A proportionality procedure I think I just invented

In discussion of the merits and demerits of various methods of achieving integer proportional mapping of a spectrum of votes (these, single choice votes as in FPTP or classic party vote PR–MMP using both in two votes) with Jefferson’s method, much promoted on this site especially, versus Hamilton’s, by far my favorite for the purpose of forming multi-member bodies as it is most inclusive, I stumbled upon what as far as I know is a new procedure.

For a capsule description:

  1. Dividing the total of all ballots cast by the number of seats to be assigned, the total of ballots cast for each party in the election shall be divided by that quota, and the set of all parties and independent candidates (henceforth referred to simply as parties, independents being regarded as parties of one candidate) that thus would obtain at least one quota seat determined and totaled; the remainder of votes cast for parties shall also be noted.

  2. These two subtotals shall be divided by the quota and the portion with greatest remainder beyond integer quota seats shall receive the remaining seat.

  3. Using the Huntington-Hill method of proportional seat allocation, all parties included in the quota set shall be apportioned seats out of the share assigned in step 2 to the quota parties.

  4. The subquota parties shall be ranked in accordance to votes received, and in default of actions described below in step 5, the top ranked ones totaling to the share of seats for subquota parties determined in step 2 shall also be elected.

  5. However, the multiple of apportioned seats for the quota parties shall be subtracted from the total votes each party received, and any party having a positive balance remaining (and thus underrepresented) shall be able to participate in asset voting with the subquota parties, able to transfer any part of this excess to any party of their choice that agrees to accept them, as the subquota parties shall also be able to transfer their received ballots in the same manner. These transfers will reorder the list, and any consolidation that exceeds the quota shall upon final confirmation entitle that party to a seat, which shall be subtracted from the available seats allocated to the subquota share of seats. Any seats remaining shall go to the parties with positive outstanding shares in the largest number per the criteria in step 4. This process shall occur in a stipulated time frame and upon the end of the time frame any tentative assignments of votes not rescinded shall be locked in place and any votes not consolidated in parties winning seats shall be deemed suspended, subject to procedures in step 6.

  6. Majority guarantee and ngoing operation of asset votes cast by parties not represented directly:
    a) Upon completion of step 5 the body composition shall be set and no seats shall be added or lost to any party by the processes here, but;
    b) All votes toward any party winning one or more seat shall be divided by the number of seats that party wins and this shall be divided by the quota, for an index of actual vote based representativeness of that party’s seats, and the entire set of all votes effectively assigned to seated parties shall be divided by total seat numbers, and multiplied by the seat majority of the body to determine a majority quota for the body.
    c) At any time during the period after the election closes until the body finally adjourns for the last time prior to the next election, any party that assigned asset votes to any other party without itself holding any seats when the process of step 5 is complete can, by default by the individual candidate not seated of such a party, or whatever party authority this candidate may have delegated this power to, withdraw them, or parties that did not assign asset votes toward any party holding one or more seat can by mutual agreement with the party add them. These subtractions and additions change the majority vote quota for the body and the indices of any seated parties affected for all matters subsequent to the change being recorded.
    d) By default all body business is conducted on the basis of one member having one unitary vote, but any majority claimed for any purpose including supermajorities of specified scale must, in the members counted toward that majority, actually exceed the majority threshold as adjusted by the members’ summed indices to be unchallengable. The body shall in its administration determine whether any votes that were close enough to the appropriate minimum threshold to be at risk of falling below that threshold and note those that might fail to meet this target. If this should happen, any member in the member count majority can protest the passage and a reasonable time limit in context shall be set for the member count majority to in some fashion raise the count above the actual majority threshold relative to represented public popular vote in the body. These shortfalls shall not be created retroactively by changes in part c that happen after the vote in question, but such changes can if they result in raising the recorded share of members above the threshold will settle the issue. Upon expiration of the deadline set failure to meet the majority criterion shall be deemed a failure of the original vote.

As laid out, these procedures are for a classic proportionally elected body such as say the legislature of the kingdom of Denmark. But they can with suitable adjustment apply more broadly, as to say my proposal to achieve PR in a body primarily elected by FPTP in single member districts via evaluation of the need for top off seats to counter overhangs. More on that, with mention of the problems of generalizing the concept of proportionality beyond single choice simple voting, in the next post.

To apply this to the general concept of Overhang Resolution, we would first perform step one and also in addition note how many quota seats any parties that won district races by any means would be entitled to proportionally.

To generalize to cardinal district voting methods, we need to develop acceptable criteria for something closely enough analogous to proportionality to be useful and meaningful but as I have suggested we can always add to an approval or score district context an additional “Prime” vote, ideally a single choice for a single party, or perhaps for ballots where the voter does not cooperate in this, inform voters that their apparent highest priority choices (top scores in score votes, all approval votes, for ordinal ranking systems the first ranked, etc) shall be interpreted as prime, and if there are ties such as two or more parties rated at the highest level the voter chose, they are deemed indifferent between those choices and their vote split fractionally and evenly among them; it may or may not prove workable to use an approval or score ballot to so apportion all ballots fractionally in proportion, or some other process of parsing their votes might come suitably close to the proportionality concept. By default, I assume either races are single choice as in FPTP single seat per district races, or that a method of indicating Prime party/candidate choice is provided and followed.

Assuming a suitable definition of proportionality or quasi-proportionality is accepted, the partition of seats between quota and non-quota parties as in step 2 should be possible, and it should be possible to arrive at the specific quotas (not complete proportional shares, but that portion of them corresponding to full quota shares) each party winning district races in any number are entitled to. For overhang compensation and top off, we subtract the quota share with the standard body size from any larger number of district seats won, and double that difference, then reiterate a new proportional quota determination at the higher seat count, and thence proceed on through step 5 to determine party shares. Then we can, using a method of ranking candidates who did not win election by district race victory that is appropriate to the district race method of voting, appoint any positive difference between the expanded proportional share computed on the larger body and the district seats won, by taking the highest merit (in plurality FPTP, simply highest vote count won) candidates. If the total proportional share of a party exceeds the total number of candidates that party ran, we can adopt various methods of augmenting the delegation, I recommend cloning the top merit individual candidates (I have a topic on Deputy Candidates that gives a very specific and generally applicable to all kinds of parties method for this).

Note that my deep concern about ironing out all overhangs, or somehow determining which ones must be ironed out, in order to prevent what I called “spurious majority,” might be dispensed with after a single pass if we have the Step 6 checks on numerical majorities of seat count corresponding to actual popular vote majorities; a party with a substantial overhang outstanding will automatically have its members’ majority-share index reduced, so coalitions including them have to come up with extra votes to avoid narrow nominal majorities from being vetoed by actual majorities. Note that an example of a persistent overhang is in recent British Parliaments even after one pass, with Scottish National and others persisting with something like 50 percent more seats even after leveling up. Note we can readily calculate the degree of total expansion of Parliament multiplied beyond 650 seats that would nail down SNP, and some smaller persistent overhangs by other smaller parties, but the magnitude is tremendous–which is one reason I liked the option of simply doubling the house size. (Even that is not absolutely guaranteed to prevent party overhangs if we have districts greatly fragmented so that the plurality winning a seat there is a small minority, say a quarter or less the total vote, so it is a good thing to have a method of neutralizing the harm done without having to expand the body infinitely!) Even without great fragmentation of party votes in hotly contested districts, persistent overhang happens with smaller parties because the top off seat shares go proportionally and so a party causing many overhang resolution top off seats might get quite few of them itself–also if the proportional-guaranteed election enables voters from all over the system to support localized parties and candidates from far outside their region, we will presumably see less of this, as Scots or people who just like SNP’s program all over the UK can register additional support beyond that available within the districts where SNP actually runs candidates).

Now to see if I can briefly outline the reasoning behind this novel (to me anyway!) process, referring to the numbered list in the original post.

It is most likely a constitutional matter as to whether a nominal majority of legislators can vote and not win (unless perhaps if you have a legislative rule implementing this, similar to the way the filibuster works, but you might need a supermajority requirement for overturning the rule to ensure the nominal majority doesn’t get rid of it immediately.)

Step one I think is pretty transparent and clear, I hope. Looking then at step 2, first of all why stop short of full application of some standard method of proportionality–as I think is the case, these all are on a spectrum between Jefferson’s method which concentrates seats the most on the biggest parties, and Hamilton’s which is the most inclusive of as many parties as possible given the fixed, more or less, number of seats available?

As noted in the OP, we were discussing the reasons that Hamilton’s rule is normally disregarded as a suitable PR seat assignment method, in particular the somewhat chaotic jittering of exact numbers of seats assigned that tends to favor parties splitting just to capture extra votes on the “seams” as it were. I was not happy abandoning Hamilton at all. And in fact Step 2 is an ultra simple application of Hamilton’s rule–we could for instance use Jefferson instead here, and the upshot of that would be to have more seats for the quota parties, and fewer left over for the Asset Vote process, which note is ultimately also resolved by Hamilton’s Rule.

I believe that by restricting application of Hamilton’s Rule to these two separate and specialized steps, the nonmonoticity “Alabama Paradox” aspect of HR is neatly sidestepped. We minimize the sequestering of available seats to the quota parties, who certainly should have the lion’s share of them, to the lower end of the range of acceptably proportional outcomes. But there can be no subdivision of viable parties entitled to multiple seats to manipulate or to chaotically throw off overall balances; the division in 2 is purely and only into two portions, one split, and it is determined with essentially all the parties that will routinely dominate body business lumped into one giant camp, in order to discern how many seats reasonably should be available for maximum inclusiveness and maximum body “footprint” on the total vote for the non quota seats.

For step 3, note that while the quota seats alone are the lion’s share of quota party delegations, they don’t proportionally comprise all of them; in fact given realistic election patterns (I have also looked at the British Parliament where while it is FPTP, there is still tremendous party diversity, and at some proportional elections that are highly fragmented like that of Brazil’s lower legislative house) the quota parties ought, per the numbers found in step 2, to have between them the vast majority of all seats. I mean to offer examples, but for a quick capsule, the US House race of 2012 would produce just 4 quota parties on the basis of 435 seats, and gain just one more, an independent from California named Bloomfield, if we slightly more than doubled the House to 871 Members! Those 4 actual parties are Democratic and Republican of course, and Libertarian and Green. With 871 members, step 2 would reserve 860 of the seats, leaving just 11 for the fractional quota competition. But 11 seats being available clearly means a great many voters–nearly 1.5 million out of 120 million, or well over 1 percent–failed to consolidate on quota-viable candidates. Customarily we simply shrug them off as not deserving representation, but my concern is to maximize positive representation.

More on why Huntington-Hill after I check replies to what is posted so far!

I think you are conflating three related but distinct C-words here–“constitutional”=/=“customary”=/=“conventional.” OK maybe the last two are almost synonyms, but constitutional definitely is not.

I’m using the lower case here to acknowledge not all nations have a written, definite Constitution as the USA does, and as its states do. Prior to the Revolution Americans were familiar with Britain’s unwritten constitution, and the word definitely did not mean a specific document that specifies a framework, but rather an evolved set of practices and values that are obeyed not because some authority says one must, but because common practice demonstrated over time why it was wise to stick to certain broadly understood principles and protocols.

So certainly once a republican body (if you will allow me to stretch that term to cover say Commonwealth parliaments–in conventional prerevolutionary British empire discourse, a key element of the British “constitution” was the idea that three different principles of government as categorized by Aristotle, namely monarchial, aristocratic and democratic were embodied in both the formal evolved structure of British government and in British society in general—the House of Commons, long before the franchise was systematically extended to all subjects in the UK, was deemed the governmental embodiment of the democratic dimension of the British governmental constitution) has formed in all modern era history known to me, what has mattered has simply been the numbers of members, deemed equal to one another, in the body–half of them plus one, barring special matters relegated to supermajority, can do anything within the parameters of any checks and balances developed by custom/common law, or by explicit Constitutional stipulation in our US written version.

Most certainly it has not then been customary, or conventional, to introduce special checks, except of course for the very long established practice of supermajority rules.

But I don’t think you will find a word in the US Constitution barring such an innovation. As an untested innovation dashed off by my brain just in the past few days, probably it needs serious critical scrutiny and might be demonstrated to be a very poor idea aside from whether people will even give it a fair hearing or not.

Then again, I suspect exactly such pettifogging considerations are embodied in many corporate charters! There it would be money–one dollar one vote–rather than people which is the fixed, quantified metric. I’m just guessing here that if I were intimately familiar with the operations of boards of directors of companies, it would be quite common that not all board members vote equally, that one or two guys who happen between them to own over half the stock pretty much dictate. This can be consistent with operations that normally go forward as though the Board were composed of equal members with equal votes–note that I didn’t suggest that every vote by a false majority (Elizabeth May of the Canadian Green party is the one I owe that pithier term than “spurious” to, from someone’s recent post here on Canadian advocacy events) would automatically be invalidated, though that actually might be a better way to proceed, and indeed for someone to be say Speaker of the House they definitely ought to be required to be elected by a non-false majority! I said that some member can opt to move to object and that the nominal majority has a limited time span to bring someone on board to augment the vote beyond doubt. Just as the majority owners on a Board of Directors might be content to let the usual parliamentary procedures roll on as though they were just a couple of the guys, and only clear their throats to indicate something specific either had better happen or had better not when it is a matter of importance to them.

As a practical matter, there are two considerations that might sidestep the whole section 6 gadgetry I dreamed up:

  1. Usually, if a faction or coalition can achieve body majority, it can press on to get some substantial margin beyond just barely doing it, and that margin, even of discounted members, can be made large enough to guarantee the majority is in fact real.
  2. I still have not got a bead on what people here mean by a proportional body, but for me it is open and inclusive and surely will become pretty diverse, and the changed political context will moot a lot of stuff we take for granted right now. I have gotten concerned about very fine balances because in our current US situation, the two-party system often does turn on very fine balances. In 2012 for instance, a properly proportional US House of Representatives would have turned on just one or two crucial members, the two parties being quite neck and neck, and both of them in the minority by PV–an ugly fact obscured by all sorts of consensus-manufacturing machinery. I still haven’t gotten to bringing out the smoking gun of the Nevada state Assembly false majorities in 2012 and 2016 sessions here; I did manage to toss out the much more infamous (though far less grossly out of balance!) case of Michigan’s state legislature.

Since you toss around the word “constitutional” now is as good a time as any to mention that this general paradigm of a closely proportional body adjusted by top-off but drawing from vanilla FPTP plurality districts cannot apply directly to the US House, and I can point to the specific Constitutional language that poses a possibly insuperable barrier.

But I actually think people often assume the Constitution says things it really doesn’t.

As it happens, there is another approach toward, not reaching but toward, a proportional US House I think would in fact be quite Constitutional and I’m meaning to get to it one of these days when I have managed to get more or less on top of other plates I have set spinning and responding to other people’s topics reciprocally. As it also happens, this not quite properly proportional approach would be quite vulnerable to factors tipping delicate balances and I therefore get pretty excited about some sweeping, go to the root of the problem approach to doing away with the problem.

For another gimmie that I am surprised no one has pounced on yet, part of these alternate versions of top off, either flexibly but unpredictably in response to specific mismatches between plurality outcomes and measurable overall proportionality, versus a fixed body size much larger than the number of districts, have floating membership; even with say 870 House members, exactly twice 435, under these proposals there would be no guarantee each district had exactly two members; some would have more than that, others be stuck with just one–statistically it works out to something like 70 percent of all districts in a body where the make-up members are recruited from district also-rans, where the total body is twice the size of the number of districts, would have exactly two members; the majority of the other 30 percent is single member districts while a lower count of districts (but much higher percentage of all members, obviously) have three or more.

Would obstructionists have a field day with such blatant disregard of the idea that each district should have equal representation to be fair? Probably–I’d counterargue the body exists to represent people, all the people of the entity it legislates for, not some count of districts, and that if each person who votes has equal influence on the net outcome, it is not their problem or concern if their district has just one but another one has 3 or 4; as long as this imbalance responds to greater polarization of a district and is not some kind of fixed, standing thing.

I do appreciate these challenges, I think you can see my thinking does evolve under these criticisms, and I am grateful for them.

The thing about consensus-biased PR methods is that, for the most part, they’re only as consensus-biased as voters want them to be. It’s only when you’re unsure whether your party can actually win a seat that you have to give some points to your less-preferred candidates. In addition, if you think your party can win 1.8 seats, you have an incentive to try to score other parties to some extent to ensure that your ballot carries weight for the remaining 0.2 seats; even here, you can support anything from a slightly less-preferred but larger party to the most compromising party in the race. So overall, they only elect a lot of compromisers when voters want that to be the case.

It’s hard to strenuously analyze your FPTP-based PR proposals when the only major voting reform organizations with the power to push for PR (the Center for Election Science, the Equal Vote Coalition, and FairVote) would never push for them, because they’re FPTP-based, and might involve Asset negotiations. However, they all look like they’d work well, and I am influenced by the ideas in them.

Yeah, it is hilarious to me how the stock canned answers I get when they can be arsed to reply at all (CES did, and so here I am) patronize me with “oh, your system is surely too complicated to implement and understand,” when I found the whole thing on people casting exactly the same kind of votes, on the same kinds of ballots, continuing to elect the same FPTP members in the same districts. It is the simplest approach of all. It simply involves counting the same votes twice! No party lists, no ranked choices. Just vote your single choice in your vote for the one you want, and know that if some decently large, by which I mean one in a hundred or so, of your fellow state residents or US citizens agrees pretty well with you, even an outlier party is on its way to the Capitol with perhaps just a handful–but that handful can be very critical in some circumstances. And the really gross contradictions of basic democratic principles are automatically countered

It seems plain that statement either refers to wanting to support quite a tiny splinter party indeed, or alternatively thinking in terms of very small levels of aggregation, like combining two districts into one with no further scope for integration and calling that “proportional.” Mind exercises I have done come much closer to proportional than FPTP does, but the main thing is to get all the people voting all together in one big batch. Figuring out how to do that while at the same time leaving any advantages to local reasonable sized districts intact felt pretty good to me.

So actually, again remembering that we can’t actually apply the system as described to the US House for actual Constitutional reasons, actually expecting 1.8 members to be elected to the House is pretty much the boat Green voters would be in–assuming no one flocked to greatly expand the third parties with their new freedom to do so that is! The thing is, if I have decided to back an outlier party, it is presumably because I think they are better than the mainstream ones. Thus if I compromised to vote for the less offensive mainstream party, I would not be getting the policies I want anyway. So wouldn’t my reasons for casting two approval choices, one for say the Greens and another for the local Democrat, be pretty much exactly the same thing as when I see a Green on the ballot (in California, before Top Two, which was introduced there after I left–no Greens on Nevada ballots! I literally had zero choices to the left of Hillary Clinton in '16 even if I had dared break duopoly ranks in a close battleground state like Nevada anyway) but cast my single choice for the Democrats? It is well to say that if one studies the election statistics afterward, wow look, a third or more of my fellow voters agreed with me instead of half a percent! But we remain footnotes.

Whereas my itty bitty two percent vote multiplies by a hundred million or so federal voters into some real seats held by some real people, when all are counted together at the center. One percent of 435 seats in Congress is 4 seats–that is the same number as “The Squad.”

This is where Score would be significantly better than Approval, because you can give 5/5 to the Green and 1/5 to the Democrat. Even the ability to indicate just half an approval for the Democrat would probably do the job. But suppose there was a “Green+” party (Greens but slightly more compromising, and thus acceptable to more voters.) Such a party likely has a better chance of winning 2 seats than just the pure Greens, so a voter could approve only Green and Green+ and still mostly guarantee proportional representation.

Do these party-list FPTP PR methods require centralized counting?

(Something something 20 characters.)

1 Like

Thank you very much!

@AssetVotingAdvocacy, all systems require some centralization. In the sense that say RCV does, needing one central count because the count can’t happen properly at all without all the ballots, no. But say we have a district that covers a whole bunch of counties; no final determination can be made without someone collating the returns from all of them.

It can depend on how we determine the district results too. If we wanted score voting to determine that, we would not require more centralization than needed for traditional US single choice ballots, but if IRV were used, it would require as much as that requires.

The determination of proportionality does require centralization in the sense that some single office must receive summarized data from all precincts in order to know the global outcome. Say ten percent of all precincts were not reporting, that throws doubt on the whole balance. What if it turns out those precincts were all lopsidedly voting one way, out of the general pattern of the rest of them? There is no way to know that until the returns are all in.

However, the partitioning between quota and nonquota parties suggests a method for the body to proceed on an interim basis I think. The population of the districts not yet reporting is known, as a proportion of the whole system; if 90 percent of all districts have returned complete results, we could compute a partial outcome and conservatively determine minimum victories for all parties, leaving some races up in the air while settling others.

While you talk about applying your system to Congress, the state level might be a better place to start. Here are some reasons why:

  1. One benefit of your system is that parties are incentivized to compete in places where they are likely to lose, since they can’t win any votes where they don’t run candidates. A significant problem with state level elections, far more so than in Congress, is the large portion of state legislators who run unopposed. This includes Nevada, where nearly a third (13/42) of the state Assembly ran unopposed. Some level of electoral competition is a pretty basic requirement for democracy.
  2. While I can understand why your system doubles the size of the US House, (it would allow for finer proportionality without sacrificing local representation), making a legislature that large can make deliberation inefficient. An 870 seat House would be larger than any single House in the world except China’s National People’s Congress, which doesn’t do a whole lot of actual governing. (The much smaller Central Committee of the Communist Party actually makes the decisions.) In contrast, Nevada is considering increasing the size of its legislature anyway. Granted, it’s to help rural communities avoid having to be mashed together, a consequence of them losing seats to Vegas, but as the Assembly is precisely double the size of the Senate, your system could work into this, for example by increasing the size of the Senate, then using the Senate districts with your system to elect the Assembly.

Running a PR system for Congress is a bad idea unless all states join in at once; a gerrymandered state can send more of its majority’s preferred candidates to the House than a PR-using state. It’s like the quandary of winner-take-all versus proportional allocation in the Electoral College.

1 Like

I use the US House of Representatives as a heuristic example of what applying these principles would look like concretely, bearing in mind that the point of the reform is to enable voters to change their behavior and change the basic dynamic. But everything I have said is meant to be general, which is why I started with the example of the Michigan state house, and mentioned Parliament in the UK.

The US House is of great interest I presume to US reformers all across the nation, and well known to all, and the election data is available in a pretty easy to read yet comprehensive form too.

But, for specific Constitutional reasons (not exactly the ones people might guess if they are not deeply familiar with what the Constitution actually says!) the general systems of topping off to get a larger body based on ongoing single member districts I have described–in other topics, this topic is about a general model for proportionality–cannot be directly applied to the US House.

The basic problem here is that the Constitution specifies certain fixed numbers of seats be apportioned to the states in proportion to their population.

The floating overhang countering addition approach, that I labeled “Overhang Resolution,” would require doing a new apportionment to account for the extra seats, and then it would be necessary for overhang countering new seats to come from the specific states that gain seats in a global reapportionment, and not from the places where they would most reasonably come from–states whose delegations were in close to proportional balance by FPTP happenstance would be thrown out of balance, make up seats would come from places they are not needed to come from and not come from places where the rebalancing is most needed. It would be a mess! In fact I think the “Alabama Paradox” in the form of a state losing seats when total seat number is raised might also come into effect, denying a perfectly reasonably elected Member a seat and throwing that state’s delegation into chaos. Furthermore other elements of the Constitution presume and depend on states having a fixed delegation between Censuses–the size of the various state Electoral Colleges for instance.

Ideally, the US House ought to be regarded as a body of for and by the voting people of the USA as a whole, without regard to state boundaries. But the Constitution as worded blocks that approach unfortunately.

Absolutely part of the whole plan! Along with hoping it will raise turnout across the board–and indeed a part of low US turnout relates to these uncompetitive races after all.

Even in the US House, there are members who were “elected” with zero votes–several Members from Florida, specifically.

But yes, the problem is pretty massive in state legislatures–in this respect, Nevada actually has a quite mild case of the disease! I gave up attempting to look at the way Louisiana’s practice of a jungle general election with later runoffs could map into the general approach when I found that something like half of all Louisiana’s lower house members just walked into office unopposed and therefore were not voted on in their home districts at all! No vote being held for unopposed candidates is a thing I noticed is true of Florida and Pennsylvania too…but appalled and daunted as I was by this, so far Louisiana’s case takes the cake.

I find it remarkably difficult to get data on these elections by the way. Obviously with FPTP being the deciding thing in US elections, states have little obligation to be more complete than simply presenting the public a list of the candidates who won the elections. Of course they had better be prepared to show the numbers that make these lists well founded, but clearly nothing legal hinges on complete data about also-rans. Each state presents its data differently and I often have great difficulty locating it–I have never found Pennsylvania’s official returns for the record for instance, and not because I did not try hard to look for them!

Every state does something weird. I think I can lay hands on Nevada’s official secretary of state office final return reports for instance, but if you look at it, the results omit to tell which party each candidate listed ran under! This is extra perverse when we recall Nevada has a closed ballot with electronic voting, so every candidate who got any votes whatsoever is unambiguously identified and any partisan affiliation is well known, so not mentioning that one candidate is a Democrat, another a Republican and the third from Independent American Party while the fourth is NPA (no party affiliation, we can’t say “independent” when one of the parties includes that label in its name!) is particularly and pointedly bull-headed. But in other respects the Nevada report is a piece of cake. Fortunately with only 42 seats it is not too terrible a chore to look up which party each candidate ran for and add that in. Consolidating Michigan’s many separated lines for returns from different counties in the same district is pretty daunting, yet easier than having to transcribe some state results by hand. Even the less difficult states take some getting used to to figure out their little tricks and quirks!

This is bad, but far from the most remarkable disturbing thing about Nevada’s outcomes. You seem to be paying attention to Nevada and know things I do not know, though I certainly noticed the unopposed seats early on. As I say other states are much worse than Nevada in that respect, but there is something else bad where I suspect Nevada may be the worst of all–it is hard to tell when locating state returns, then collating them intelligibly and usefully is a chore like pulling teeth!

Well, I actually have a reason perhaps you might not guess at, though there are clues in this post already, why I talk about an 870 or more often 871 (to keep an odd number) US House. To be sure, originally I was just doubling to get a neat margin for proportionality and because that meant an average of two seats per district. But given Constitutional constraints, I think the House must be sized at least 773 members to open the way for proportionality with equal protection under the law.

Parliament which I have mentioned has 650 members–currently that is, as I note the body size has fluctuated over the course of the 20th century. That means that 871 would not be a tremendous increase in scale relative to that, and of course the USA is a lot more than 50 percent more populous than the UK!

But yeah, if there is zero precedent for a working body of 871, that could be a problem, and I think would require a Constitutional amendment to enable proportionality. Now there is a bill in the House to mandate RCV for all states; if this is for IRV only then each district remains similar to every other, but I believe the plan is to enable STV for states with more than one Representative, and I suspect if there were the slightest chance this bill could become law in current circumstances, it would be struck down immediately by the courts for failing to provide equal processes–namely because under RCV, the people of a state with a single Representative would be voting only IRV with no proportionality at all, and that is not equal protection under the law I would think.

If we can raise the House size above 773, there is an approach that would allow for some approach to proportionality but would conform to Constitutional mandates and pass muster, I believe, on the front of equal protection. Of course obstructionists could probably howl about a lot of things but these people are often not arguing in good faith, and must be overruled somehow if one is to have any useful, meaningful reform at all.

Why 773? Because that is the minimum number of seats needed to proportionally, under Huntington-Hill anyway, it might be a little lower with Hamilton, give the smallest state, Wyoming, two seats. The key is to have as many or more seats available for proportional adjustment as there are districts, and so the minimum state apportionment must be two. Note that this would also permit STV to be mandated in the form of keeping the current 435 districts but requiring each district elect two Representatives. Doing that would be a poor approach to proportionality overall since with no cross-aggregation between districts third party options remain poor. But it would address the lack of equal protection the current RCV based proposal in Congress fails. And that would take us directly to 870 seats, with no way to get an odd total.

Whereas technically somewhat less than 871 would do to enable each state to have at least 2 seats; mandating then that the states shall have districts numbering half the apportionment rounded down (so if California gets 105 total members, it is also apportioned 52 districts with 53 more members making up the top-off delegation members). But it is a bad idea to work with the smallest number possible, since some future fluctuation could put Wyoming farther down the relative share of total US population and thus raise the minimum needed to make its share of 2 proportional. Better to have some margin, and going with double the current number of seats plus one to make it odd means about half the individual state district apportionment would be the same as current–there would actually be significantly less than 435 districts due to rounding down odd state apportionments.

So nowadays, that is my major reason for aiming for 871 or higher, though certainly US voters are pretty badly underrepresented by international standards. Again if Parliament can work with 650 members, I don’t see 871 as impossible.

To be sure, one tack obstructionists could easily take would be logistical–I don’t know if it is possible to cram 871 Representatives onto the Capitol floor of their chamber! We’d have to double all the office space and staff too, and possibly have to build a whole new Capitol building. And that same objection applies to all the state Houses too.

This motivated my hitting upon top off based on overhang resolution, which achieves proportionality at a much lower “cost” in terms of cramming more people into the existing facilities. Very sadly it results in unequal districts, but a good version of make up proportionality does that anyway–not all districts ought to have a second representative and some might merit three or more.

We could on paper declare that the anti-overhang top off members, who after all all failed to win a FPTP plurality district seat, are not district representatives at all, but at-large.


Is that just the House of Commons, or the whole Parliament? If the former, please join my movement to stop calling the House of Reps “Congress” and the House of Commons “Parliament”.

I am pretty careful not to call the House of Representatives “Congress” but I guess I have no idea what the House of Lords do all day that anyone should be mindful of them; all British government actions I ever hear of are done by Commons. So yes, I meant the House of Commons.

Florida is not absolutely the only state that elects Representatives with no votes; Jim Bridenstine of Oklahoma was also elected with zero votes in 2014 and 2016 to the House. (He is the NASA administrator now).

The how is pretty simple; no one ran against them in their district. If you don’t want to trust me or Wikipedia on this, here is the page where the House of Representatives’ own record of election statistics, drawn up by the Clerk of the House, records the returns the House itself regards as official. If you scroll down to Florida and in the 2014 and 2016 records, Oklahoma, you can see there some little footnote marks in the party column of the recap table and the footnote says that according to respective state law unopposed candidates are not printed on the ballot at all.

1 Like