce an artificial hurdle above and beyond the simple quota of votes cast to equal one seat won, such as say 5 percent, in various forms--5 percent of the local district vote, 5 percent of national totals, and falling short of that a party is either dropped completely or restricted in the manners in which it can win seats. I see absolutely no good reason for this kind of policy and it is not needed in the Hamilton system in any way. For other PR systems it arguably is necessary to have a threshold but I see no reason for it to be larger than the number of votes needed to earn one seat. In the Hamilton system, a rough natural threshold of approximately half of the quota emerges. But it is not hard and fast. My belief is that proportionality accomplishes several goals, and one of these is to prevent outlier views from being completely excluded. Thirty or forty seats out of 871 are not going to be decisive (unless the other seats are divided between two parties very evenly, and these voters are decisive swing influences!) But they are enough to voice perspectives and views that otherwise might go completely without a hearing, and insofar as the public is watching, if an outlier third party voice makes a very good rational point, it might result in that party gaining support and mainstream ones flouting this consideration losing them. Thus I think a PR system should be inclusive for best results which is why I favor Hamilton's system for filling a legislature. Extra hurdles above and beyond the natural limits it is necessary to clear to get a seat at all do nothing but harm in my view; they discourage voters from confidence their votes are being counted to the full extent possible and entrench established interests, while serving no necessary purpose whatsoever. If the system is at all proportional, any evils that hurdles might be purported to prevent are actually going to be present anyway, as they can be with first past the post systems--such as the current Parliament in Britain for instance, that aren't proportional at all. Being fully inclusive does not add to the evils and offers some useful mitigation of other evils.

So yes as I said in my first post, it could be classified in the same category as MMP, but it is a distinctly variant form of it mainly in that I would have one vote stand for all, and forbid any artificial hurdles. I would also permit people to vote for candidates running in other districts, or as a fallback for a party while refusing to specify a candidate (but not a candidate while refusing to endorse a party--if you like a candidate you should like their party too in this system, since it tends to be honed to a distinct set of interests). These features differ from established MMP as far as I know. I am also not sure that most MMP, or anyway not all of them, specify the party list winners as the biggest vote getters. And I don't think anyone uses Hamilton's method although it is pretty simple to understand and implement and very plainly fair in its operations. Again I attribute that to the effect of conservative politics that seeks to keep back from the extremes of democracy, since other popular methods like d'Hondt are slightly more restrictive in their outcomes.
 
The interesting thing about this little exercise I just did is that even though the vast majority of American voters never dare to vote for a third party since their vote would be "wasted," in reality enough did so anyway to merit some 24 out of 871 seats (and thus more realistically some 12 out of 435) to be given to third parties and even independents who failed to win local pluralities. In the USA, it is quite rare for independents to win. In my system even very small parties have an advantage over true Independents in that the latter have no net cast for votes outside their own district, thus the handful who won seats in my version of the system accomplished a rather extraordinary showing. Another 4 seats could have gone based on voters counted as "independent" who failed to be high up enough on the list for their candidates to win on their own, but had their voters all voted for a single party that all these independents belonged to, they would have some 8-9 Representatives altogether. Of course in real life they would not be one party but several since different candidates stood for quite different things.

In real life, if such a system were implemented in the USA, we would expect an order of magnitude more votes for third parties and better consolidation of small parties along specific interest lines and the dominant two parties would be quite broken in their current roles. They would have to adapt to quite a different system.

But note how already, the tiny number of third party voters of OTL were enough to deny either party an absolute majority!
 
The reason for thresholds is largely to avoid giving excessive power to otherwise tiny or unpopular parties in a position to make or break coalitions or to avoid the problem of parliaments being so fragmented on partisan lines that it is impossible to form a working coalition-i.e. a need to have some kind of consolidation.
 
The reason for thresholds is largely to avoid giving excessive power to otherwise tiny or unpopular parties in a position to make or break coalitions or to avoid the problem of parliaments being so fragmented on partisan lines that it is impossible to form a working coalition-i.e. a need to have some kind of consolidation.
But those are fables. Consider the position of the current British Parliament. They have first past the post voting, but it is an interesting fact that British voters defy the logic that prevails so overwhelmingly on US voters, who are almost entirely deterred from voting for third parties. In Britain they vote for third parties anyway, and more often than not there are substantial numbers of third party MPs in any Parliament. And in the current one, the largest party, the Conservatives, falls short of a majority. Exactly the kind of bargaining that alarms you is exactly what the Tories had to do to form a working coalition. In many other countries--countries rated very high in terms of metrics of successful democratic governance, not in a chaotic meltdown, such as Norway, generations have gone by with perhaps a handful of years in which any single party held a majority, yet these nations manage to get along. Forming a coalition is exactly what they have to do, and that means the leaders of large parties meet with selected smaller parties to hammer out how they are going to compromise to advance each other's agenda in a reasonable way. This is democracy in action. Voters can observe their elected representatives and shift their votes to make their wishes more effectively known if they think they are being fobbed off, but voters also have to learn some realism about the kinds of goals that can be met with easy win win policies, and which they have some homework to do persuading their fellow citizens they ought to take for priorities too.

A small party being able to hold a big one hostage is exactly the situation the Tories face now, and their decision to go with the Ulster conservatives rather than to say reach out to the Liberals (who to be sure have far fewer seats than they ought to) or reach out to say the Scottish Nationalists reflects their own values. If the biggest party finds itself ideologically isolated but a coalition of smaller parties can both find common ground and outvote the biggest, then clearly plurality status must give way to the actual governing majority, which can be formed from many parties as well as one.

In practice, we should get away from the whole idea that one party deserves to govern untrammeled. Policy should be set by individual items of legislation which can command majorities to be voted in, and these can be expected to be passed by different coalitions based on the exact relationship of a new law with each party's priorities.

The OP was asking about how to govern the USA (or apparently optionally also, Mexico). Both the USA and Mexico are not parliamentary but are Presidential republics, in which the executive is elected separately from the legislature, so the question of "forming a government" does not arise at all here. The government is what the President appoints, with the legislature having little direct say in it. Meanwhile though--I mentioned Norway. Or Denmark. Israel too lacks individual parties that can expect to rule with a simple majority, and is run by a prime minister, but they go ahead and have decisive government anyway. The Scandinavians get along fine.

Cherry picking horror stories, it is easy to blame dysfunctional governments on the machinery, but nations are capable of governing themselves well or poorly based on any system, depending on how willing their voters are to elect a government they can live with. If a nation has a proportionally elected parliament but removes the arbitrary hurdles that exclude a half dozen or so fringe parties, the only difference is that the largest two or three parties lose a small percentage of their holdings, and the smaller ones are now present. But they hold just a handful of votes and will never be pivotal in forming a coalition, unless a coalition is extremely marginal to begin with. If they had walled out the large percentage of the voters that these hurdles tend to block, the essential crisis would still exist, whereas in a more usual situation a given spectrum of views can harmonize well enough for large and stable margins formed by a few large and medium parties--and then of course ideologically compatible small parties will join too, why not?

Indeed I think that including the small parties and independents is an excellent way of increasing the options for a legislature to form a working coalition.

You say that a nation is held hostage to a few ideolouges, but the same thing happens in two party first past the post nations, with key factions within the large duopoly parties similarly gaining seemingly unbreakable leverage over the majority. I think it is better if those ideologues maneuver as their own party, and see whether their peculiar viewpoint turns out to be valuable to some coalition or other, than to create great uncertainty about what exactly a voter is voting for when they vote for a candidate of one of two duopoly parties.

The logic that you have brought to the conversation is an argument against having proportional systems at all it seems to me, and not one against allowing the logic of full representation play out fairly. Either we want democratic representation or we don't, is how it seems to me.
 
Germany's 5% threshold is there for the explicit reason of keeping out the neo-Nazis. New Zealand's 5% threshold is there because it modelled its voting system on (West) Germany's.
 
Now voters can vote for one person for Congress. A vote for a given candidate is also counted as a vote for their political party. Whoever wins the plurality in their district is given the office of Representative, no run-offs--this is normal in most states, indeed I am not 100 percent sure a state can require a runoff for Congress if no one gets a majority, but even supposing they can, few do.


Afaik, they can unless Congress expressly forbids it by legislation.

The Elections Clause of the Constitution leaves "the time, place and manner" of Congressional elections to the State Legislature, but empowers Congress to alter such regulations at any time. So Congress has the power to forbid runoffs - but equally the power to mandate them if it so chooses.
 
@Shevek23, are there successful instances of full-proportional legislative elections in a presidential system? As you said yourself, in a presidential system the the government is independant of the legislature. How is the government going to successfully pass laws if the legislature is completely fragmented and the governing party do not have a majority?
 
@Shevek23 To be clear, I am aware of that but I was thinking more of very small microparties proliferating-in the UK, FPTP keeps parties somewhat consolidated so it's a smaller number of major parties and microparties and only two can be realistic majority or coalition leaders. Likewise, I wasn't talking about larger coalition governments so much as the equivalent of say 5-10 parties needed to hold a coalition down-less Lib-Lab, Lib-Con, or the current supply and confidence deal, than say trying to hold a government of "the cons, a UKIP-esque splinter part, some kind of Red Tory party, a few minor parties, and the Libs".
 
@Shevek23 To be clear, I am aware of that but I was thinking more of very small microparties proliferating-in the UK, FPTP keeps parties somewhat consolidated so it's a smaller number of major parties and microparties and only two can be realistic majority or coalition leaders. Likewise, I wasn't talking about larger coalition governments so much as the equivalent of say 5-10 parties needed to hold a coalition down-less Lib-Lab, Lib-Con, or the current supply and confidence deal, than say trying to hold a government of "the cons, a UKIP-esque splinter part, some kind of Red Tory party, a few minor parties, and the Libs".
That was clear enough, but let's look at the actual British Parliament as it is first of all. What happened in the last election? The Conservative party held a majority, with 330 seats, until they themselves, hoping to improve their majority, held a snap election in 2016. The Westminister Parliament is of course first past the post, like the US Congress in that respect. It has 650 Members, but one of these, the Speaker, has a special status that is supposed to be neutral, and votes only according to certain customary rules to break ties, and their election is supposed to be uncontested, so really 325 Members are sufficient to hold a majority. The dynamic then is equivalent to 649 competitive seats.

The Conservatives having a decisive majority in Parliament in 2015 by the way was not reflective of the votes cast by the electorate; in truth they only held just under 37 percent of the votes cast--but with those, gained control of just under 51 percent of all seats in Parliament! This is one of the glaring evils of FPTP, from a point of view of effective accountability of the legislature to popular vote. Or so it seems to me anyway; the suggestion that a lot of little parties would prevent good governance and the suggestion that perhaps it is a good thing for direct democracy to be filtered through lenses that distort the bodies in this fashion are pretty closely related. The implication is that somebody or other knows better than the people or their chosen representatives, that processes that distort that raw public will are somehow desirable. A lot of people do believe that. But let's be clear what we want when we want something to override the will of the people!

In 2016 the Tories did not gain seats, they lost them. When the dust settled they had only 317 seats left in the OTL Parliament. In order to form a government they had no choice but to ask some other party to cooperate with them. With a gap of 8 seats they had two viable choices who to ask, given that other, larger parties, they could approach were more left wing than the alternatives they might prefer. The second party, Labour, had 262 and given the outcomes for other left wing parties had no shot OTL at forming a ruling coalition, nor would the Tories wish to partner with them. The third party was the Scottish National party with 35, but as I understand it they veer left too--in general Labour works better with the various "national" parties like Plaid Cymru. This left them a choice of asking either the Liberal Democrats, with 12 seats, or the Democratic Unionist party of Northern Ireland with 10. They chose the latter, smaller party.

Thus the very nightmare scenario you fear is close to reality despite the strong mechanism of FPTP allegedly "guaranteeing" voters will concentrate in large parties to have a chance to win! The Tory regime depends on keeping a pretty right wing Irish party happy with them.

Now suppose the election of 2016 had for the first time incorporated a Hamilton style maximum inclusion of small parties system such as I like, but people voted in exactly the same way for the same parties as OTL. In that case the Conservatives would still wind up the largest party with 275 MPs--which by the way is 42.31 percent of the total, a substantial gain over the mere 239 which would have been their proportional share in 2015! In fact, the Tories did gain actual support in the snap election, but lost their ill-gotten lead of 91 seats beyond what proportion would entitle them, seeing it whittled down by the semi-chaos of FPTP to merely 42! All of a sudden the need for third party support for the leading party to rule would be drastically multiplied, by a factor of 5!

But this reflects the will of the British electorate. Labour OTL gained 30 seats, up to 262, but proportionally should have had 2 less, 260. Note that in 2015 they had 232 but their proportional share would have been just 198, so ironically this snap election had the effect of calling a lot of voters who had gone "astray" with third parties home to the dominant two; between them their share of votes rose from 62.33 percent to 82.44! And yet it was the 2016 outcome that required the Tories to seek coalition. Remind me again how stepping away from the will of the people brings stability and order?

With PR in 2016, a coalition would also be needed. But the outcomes would be quite different for the upper parties--interestingly less so for many of the minor ones! The minor ones in FPTP have apparently solid consitituencies in certain regions that somehow enable them to exist--indeed the OTL third party, Scottish National, would be downgraded in outcome to 20 versus 35. And further eclipsed by the fourfold surge of the position of the Liberal Democrats, who would rise from 12 seats to 48! The Greens would rise from 1 seat to 11, but Plaid Cymru drops but only by 1, to 3.

Here is the complete list with Hamiltonian outcomes:

Conservative 275--down 42 vs OTL
Labour 260--down 2
Lib Dem 48--up 36
Scottish N 20--down 15
UKIP 12--up from zero OTL
Green 11--up 10
Democratic Unionist 6--down 4
Sinn Fein 5--down 2--meanwhile I am told SF never takes their seats on principle.
Plaid Cymru 3--down 1
Eight of these 9 gained seats in OTL, except UKIP.
Social Dem & Labour (Northern Ireland, caucuses with Labour) and Ulster Unionist party would each hold 2 seats having gained none OTL
A single Northern Irish "Independent Unionist" gained her seat OTL FPTP and would hold that seat here as well
Three other parties--Alliance Party of NI, Yorkshire Party, National Health Action--would each hold a single seat.

I note that this ragtag scattering of small and regional parties is not unusual in the OTL FPTP system; it is quite common for a handful of single MP factions to show up there (unlike the US Congress). So to be clear, in Britain's case you are sounding an alarm about something that actually happens there all the time, or rather would if the true will of the British voter could be expressed in accurate sizing of the leading parties. The major thing that prevents a ragtag group of 5 or six small parties from holding the whip hand is not, let us be clear, that these are excluded from Parliament completely--rather, it is that generally speaking, the big parties get substantially more seats than are properly due them, which tends to guarantee one or the other holds a bogus majority, as was so grossly and egregiously the case in 2015 where the Conservatives, until they foolishly grasped for yet more, could legislate and govern unchecked despite not even managing as much as 40 percent of the vote!

Observe how grossly screwed the Lib Lab party, which by rights should be the third party in the current British system, was by the FPTP dynamics!

Consider how many British voters who cast votes in the FPTP system got no MPs at all for their pains: adding up all the votes cast for parties that failed completely to get any seats whatsoever, we have just over a million out of 32 million. Considering the lopsided manner in which some parties got far more than their share while others get far less, we can more systematically measure how distorted the outcome is with something called the Gallagher Index, which is a least squares measure of a type familiar to science students--square the deviation between the share of votes cast and seats received (as a quota of votes cast), sum them, divide by two and take the square root, and we have a number representative of how far off the seating is from proportional overall, accounting for all factors. For the Parliament of 2016, this GI is 6.49 percent. For the Hamilton apportionment (no system can be perfect, and under Hamilton's system about 185,000 voters still fail to get any PMs, and small deviations from perfect proportionality exist among the others that do get seats) the GI is on the other hand 0.4 percent!

Now look at the dynamic of forming a government facing the Tories--and also Labour! The Conservatives need 50 more members of a coalition, and their preferred choice of OTL is down 4, leaving only 6 toward that goal. Perforce the Tories must turn to parties that actually got a lot more support from voters than their OTL preference. (Remind me why this is a bad thing for democracy?) They cannot form a government without asking the Liberal Democrats to support them. Nor is this quite adequate, they need two more seats from somewhere! Aha, the dread specter of the small party holding the big one hostage emerges! But does it really? If the Lib Dems agree, then the Democratic Unionists are still there, to enable this coalition by a slightly larger margin than OTL to govern. Of course now the Tories must reconcile their policies with the Lib Dems to survive--but that I believe means that they must moderate their conservatism, which is what the British people voted for I think. To be sure there is the headache of simultaneously keeping DU on board as well, but how strictly necessary is that?

Not very! The Ulster Unionist Party has 2 seats (in the PR ATL) and can just fill the gap for an alternative 3 party coalition. There is that single independent Unionist who could cut it down to one seat that could be filled by the single Alliance Party Member. UKIP has 12 members to offer. The Conservative Party thus could seek to draw all of these into its coalition and thus hold a margin such that no single third party withdrawing (except the vital LibLabs) would cost it government.

But now consider the position of Labour. It does not trail the Tories by a large amount. Gaining the support of the Lib Labs, there are a lot of choices available for closing the 27 seat gap that would remain. The Scottish Nationalists would fill most of it, and leave just 7 more seats to be picked up. The Greens alone could more than do it, meanwhile Labour too has many options in that most of the small factional parties remaining veer left. Perhaps they could persuade Sinn Fein to take their seats and help govern? But they don't have to do that--Plaid Cymru, perhaps the Yorkshire Party, I would assume the single National Health member, all seem likely to work well with Labour, and I am told the Irish SD&L Party is really just Labour with an Irish accent.

It boils down then to the LibLabs being the "tail that wags the dog," for that party determines which of the two large ones will govern and can name its own terms. But is this a bad thing? LibLab is presumably a very moderate party, with possible interests in working with either Conservatives or Labour. At this point, it becomes the job of the political leadership of all three parties to craft arrangements appealing to common values that taken together, represent the will of a solid majority of British voters. With prior and broad agreement, the dominant party will modify its platform to appeal to LibLab as well, and together they govern as one.

None of the ragtag small parties are in a position to veto or upset this balance. Each coalition must include a third party, but that third party is expendable! It can be replaced by others, and what I would do, as Tory or Labour leader, is drum up as many as my party and I judge the LibLabs can tolerate, for maximum votes and maximum flexibility, so if one or two throw a fit they can leave without changing the dynamic. Only the LibLabs are in a position to throw it one way or the other singlehanded, and they must appreciate that despite their power in this they remain a minority. If they choose to break a working coalition over an issue that the majority of voters do not support, they are likely to lose some votes while the wronged coalition big partner may gain them. Or both may go down in favor of the opposition big party and other small parties that gain in the face of this spectacle.

The notion that government is held hostage to irresponsible little parties is clearly false in this case.

But what happens if we listen to the siren song of hurdles, and put up additional barriers to achieving membership in Parliament? Note that this would be a deviation from British custom first of all, in that quite small parties are already able to get seats in the current system.

How big a hurdle do we want? Let's start with 1 percent, which is quite low. At a stroke, all parties getting less than the Green Party in 2016 would be out--and by the way, it would be necessary to exempt the Speaker's seat because they generally have very low turnout in their customarily noncompetitive elections. 1,777,000 votes, 130,000 more than under FPTP voters would have their choices tossed away as though they had not voted at all. The effect of this on the Gallagher index, a compound of overrating the performance of the top several parties and dropping the majority of parties from consideration completely, is to raise it from 0.4 percent to 1.9 percent, nearly a factor of five. It is still far below the 6.5 resulting from the irrationalities of the FPTP system to be sure, but even more voters have effectively lost their influence in Parliament. What have we gained for this sacrifice?

The Tories have lost their third party, DU, as supporters, but they gain more of their own members--now up 11 to 286, while their vital coalition partner, the LibLabs has gained two as well, so a conservative-moderate coalition is simpler and stronger. Meanwhile Labour gains 10 seats, so that with LibLab their alternative coalition also has a margin of 5 without needing other members. The Scottish Nationals and Greens remain to pad a Labourite government in large coalition with 31 more seats, whereas the Tories have available to them just UKIP for 12 more seats. The left leaning total coalition would be larger by 4 seats.

The German hurdle of 5 percent is being bandied about as reasonable. What would it do to the 2016 election?

Well, only three parties--Tory, Labour, and LibLab, cleared that bar in 2016. Clearly the coalition dynamic is simplified down to the ultimate. The fundamental dynamic is still the same though--someone must court LibLab. Meanwhile, the number of voters cast aside as insufficiently compromising of their values to qualify for effective control rises by nearly 2.1 million, tripling the number excluded in FPTP OTL, to 3.27 million or ten percent of the whole electorate. Accordingly the Gallagher Index rises to 5.5 percent. We have a somewhat less awful deviation from voter desires than with FPTP, but at the cost of excluding far more of them.

Basically a hurdle as large as 5 percent comes close to replicating the overall irrationality of FPTP. A smaller hurdle does less harm, but harm is done for no good reason. It is simply not true that the small parties we get with maximum inclusion derail coalition politics. A very small party is critical only if the nation is finely on a balance, and in that case it is probably for the best if a breakdown in coalition forces a new election or a period of weak government, since this focuses the electorate on the issues at hand. Papering over the issues by steamrollering out 20 times more effective votes than is strictly necessary strikes me as a cure worse than the disease, where the disease is precisely the failure of publics to come to workable agreement.

I would also note that the optics of the outcome is skewed toward the OTL apparent conservative dominance, while with more inclusive outcomes, the fact that British voters on the whole skewed a lot less rightward is more apparent. A Tory-LibLab coalition commands more members than a Labour-LibLab coalition, while sorting the parties left and right and combining all of them with the lower hurdle or the no-hurdle Hamilton method gives the leftist coalition the edge.

What would the options have been in 2015 by the way, when OTL the Tories held a thin but comfortable victory margin in their own right? Under PR they and Labour both would suffer considerably! As it happened in that election, the separatist from EU UKIP party enjoyed a tremendous level of success in polling, but got just one seat in Parliament by FPTP. Proportionally though they would have gotten 82 seats by Hamilton's Method. Without including UKIP in their coalition the Conservative Party would have been barred from forming any majority coalition. With it they would need only a handful more they could hope to get from many sources. However, Labour, assuming that party would never agree to combine with UKIP, has no viable path to majority whatsoever. The LibLabs, getting 51 seats, and Scottish National with 31, and Greens with 25 would not be enough to close the gap with the Tories and no amount of scrambling for the smallest parties on any conditions could get them victory. The upshot is that as government without the LibLabs would have been impossible after the 2016 election, so UKIP would control whether the UK could have a viable government or not in 2015, only more strongly, and they would have no option but to do so with the Tories.

Would raising the hurdles fix this? No, it would not. A drastic 5 percent hurdle would include only the top 4 parties, skewing the choices rightward as in 2016, limiting them in order to Tory, Labour, UKIP and LibLab. Labour would still have no prospect of governing without dealing with UKIP. The Tories would indeed pick up a new option, which would be to govern with a slim majority in coalition with LibLab, leaving both the middle parties out.

So in this case, having high hurdles would indeed give the Conservatives an option other than aligning with a party that some people (me included) considers an unsavory far right wing as well as lunatic bunch. But the reason UKIP would stand so high in 2015 is a that a whole lot of Britons voted for it. Without that brand of right wing lunacy no single majority bloc would exist, and the only one possible would be right wing. Meanwhile it is not clear to me that the Tories would actually prefer to favor the respectable LibLabs if their basic agenda is more strongly supported by the radical UKIP. Thus the value of the high hurdle is dubious at best. Voters in 2015 voted a range of options that veered hard right, much as they did in America the next year. Again the only reason the Tories have better options with the hurdle is that they themselves would gain 33 seats at the cost of voters for smaller parties who presumably favored less extreme options--it is this disproportionate padding of a mainstream party that gives it the freedom to choose which ideological direction it might veer in and all the recent history of that party suggests to me they would choose the harder right option--for they could have built a stronger coalition with LibLab in 2016 OTL that with DU, but they chose the latter.

It is my belief that if Britons approached each election knowing they could vote for the candidate and party they most strongly believed in, without fear of it being cast aside by the rough and unpredictable FPTP nor by an arbitrary hurdle, they would pick a faction and, assuming that the leadership and MPs continued to pay close attention to their deep concerns, stick with it. As issues change over time, some parties would lose support while others gain, parties will die out while new ones form. But by and large the voters will pick stable factions, and the factions will form stable alliances. We would not have seen the rise of a party like UKIP in the first place; rather certain parties already in place would hae gained some strength and the showdown on the issue of separation from the EU and whatever else motivated UKIP voters of OTL would have been aired out earlier. Other conservatives along with people to the left who had their own issues with EU would have to show their hands, showing where they stand, earlier, and be judged well or badly by their various constituencies.
 
Now voters can vote for one person for Congress. A vote for a given candidate is also counted as a vote for their political party. Whoever wins the plurality in their district is given the office of Representative, no run-offs--this is normal in most states

Of course, if open Primaries were used rather than party ones, the point would be moot, as every November election would be a "run-off" between those who placed highest in the Primary contest.
 
Of course, if open Primaries were used rather than party ones, the point would be moot, as every November election would be a "run-off" between those who placed highest in the Primary contest.
Perhaps you should clarify what you mean by "Open Primaries?" (For the record I am against both of the alternatives below):

1) It can refer to having multiple registered parties in the race, along with options for independents, for the official November election in which the winners for office are finally chosen, but each party has a primary in which a registered state voter can vote in any one of them however they choose, regardless of their registered affiliation if any. That is, there is a Democratic primary, a Republican primary, a Libertarian primary, etc for as many parties as the state recognizes. But while voters are registered in one of these or none as nonaffiliated or "nonpartisan," any voter can cast a vote for any single primary candidate for any office. Registered Republicans may forgo their opportunity to influence who becomes the Republican candidate by instead meddling in the Democratic primary, essentially trolling it since they have no intention of voting for any Democrat in November even if the one they vote for wins. This kind of thing is popular in Dixie and has been used to destroy many a strong Democrat in the South, by primarying them out with a bunch of votes that will never vote for any Democrat when the chips are down. This is what is typically called an "open primary."

But it can't be what you are talking about because in this case, there are going to be more than two candidates to vote for in November just the same, unless the state restricts legal recognition to just two parties, or makes some unheard of rule that the two parties with the nominees who got the highest and second highest number of votes are the only two that people are allowed to vote on, which would be a new twist in America I think, though perhaps it has been tried historically.

This is a step toward:

2) Something entirely different seems to be what you are talking about. First adopted I believe in Louisiana, where single party rule has remained the norm despite suddenly switching from one party (Democratic, in a mainly Jim Crow way with weird nuances of populism that sometimes even involved nuanced and ambiguous outreach to the African Americans who make IIRC an actual majority, or anyway are a very very large minority, to Republican in the course of a decade or so) the rationale is that since the real election on the presumption that one party will win all the November races is the primary, the rules should be rewritten to guarantee a real contest in November, unless the primary itself is decisive. Instead of partisan primaries, all candidates for a given office must run in one combined primary where all voters vote on the same list of all comers. That way if the November race is really going to boil down to Republican versus Republican, the primary is seen as mainly the opportunity to identify two and only two candidates that voters must choose between in November, giving the entire electorate the opportunity to decide which of them is the lesser evil. This guarantees that one will win with a clear majority mandate for in a two way race plurality equals majority. It guarantees that most voters must compromise in advance and decide which candidate is closer to their values, granting that both might be mere microns apart on most or even all issues, and light years away from many voters' value sets. Everyone gets a degree of influence that is in principle unpredictable, as it depends on perceptions of where two candidates who emerge from the primary each stand, and how relevant the differences between them are. The only answer for a voter who hates both is just to stay home, or show up but skip that vote. Here is a case where a "None of the Above" option would get my grudging approval, provided it had some teeth, so that voters voting for that option above a certain threshold trigger a different election process, one that bans both these candidates from appearing perhaps.

In truth there is no guarantee there will be a November election for every office at all, for if one candidate gets over 50 percent in the "primary" I believe they are just declared the winner and will take office with the ones who must go on to win in November. This is how they do it in France I believe; second round only goes if the plurality winner in the first round falls short of 50 percent. French voters are accustomed to vote for many parties in the first round and pretty confident that they won't be cut out of the second round loop by a rival candidate winning a true majority the first round, but some of them are anyway every election. Just a few. But if we adopted such a system in America comprehensively, I believe the whole thing would end right there in the "primary" with a majority of candidates getting majorities right off, and the minority who did not support them does not get a second shot to elevate the number two candidate. It all depends on how the rules are written; a November race can always be guaranteed by simply saying that the top candidate and the runner up get to face off in November, even if the top candidate has 89 percent and the second runner up only 5 percent. Presumably the first one will win easily in November but in principle we don't know until the vote is held.

But I don't think states that adopt this, unflatteringly called a "Jungle Primary," do that; they tend to be all about decisiveness and less about concerns about the ability of each voter to achieve the closest thing they can to a fair share of power. It is pretty decisive that there is a chance the whole thing is settled months before November.

Much to my amazement and dismay, California too has adopted the "Jungle Primary." I disapprove heartily of course, but there it is. Presumably this was some kind of response to Democrats gaining the upper hand and winning the majority of elections. I think it is a bloody nightmare myself.

Assuming that ignoring the siren song of decisiveness at the first opportunity, even majority winning primary candidates must face off one more time against their front-running runner-up in November, that fits your description a lot better. But it is a terrible system. You would do better to suggest that two candidates among the eligible citizenry are chosen by lottery, with each one chosen having the right to refuse, so that the state just keeps rolling the dice until two citizens, random except insofar as the factor of being willing to take the risk of a campaign from which one will emerge with real power while the other is dropped into obscurity again, are up for election in November. Hell, I think nomination by lot like that is infinitely better than something like the Jungle Primary you seem to be advocating.
 
If there are several Democrats and seceral Republicans contesting the Primary, it's surely difficult for any one eof them to get 89% - or even 50%. So no real problem.
 
Instant runoff voting would be great. Everyone ranks the candidates in order of preference. The last place candidate is dropped and those ballots are redistributed to the second choice. Rinse and repeat until you have a winner.

Libertarian or Green or Constitution voters wouldn't need to vote the lesser of two evils. They can still rank the Democrat ahead of the Republican or vice versa.
 
Instant runoff voting would be great. Everyone ranks the candidates in order of preference. The last place candidate is dropped and those ballots are redistributed to the second choice. Rinse and repeat until you have a winner.

Libertarian or Green or Constitution voters wouldn't need to vote the lesser of two evils. They can still rank the Democrat ahead of the Republican or vice versa.

This would also make primaries unnecessary. All the Republican and Democratic aspirants could just go on the November ballot, and let the voters sort it out. In general, one assumes, Republicans would give their second etc preferences to fellow-Republicans other than their first choice, and Democrats similarly to fellow Dems.

The parties might still hold Conventions, but these would only be to draw up platforms, and maybe give recommendations (which the voters could ignore if they chose) as to which candidates they should support.
 
Top