“Effective Seat Product” for two-tier PR (including MMP) and MMM

The seat product for a simple electoral system is its assembly size (S) times its mean district magnitude (M) (Taagepera 2007). From this product, MS, the various formulas of the Seat Product Model (SPM) allow us to estimate the effective number of parties, size of the largest, disproportionality, and other election indicators. For each output tested in Shugart and Taagepera (2017), Votes from Seats, we find that the SPM explains about 60% of the variance. This means that these two institutional inputs (M and S) alone account for three fifths of the cross-national differences in party system indicators, while leaving plenty for country-specific or election-specific factors to explain as well (i.e., the other 40% of the variance).

The SPM, based on the simple seat product, is fine if you have a single-tier electoral system. (In the book, we show it works reasonably well, at least on seat outputs, in “complex” but still single-tier systems like AV in Australia, majority-plurality in France, and STV in Ireland.) But what about systems with complex districting, such as two-tier PR? For these systems, Shugart and Taagepera (2017) propose an “extended seat product”. This takes into account the basic-tier size and average district magnitude as well as the percentage of the entire assembly that is allocated in an upper tier, assumed to be compensatory. For estimating the expected effective number of seat-winning parties (NS), the extended SPM formula (Shugart and Taagepera, 2017: 263) is:

NS=2.5t(MB)1/6,

where MB is the basic-tier seat product, defined as the number of seats allocated in the basic tier (i.e., assembly size, minus seats in the upper tier), and t is the tier ratio, i.e., the share of all assembly seats allocated in the upper tier. If the electoral system is simple (single tier), the equation reduces to the “regular” seat product model, in which MS=MB and t=0.

(Added note: in the book we use MSB to refer to what I am calling here MB. No good reason for the change, other than blogger laziness.)

We show in the book that the extended seat product is reasonably accurate for two-tier PR, including mixed-member proportional (MMP). We also show that the logic on which it is based checks out, in that the basic tier NS (i.e., before taking account of the upper tier) is well explained by (MB)1/6, while the multiplier term, 2.5t, captures on average how much the compensation mechanism increases NS. Perhaps most importantly of all, the extended seat product’s prediction is closer to actually observed nationwide NS, on average, than would be an estimate of NS derived from the simple seat product. In other words, for a two-tier system, do not just take the basic-tier mean M and multiply by S and expect it to work!

While the extended seat product works quite well for two-tier PR (including MMP), it is not convenient if one wants to scale such systems along with simple systems. For instance, as I did in my recent planting on polling errors. For this we need an “effective seat product” that exists on the same scale as the simple seat product, but is consistent with the effect of the two-tier system on the effective number of parties (or other outputs).

We did not attempt to develop such an effective seat product in Shugart and Taagepera (2017), but it is pretty straightforward how to do it. And if we can do this, we can also derive an “effective magnitude” of such systems. In this way, we can have a ready indicator of what simple (hypothetical) design comes closest to expressing the impact of the (actual) complex design on the party system.

The derivation of effective seat product is pretty simple, actually. Just take, for the system parameters, the predicted effective number of seat-winning parties, NS, and raise it to the power, 6. That is, if NS=(MS)1/6, it must be that MS=NS6. (Taagepera 2007 proposes something similar, but based on actual output, rather than expected, as there was not to be a form of the seat product model for two-tier systems for almost another decade, till an initial proposal by Li and Shugart (2016).)

Once we do this, we can arrive at effective seat products for all these systems. Examples of resulting values are approximately 5,000 for Germany (MMP) in 2009 and 6,600 for Denmark (two-tier PR) in 2007. How do these compare to simple systems? There are actual few simple systems with these seat products in this range. This might be a feature of two-tier PR (of which MMP could be considered a subtype), as it allows a system to have a low or moderate basic-tier district magnitude combined with a high degree of overall proportionality (and small-party permissiveness). The only simple, single-tier, systems with similar seat products are Poland (5,161), with the next highest being Brazil (9,747) and Netherlands before 1956 (10,000). The implication here is that Germany and Denmark have systems roughly equivalent in their impact on the party system–i.e., on the 60% of variance mentioned above, not the country-specific 40%–as the simple districted PR system of Poland (S=460, M=11) but not as permissive as Brazil (S=513, M=19) or pre-1956 Netherlands (M=S=100). Note that each of these systems has a much higher magnitude than the basic-tier M of Germany (1) or larger assembly than Denmark (S=179; M=13.5). Yet their impact on the nationwide party system should be fairly similar.

Now, suppose you are more interested in “effective district magnitude” than in the seat product. I mean, you should be interested in the seat product, because it tells you more about a system’s impact on the party system than does magnitude alone! But there may be value in knowing the input parameters separately. You can find S easily enough, even for a complex system. But what about (effective) M? This is easy, too! Just take the effective seat product and divide it by the assembly size.

Thus we have an effective M for Germany in 2009 of 7.9 and for Denmark in 2007 of 36.9. These values give us an idea of how, for their given assembly sizes, their compensatory PR systems make district magnitude “effectively”–i.e., in terms of impact on the inter-party dimension–much larger than the basic-tier districts actually are. If we think low M is desirable for generating local representation–a key aspect of the intra-party dimension–we might conclude that Germany gets the advantages M=1 in local representation while also getting the advantages of the proportionality of 8-seat districts. (Best of both worlds?) By comparison, simple districted PR systems with average M around 8 seats include Switzerland and Costa Rica. (The Swiss system is complex in various ways, but not in its districting.) Eight is also the minimum magnitude in Brazil. Denmark gets whatever local representation advantages might come from an actual mean M of 13.5, yet the proportionality, for its assembly size, as if those districts elected, on average, 37 members. Actual districts of about this magnitude occur only in a relatively few districts within simple systems. For instance, the district for Madrid in Spain has M in the mid-30s, but that system’s overall average is only 6.7 (i.e., somewhat smaller than Germany’s effective M).

Now, what about mixed-member majoritarian (MMM) systems. Unlike MMP, these are not designed with a compensatory upper tier. In Votes from Seats, Taagepera and I basically conclude that we are unable to generalize about them. Each system is sui generis. Maybe we gave up too soon! I will describe a procedure for estimating an effective seat product and effective magnitude for MMM systems, in which the basic tier normally has M=1, and there is a list-PR component that is allocated in “parallel” rather than to compensate for deviations from proportionality arising out of the basic tier.

The most straightforward means of estimating the effective seat product is to treat the system as a halfway house between MMP and FPTP. That is, they have some commonality with MMP, in having both M=1 and a list-PR component (not actually a “tier” as Gallagher and Mitchell (2005) explain). But they also have commonality with FPTP, where all seats are M=1 plurality, in that they reward a party that is able to win many of the basic seats in a way that MMP does not. If we take the geometric average of the effective seat product derived as if it were MMP and the effective seat product as if it were FPTP, we might have a reasonable estimate for MMM.

In doing this, I played with both an “effective FPTP seat product” from the basic tier alone and an effective FPTP seat product based on assuming the actual assembly size. The latter works better (in the sense of “predicting,” on average for a set of MMM systems, what their actual NS is), and I think it makes more logical sense. After all, the system should be more permissive than if were a FPTP system in which all those list-PR component seats did not exist. So we are taking the geometric average of (1) a hypothetical system in which the entire assembly is divided into a number of single-seat electoral districts (Eeff) that is Eeff = EB+tS, where EB is the actual number of single-seat districts in the basic tier and S and t are as defined before, and (2) a hypothetical system that is MMP instead of MMM but otherwise identical.

When we do this, we get the following based on a couple sample MMM systems. In Japan, the effective seat product becomes approximately 1,070, roughly equivalent to moderate-M simple districted PR systems in the Dominican Republic or pre-1965 Norway. For South Korea, we would have an effective seat product of 458, or very roughly the same as the US House, and also close to the districted PR system of Costa Rica.

Here is how those are derived, using the example of Japan. We have S=480, with 300 single-seat districts and 180 list-PR seats. Thus t=0.375. If it were two-tier PR (specifically, MMP), the extended seat product would expect NS=3.65, from which we would derive an effective seat product, (MS)eff=3.666 =2,400. But it is MMM. So let’s calculate an effective FPTP seat product. Eeff = EB+tS=300+180=480 (from which we would expect NS=2.80). We just take the geometric mean of these two seat-product estimates: (2400*480)1/2=1,070. This leads to an expected NS=3.19, letting us see just how much the non-compensatory feature reduces expected party-system fragmentation relative to MMP as well as how much more permissive it is than if it were FPTP.

How does this work out in practice? Well, for Japan it is accurate for the 2000 election (NS=3.17), but several other elections have had much NS lower. That is perhaps due to election-specific factors (producing huge swings in 2005 and 2009, for example). As I alluded to above already, over the wider set of MMM systems, this method is pretty good on average. For 40 elections in 17 countries, a ratio of actual NS to that predicted from this method is 1.0075 (median 0.925). The worst-predicted is Italy (1994-2001), but that is mainly because the blocs that formed to cope with MMM contained many parties (plus Italy’s system had a partial-compensation feature). If I drop Italy, I get a mean of 1.0024 (but a median of only 0.894) on 37 elections.

If we want an effective magnitude for MMM, we can again use the simple formula, Meff=(MS)eff/S. For Japan, this would give us Meff=2.25; for Korea Meff=1.5. Intuitively, these make sense. In terms of districting, these systems are more similar to FPTP than they are to MMP, or even to districted PR. That is, they put a strong premium on the plurality party, while also giving the runner-up party a considerable incentive to attend to district interests in the hopes of swinging the actual district seat their way next time (because the system puts a high premium on M=1 wins, unlike MMP). This is, by the way, a theme of the forthcoming Party Personnel book of which I am a coauthor.

(A quirk here is that Thailand’s system of 2001 and 2005 gets an effective magnitude of 0.92! This is strange, given that magnitude–the real kind–obviously has a lower limit of 1.0, but it is perhaps tolerable inasmuch as it signals that Thailand’s MMM was really strongly majoritarian, given only 100 list seats out of 500, which means most list seats would also be won by any party that performed very well in the M=1 seats, which is indeed very much what happened in 2005.)

In this planting, I have shown that it is possible to develop an “effective seat product” for two-tier PR systems that allows such systems to be scaled along with simple, single-tier systems. The exercise allows us to say what sort of simple system an actual two-tier system most resembles in its institutional impact on inter-party variables, like the effective number of seat-winning parties, size of the largest party, and disproportionality (using formulas of the Seat Product Model). From the effective seat product, we can also determine an “effective magnitude” by simply dividing the calculated effective seat product by actual assembly size. This derivation lets us understand how the upper tier makes the individual district effectively more proportional while retaining an actual (basic-tier) magnitude that facilitates a more localized representation. Further, I have shown that MMM systems can be treated as intermediary between a hypothetical MMP (with the same basic-tier and upper-tier structure) and a hypothetical FPTP in which the entire assembly consists of single-seat districts. Again, this procedure can be extended to derive an effective magnitude. For actual MMP systems in Germany and also New Zealand, we end up with an effective magnitude in the 6–8 range. For actual MMM systems, we typically get an effective magnitude in the 1.5–3 range.

I will post files that have these summary statistics for a wide range of systems in case they may be of use to researchers or other interested readers. These are separate files for MMM, MMP, and two-tier PR (i.e, those that also use PR in their basic tiers), along with a codebook. (Links go to Dropbox (account not required); the first three files are .CSV and the codebook is .RTF.)

Added note: In the spreadsheets, the values of basic-tier seat product (MB) and tier ratio (t) are not election-specific, but are system averages. We used a definition of “system” that is based on how Lijphart (1994) defines criteria for a “change” in system. This is important only because it means the values may not exactly match what you would calculate from the raw values at a given election, if there have been small tweaks to magnitude or other variables during an otherwise steady-state “system”. These should make for only very minor differences and only for some countries.

Does the electoral system affect polling errors, and what about presidentialism?

I will attempt to answer the questions in the title through an examination of the dataset that accompanies Jennings and Wlezien (2018), Election polling errors across time and space. The main purpose of the article is to investigate the question as to whether polls have become less reliable over time. One of their key findings can be summarized from the following brief excerpt:

We find that, contrary to much conventional wisdom, the recent performance of polls has not been outside the ordinary; if anything, polling errors are getting smaller on average, not bigger.

A secondary task of Jennings and Wlezien is to ask whether the institutional context matters for polling accuracy. This sort of question is just what this virtual orchard exists for, and I was not satisfied with the treatment of electoral systems in the article. Fortunately, their dataset is available and is in Stata format, so I went about both replicating what they did (which I was able to do without any issues) and then merging in other data I have and making various new codings and analyses.

My hunch was that, if we operationalize the electoral system as more than “proportional or not”, we would find that more “permissive” electoral systems–those that favor higher party-system fragmentation and proportionality–would tend to have larger polling errors. I reasoned that when there are more parties in the system (as is usually the case under more permissive systems), voters have more choices that might be broadly acceptable to them, and hence late shifts from party to party might be more likely to be missed by the polls. This is contrary to what the authors expect and find, which is that mean absolute error tends to be lower in proportional representation (PR) systems than under “SMD” (single-member districts, which as I always feel I must add, is not an electoral system type, but simply a district magnitude). See their Table 2, which shows a mean absolute error in the last week before electoral day of 1.62 under PR and 2.28 under “SMD”.

The authors also expect and show that presidential elections have systematically higher error than legislative elections (2.70 vs. 1.83, according to the same table). They also have a nifty Figure 1 that shows that presidential election polling is both more volatile over the timeline of a given election campaign in its mean absolute error and exhibits higher error than legislative election polling at almost any point from 200 days before the election to the last pre-election polls. Importantly, even presidential election polls become more accurate near the end, but they still retain higher error than legislative elections even immediately before the election.

This finding on presidential elections is consistent with my own theoretical priors. Because presidential contests are between individuals who have a “personal vote” and who are not necessarily reliable agents of the party organization, but are selected because their parties think they can win a nationwide contest (Samuels and Shugart, 2010), the contest for president should be harder to poll than for legislative elections, all else equal. That is, winning presidential candidates attract floating voters–that is pretty much the entire goal of finding the right presidential candidate–and these might be more likely to be missed, even late in the campaign.

To test my own hunches on the impact of institutions on polling errors, I ran a regression (OLS) similar to what is reported in the authors’ Table 3: “Regressions of absolute vote-poll error using polls from the week before Election Day.” This regression shows, among other results, a strong significant effect of presidential elections (i.e., more polling error), and a negative and significant effect of PR. It also shows that the strongest effect among included variables is party size: those parties that get more than 20% of the vote tend to have larger absolute polling errors, all else equal. (I include this variable as a control in my regression as well.)

The main item of dissatisfaction for me was the dichotomy, PR vs. SMD. (Even if we call it PR vs. plurality/majority, I’d still be dissatisfied). My general rule is do not dichotomize electoral systems! Systems are more or less permissive, and are best characterized by their seat product, which is defined as mean district magnitude times assembly size. Thus I wanted to explore what the result would be if I used the seat product to define the electoral system.

I also had a further hunch, which was that presidential elections would be especially challenging to poll in institutional settings in which the electoral system for the assembly is highly permissive. In these cases, either small parties enter the presidential contest to “show the flag” even though they may have little chance to win–and hence voters may be more likely to defect at the end–or they form pre-election joint candidacies with other parties. In the latter case, some voters may hedge about whether they will vote for a candidate of an allied party when their preferred party has no candidate. Either situation should tend to make polling more difficult, inflating error even late in the campaign. To test this requires interacting the seat product with the binary variable for election type (presidential or legislative). My regression has 642 observations; theirs has 763. The difference is due to a few complex systems having unclear seat product plus a dropping of some elections that I explain below. Their findings hold on my smaller sample with almost the precise same coefficients, and so I do not think the different sample sizes matter for the conclusions.

When I do this, and graph the result (using Stata ‘margins’ command), I get the following.

I am both right and wrong! On the electoral system effect, the seat product does not matter at all for error in legislative elections. That is, we do not see either the finding Jennings and Wlezien report of lower error under PR (compared to “SMD”), nor my expectation that error would increase as the seat product increases–EXCEPT: It seems I was right in my expectation that error in presidential contests increases with the seat product of the (legislative) electoral system.

The graph shows the estimated output and 95% confidence intervals for presidential elections (black lines and data points) and for legislative (gray). We see that the error is higher, on average, for presidential systems for all seat products greater than a logged value of about 2.75, and increasingly so as the seat product rises. Note that a logged value of 2.75 is an unlogged seat product of 562. Countries in this range include France, India, the Dominican Republic, and Peru. (Note that some of these are “PR” and some “SMD”; that is the point, in that district magnitude and formula are not the only features that determine how permissive an entire national electoral system is–see Shugart and Taagepera, 2017.)

I have checked the result in various ways, both with alternative codings of the electoral system variable, and with sub-sets, as well as by selectively dropping specific countries that comprise many data points. For instance, I thought maybe Brazil (seat product of 9,669, or a logged value just short of 4) was driving the effect, or maybe the USA (435; logged =2.64) was. No. It is robust to these and other exclusions.

For alternatives on the coding of electoral system, the effect is similar if I revert to the dichotomy, and it also works if I just use the log of mean district magnitude (thereby ignoring assembly size).

For executive format types, running the regression on sub-samples also is robust. If I run only the presidential elections in pure presidential systems (73 obs.), I still get a strong positive and significant effect of the seat product on polling error. If I run only on pure parliamentary systems (410 obs.), I get no impact of the seat product. If I restrict the sample only to semi-presidential systems (159 obs.), the interactive effect holds (and all coefficients stay roughly the same) just as when all systems are included. So it seems there is a real effect here of the seat product–standing in for electoral system permissiveness–on the accuracy of polling near the end of presidential election campaigns.

I want to briefly describe a few other data choices I made. First of all, legislative elections in pure presidential systems are dropped. The Jennings and Wlezien regression sample actually has no such elections other than US midterm elections, and I do not think we can generalize from that experience to legislative vs. presidential elections in other presidential systems. (Most are concurrent anyway, as is every presidential election in the US and thus the other half of the total number of congressional elections.)

However, I did check within systems where we have both presidential and legislative polls available. All countries in the Jennings-Wlezien regression sample that are represented by both types of election are semi-presidential, aside from the US. In the US, Poland, and Portugal, the pattern holds: mean error is greater in presidential elections than in assembly elections in the same country. But the difference is significant only in Portugal. In Croatia the effect goes the other way, but to a trivial degree and there are only three legislative elections included. (If I pool all these countries, the difference across election types is statistically significant, but the magnitude of the difference is small: 2.22 for legislative and 2.78 for presidential.)

The astute reader will have noticed that the x-axis of the graph is labelled, effective seat product. This is because I need a way to include two-tier systems and the seat product’s strict definition (average magnitude X assembly size) only works for single-tier systems. There is a way to estimate the seat product equivalent for a two-tier system as if it were simple. I promise to explain that some time soon, but here is not the place for it. (UPDATE: Now planted.)

I also checked one other thing that I wanted to report before concluding. I wondered if there would be a different effect if a given election had an effective number of parties (seat-winning) greater than expected from its seat product. The intuition is that polling would be tend to off more if the party (or presidential) contest were more fragmented than expected for the given electoral system. The answer is that it does not alter the basic pattern, whereby it makes no difference to legislative elections (in parliamentary or semi-presidential systems). For presidential elections, there is a tendency for significantly higher error the more the fragmentation of the legislative election is greater than expected for the seat product. The graph below shows a plot of this election; as you can probably tell from the data plot, the fit of this regression is poorer than the one reported earlier. Still, there may be something here that is worth investigating further.

MMP in NZ: An example of “best of both worlds” in action

In Shugart and Wattenberg (2001) we ask if mixed-member systems offer a “best of both worlds.” That is, do they allow simultaneously for the benefits of local representation and individual-member accountability that are the (supposed) advantages of single-seat plurality (FPTP) and the representation of smaller national parties that might struggle to win districts but would be represented under proportional representation (PR).

There was a question mark in the book’s subtitle. Over time, I have come to believe that indeed the proportional type (MMP) does have a strong tendency to offer the best of both worlds. The reason is that members elected in districts have incentives to behave as local representatives at the time that there is close approximation between party vote and seat shares (assuming compensation is carried out nationwide or in large regions). The majoritarian type (MMM, as in Japan and Taiwan) probably does not; it is much closer in its overall incentive structure to FPTP, even though it does indeed permit smaller national parties to win seats.

For MMP, the “best of both worlds” argument assumes that parties nominate dually–meaning many elected members will have run in a district and had a (realistically electable) list position simultaneously. If they do, then even the list-elected members will have a local base, and should have incentives to act as the local “face” of the party, including possibly by offering constituent services. Both prior anecdotes I have shared from New Zealand (e.g., “shadow MPs” who win from the list and maintain a local office) and my forthcoming coauthored book, Party Personnel, offer further evidence that MMP does indeed work in this way.

Now comes a terrific anecdote from New Zealand’s 2020 election. In this election, Labour won a majority of seats (64/120) with 49.1% of the nationwide party list vote. In the nominal tier of single-seat districts (electorates) it won 43 of the 72 available seats. Its win included some districts that are normally strongholds of the center-right National Party (which won 35 seats overall and just 26 districts).

Commenting on some of the Labour wins in mostly rural districts, Federated Farmers president Andrew Hoggard said:

in some “flipped” electorates Labour list MPs had worked hard to raise their profile and get involved with the community and this had paid off when they campaigned for the electorate.

This is an ideal description of how the “best of both worlds” argument works: list-elected members have incentives to attend to local needs of the district in which they ran for the nominal seat (but “lost”) in hopes of capturing the local plurality in the next election.

Of course, there were other factors at work as well. I will offer another planting about one of those factors separately. There is also some uncertainty at this stage just exactly the degree to which rural voters flipped, as the wins may have come in significant part from very large swings in the town areas within districts that also include large rural areas. Regardless, MMP offers the key advantage of giving most elected members, if dually nominated, a tie to a local constituency while ensuring close approximation of overall seat totals to party-list votes.

Academics and journalists

This is such an interesting comment about academics and journalists by Andrew Gelman, in response to question as to whether he and Nate Silver might do a joint podcast or other discussion about election forecasting (Gelman says he’s asked and Silver has not responded):
____
The more general question, maybe, is how journalists and academics can interact. A traditional model is that the academic does the research and the journalist writes about it. Or the academic does the work and the journalists writes about it with a critical eye, Felix Salmon style. A different model is that the journalist and the researcher are the same person: that’s what Nate [Silver] is doing. Maybe a better way to put this is that the “journalist” and “academic” roles have been erased and replaced by the analyst, who does both. Bill James was a pioneer in this. Finally, there’s the model in which the academics and journalists collaborate, which is what Merlin and I are doing with Elliott [Morris]. At this point, you might ask, why do Merlin and I need Elliott at all: why would a forecast by two political scientists be improved by a journalist? The immediate answer is that the Economist forecast is Elliott’s baby: he came to us to ask for help. The longer answer is that 3 people are better than 2, and the distinction between academic and journalist is not always so clear. I do a lot of writing, Elliott does a lot of programming, and we both have thought a lot about politics. I’ve found that collaboration almost always makes things better, as long as the collaborators can get along.
Anyway, Nate seems pretty set in his go-it-alone, don’t involve academic researchers approach, and I really like to collaborate, so maybe that’s one reason we’re having difficulty communicating.
Also, unrelatedly, Nate is a public figure and so he suffers from what I’ve called the David Brooks or Paul Krugman problem: he gets so much low-quality criticism from randos on the internet, that he’s developed a way of pattern of ignoring or firing back at criticism, rather than engaging with it directly. It can be hard to have a conversation, public or private, with someone who’s gotten into the habit of considering outside criticism as a nuisance rather than a source of valuable input.

Dividing the Rulers: How Majority Cycling Saves Democracy

The following is a guest planting by Dr. Yuhui Li. I suggested Dr. Li draft something for this blog about his recently published book. (Note: I was the Ph.D. committee chair for Huey at UC Davis.)

You can buy the book by clicking here, and taking advantage of a discount! You can get 30% off by entering the code, UMCYCLING (for a limited time).


I’m grateful that Matthew offers to post this introduction of my book Dividing the Rulers: How Majority Cycling Saves Democracy. I hope it may interest some fellow political scientists.

The initial thought about the project originated from a debate I had years ago on Chinese social media regarding the choice of political systems. It appeared to me that many people were skeptical about the idea of democracy out of the fear that the majority, at least in theory, could be as tyrannical as individual dictators. But years later, when I learned about the social choice theories as a PhD student, I noticed an almost opposite criticism of democracy from academia, namely the instability of social choice. Both arguments sound convincing, but also at odds with each other: It’s hard to imagine a decision-making body that is both cyclical and tyrannical. In the process of solving the puzzle, I came across Nicholas Miller (1983) and Anthony McGann (2006) and realized that the two arguments can be reconciled. Cycling and the tyranny of the majority can both exist, but they can be negatively correlated with each other. As cycling is clearly the lesser of the two evils, it may be exactly the reason that democracies tend to be less tyrannical than alternative systems.

So building on Miller and McGann, I started to develop a theory and design an experiment to untangle the process of how cycling can actually be a good thing and help the temporary losers of an electoral game. I argue that in a voting body, the key factor that prevents cycling is the cost incurred on those who defect from the winning coalition. If a country’s legislative body has a low “defection cost”, cycling is more likely, and the distributive outcome more equal.

I have to admit, unfortunately, that my modeling skills are not enough to formalize a game with continuous options and highly unstable equilibria, but I did present a rather convincing strategic process showing that in a three-player committee, if the defection cost is higher than 50% of the distributable benefits in a giving round, cycling cannot happen. I then conducted an experiment by grouping respondents into three-player committees to verify that process. The experiment results turned out to be more variant than I had expected, but largely confirmed the hypothesis that a high defection cost can deter cycling and result in a more “tyrannical” outcome, with less frequent power alternation. I show that while cycling introduces uncertainty into the policy outcome, it is exactly that short-run uncertainty that creates the long-run equality by reshuffling winners and losers.

In the second half of the book, I connect such a phenomenon to the design of electoral institutions. I argue that low defection costs explain the favorable distributive outcome in countries in which the parliament does not have a majority party and the executive is subordinate to that parliament. This way, not only is the winning coalition more inclusive, but more importantly, it is vulnerable to defection and gives the losing side a better bargaining position.

While it is a common belief in the literature that the outcome of electoral system design cannot be predicted with high accuracy, I construct a comprehensive dataset on countries’ largest party vote shares and show that a no-majority party system can be guaranteed in almost any country as long as the electoral system is sufficiently proportional. I explain such a phenomenon by developing a demand and supply theory of political parties, explaining why, with very few exceptions such as South Africa’s ANC, a majority party’s politicians and voters both have strong incentives to split as long as the electoral system allows small parties to survive. And therefore the aforementioned theory that cycling leads to equality is not merely a thought exercise, but an attainable outcome giving appropriate institutional design.

There are two features of this book that I think are worth noting. First, it goes beyond positive empirical study of political institutions and offers a clear normative objective for institutional design, which is to ensure an unstable winning coalition to guard against a tyrannical government, whether it represents a majority or not. And second, the book is accessible for readers with minimal political science training, as I strive to convince as many people as possible by using intuitive methods and providing explanations to advanced concepts.

Why so much “high policy” in the LDP?

I am close(-ish) to finishing up a book on Party Personnel, coauthored with Matthew Bergman, Cory Struthers, Ellis Krauss, and Robert Pekkanen. The short version of what the book is about: How does the electoral system (and electoral reform) shape how parties deploy their “personnel” (i.e. elected legislators) to legislative committees to allow them to engage in activities that support the party organizational goal of seat-maximization?

(Quick note: While the book is coauthored, I should make clear that this post is just my musing about a puzzle, and not a piece of the book. Nor do my coauthors bear any responsibility for what I am writing here, or conclusions I attempt to draw.)

The outcome variable of interest in the Party Personnel project is the “type” of committee assignment a legislator receives–high policy, public goods, or distributive. This typology first appeared in Pekkanen, Nyblade, and Krauss (2006). In most of the countries covered in the book, a party typically has 50%-60% of its legislators sitting on high policy in any given term of parliament (not necessarily all at once, as members may be rotated). But, in Japan, the Liberal Democratic Party often has 75% or more on high policy, and in some years over 90%!

What explains the unusually high rate of high policy committee assignments in the LDP? I do not think we currently have a good explanation. We have floated some in internal discussions (mostly internal to my own brain), but as I will show in this post, they are not adequate. Before we get into trying to answer the question posed in bold at the start of this paragraph, let’s establish which committees are classified as high policy. These are committees charged with involvement in policies that are about the management of the economy and other matters of state–this is the sense in which they are “high”. Examples are finance, economy, budget, justice, defense, and foreign affairs.

Japan underwent a major electoral reform prior to the 1996 election. The old system was single non-transferable vote (SNTV). The post-reform system is mixed-member majoritarian (MMM). If we look only at averages by electoral-system era, it looks like a case of electoral reform resulting in a large increase in the percentage of LDP members of the House of Representatives obtaining high policy (HP) committee assignments:

SNTV era: 73% of legislators on HP.

MMM era: 86% of legislators on HP.

The difference in means when comparing eras is statistically significant. So, it is the electoral reform, right? (Never mind that even 73% pre-reform would be high, compared to other cases.) Is there a reason why MMM would lead a party to want to emphasize the experience of its caucus in HP more than would be the case under SNTV? Amy Catalinac‘s book, Electoral Reform and National Security in Japan (2016), suggests a reason why the answer might be yes. She analyzes individual candidates’ campaign manifestoes and finds that they are more likely to mention defense under MMM than they were under SNTV. The reason she gives is that members’ having a reputation for high policy areas like defense and foreign affairs was not useful under SNTV, when they were overwhelmingly concerned with “pork” for which they could get individual credit. By contrast, under MMM, each member is the sole standard-bearer in a single-seat district (and is often also running on a closed party list), and thus the pork incentive is greatly diminished. As national security is a key component of high policy, a similar effect might also account for the higher rate of HP committee assignments after electoral reform. That is, HP committees are assigned to LDP legislators at a greater rate under MMM because the party as a whole has a stronger interest in appearing credible on high-policy issues, including national security. This is one plausible explanation of the changes in era averages. For reasons I will turn to later, I am not satisfied with this attempt to explain the patterns in committee assignment.

An initial idea I had was that perhaps the high rate of HP, post-reform, is a legacy of a high rate under SNTV, and will be seen to have declined under MMM. In other words, this is a totally competing explanation to the one that could be derived from a logic similar to that of Catalinac. I would expect more delegation to the cabinet under MMM, in the sense of its being more Westminster-like than under SNTV. (This is a point that Catalinac explicitly argues, in justifying her disagreement with some Japan specialists who, upon the electoral reform, thought pork would remain just as important, only district-focused rather than more narrowly focused as under SNTV.) The problem with expecting a shift towards more HP committee membership under a supposedly more Westminster-like system, or just generally a more policy-centric system, is that it does not comport well with our comparative evidence.

We know that in two Westminster systems covered in the book, Britain and pre-reform New Zealand, the rate of HP committee membership is not especially high (30% in UK Conservatives, 33.5% in Labour; NZ National 54%, Labour 68%). All of these percentages are lower–most are much lower–than the MMM-era mean for the LDP. Nor is it high in Germany (near 60% in both major parties) or post-reform NZ (46% National, 57% Labour), our two MMP cases. These are all systems in which we expect party policy reputation to matter more than individual reputation. So, if Japan moves from SNTV, with its strong focus on the individual, to MMM, with greater focus on the party as a “team” seeking to take on (or hold on to) governing, why would HP be high under MMM?

With this as a puzzle, it seemed likely it might have been just a legacy of pre-reform SNTV. Perhaps, under SNTV, there actually is a logic to getting nearly everyone on HP because your electoral system makes everyone need to have a unique personal reputation. While this does, as noted, run up against the problem that the HP percentage is higher post-reform, it is worth entertaining why its being high under SNTV might not itself be inconsistent with the incentives of the system.

Why might HP for almost everyone be a strategy compliant with SNTV? Maybe the former SNTV system led the party to want most of its members to gain experience in high-policy areas because each member needs to be “his own party”. (Under SNTV, especially, LDP legislators have been overwhelmingly male.) This possible explanation seems at first to be in tension with common expectations that members under SNTV have to differentiate themselves, such as by credit-claiming for pork. Because votes are not pooled (or transferred) among co-partisans, the party needs a way to divide the vote efficiently in order to maximize seats. The literature on vote-division emphasizes how members need to be distinct from one another, so why have almost everyone on high policy?

Actually, this piece of the larger puzzle turns out not to be such a puzzle at all, even though it initially struck me as odd. Even if all legislators (2 or 3) from a given district are on HP, they can still differentiate by being on different HP committees (one on budget, one on defense, for example). Thus having many members on HP and having them develop their own personal reputations are not in any way contradictory. However, differentiating on sub-categories within HP may still not be as beneficial to claiming credit for things uniquely attributable to a specific politician as are pork-related benefits. A common expectation in the vote-division literature (including a key point of Catalinac’s thesis) is that pork is more useful for vote-division than high policy.

Even if we accept that HP is likely suboptimal for vote-division purposes, having everyone on HP does not preclude legislators also being on more specifically district-focused committees in the areas of public goods (e.g., health or education) or distributive (e.g., construction, agriculture, or transport). In fact, the LDP also has an unusually high rate (relative to other parties among the book’s cases) of category overlap. The average member is on about 1.6 of our three categories, unlike in most other countries where the figure is in the range of 1.2 to 1.4. In other words, by our categories, it is members in other countries that are the specialized ones. In the LDP, they are not; they are actually closer to being generalists, at least in this sense.

Having assignments in multiple categories allows each legislator to build a personal portfolio, almost like a micro-party, in which they are involved in some HP task and also frequently a task in either public goods or distributive during the same term. This kind of portfolio could be very useful for building the personal electoral coalition each individual needs to ensure election under an electoral system that pits members of the same party against each other in multi-seat districts, with no party-level vote-pooling. In other words, SNTV.

If the building of a personal portfolio, including but not limited to HP, due to SNTV incentives were the explanation, we should see a decline after the change to MMM. In the graph below, we will see that we do! So that is good. However, we still face the problem that it should not be higher on average during the era of MMM (so far) than it was during the era of SNTV. Yet that is precisely what we saw in the era averages shown above. Now, let’s disaggregate this thing called an electoral-system era. The graph shows the percentage on HP in each election from 1980 through 2009. The vertical line demarcates the eras, marking the first election under MMM.

The decline under MMM is certainly consistent with the notion that HP is less important to the individual legislator over time, given an electoral system that has eliminated intraparty competition in multi-seat districts. Additionally, the 1990 and 1993 elections show exceptionally high levels of personal portfolios including HP at the end of the SNTV era. Yet look how low the rate is before 1990. Thus, while we might be able to say that adaptation to MMM is leading to an expected decline in HP service after 1996 (albeit perhaps too slowly by 2009), we obviously should not conclude that it started so high because of SNTV! It was low under SNTV… until it was high. It then peaked in the first MMM election, before beginning a decline.

So what changed to lead the LDP suddenly to want nearly everyone to have high policy in their portfolio in 1990 and 1993? This–finally–is the question I am crowd-sourcing. The context of the 1990 election is one in which the LDP had just lost its majority in the second chamber (the House of Councillors). There was also a divisive debate at the time on enacting a national consumption tax, around the same time that real estate and stock market valuations were unsustainably high (leading to a subsequent period of economic stagnation). Also in 1993, due to splits over various issues including electoral reform, the LDP actually failed to win a parliamentary majority and found itself temporarily in opposition. How these would explain a surge in HP is not clear to me. So I am not proposing an explanation, but that is the context and perhaps points towards an explanation.

The pattern is similar if we look at category overlap. Again, this simply measures how many categories–out of high policy, public goods, and distributive–a member sits on. So, ignoring anyone who is on no committees in these categories (and I am ignoring such rare birds here), it can range from 1 to 3. The next graph shows this averaged across LDP legislators by election year. Again, the vertical line marks 1996, the first election under MMM. As we see, it is only 1990, 1993, and 1996 when the average number of committee categories per legislator approaches or exceeds 2.0. While lower in the other seven election, it is still above 1.5 in all elections except the three earliest SNTV elections in the sample (1980 through 1986) and then again in 2003 and 2009 of the MMM era. In the five just mentioned, it is between 1.3 and 1.4.

The pattern is again consistent with the LDP deciding for some reason before the 1990 committee allocations that it needed each member to have both high policy and at least one committee assignment from another category. After 1996, this category overlap becomes markedly less common (but still is higher than in most other parties we cover in the book).

A final graph breaks the HP category down and shows three of the main committees within it.

It is clear that the pattern of surging HP membership in 1990 is largely about the budget committee. More than half the party’s legislators sat on this committee in the legislatures elected in 1990, 1993, and 1996. The other two included here (economy and defense) show smaller bumps as well, just not rising as high. This again would be potentially consistent with my proposed explanation that “everyone needs a diverse portfolio under SNTV”, provided we can add a condition as to why this need is only actually realized at the end of the SNTV period and not the entire era. So, the electoral system explanation needs to be augmented by some political factor, like fracturing of the party and greater pressure from the opposition.

It is further worth noting that of all “high policy” committees, it is budget and economy that would have the greatest “pork” potential. While I would not want to reclassify committees dealing with such clearly aggregate national matters to the distributive category–they are clearly “high” topics–they do allow for opportunities to claim credit. Did the LDP simply need this even more in the 1990s than in the 1980s? The patterns certainly suggest they need it less under MMM, at least after the first MMM election of 1996. The patterns also show that defense committee assignments very quickly went back down in the 2000s after their peak with the first MMM election. So, while they may continue to talk about national security in the individual campaign manifestoes (per Catalinac), few LDP members by 2009 are sitting on committees where they can actually be involved in such policy discussions.

This has been a long post. Thank you to anyone who made it this far! I hope someone has some suggestions for why these patterns are seen in the data. The book does not look at time factors in committee assignments, except for pre-reform and post-reform eras where there has been a change of electoral system (Japan and New Zealand, among our cases). However, any attempt to explain the anomalously high averages in high policy assignment and committee-category overlap in Japan has to grapple with the fact that there is a within-era variance in the LDP that is quite stark. It is not just a story about two different electoral systems, nor is it something immutable about the LDP.

Canada and UK 2019: District level fragmentation

With two of the big Westminster parliamentary democracies having had general elections in 2019, we have a good opportunity to assess the state of district-level competition in FPTP electoral systems.

(Caution: Deep nerd’s dive here!)

Before we turn to the district level, a short overview of what is expected at the national level is in order.

As noted previously, Canada’s election produced a nationwide seat balance that was extremely close to what we expect from the Seat Product Model (SPM), yet the nationwide votes were exceedingly fragmented (and, anomalously, the largest seat-winning party was second in votes). The UK election, on the other hand, was significantly less fragmented in the parliamentary outcome than we expect from the SPM, even if it was in key respects a “typical” FPTP outcome in terms of manufacturing a majority for a party with less than a majority of the vote.

In general, over decades, Canada tends to conform well to the SPM expectation for the shape of its parliamentary party system, whereas the UK is a more challenging case from the SPM’s perspective.

The SPM states that the effective number of seat-winning parties (NS) should be the seat product, raised to the power, 1/6. The seat product is the assembly size, times the mean district magnitude. The SPM predictions for NS explain around 60% of the variance in actual outcomes for elections around the world under a wide variety of electoral systems. SPM predictions for other output quantities also explain in the neighborhood of 60%. So the SPM is both successful at explaining the real world of seat and vote fragmentation, and leaves plenty of room for country-specific or election-specific “other factors” (i.e., the other 40%). The SPM is based on deductive logic, starting from the minimum and maximum possible outcomes for a given number of seats at stake (in a district or an assembly). The logic is spelled out in Votes from Seats.

In the case of a FPTP system, the SPM makes the bold claim that we can understand the shape of a party system by knowing only the assembly size. That is because with district magnitude, M=1, the seat product is fully described by the country’s total number of seats, S, which is also the number of districts in which the voting is carried out. Thus we expect NS=S1/6. Let’s call this “Equation 1.”

For Canada’s current assembly size (338), this means NS=2.64, as an average expectation. Actual elections have tended to come pretty close–again, on average. Of course, individual elections might vary in one direction or the other. (The assembly size was also formerly smaller, but in recent times, not by enough to concern ourselves too much for purposes of this analysis.) For the UK, the corresponding expectation would be 2.94 based on a seat product of 650.

The actual Canadian election of 2019 resulted in NS=2.79; for the UK it was 2.39. Thus for Canada, we have a result very close to the expectation (ratio of actual to expected is 1.0578). For the UK, the actual result was quite short (ratio of 0.8913). As I said, the UK is a challenging, even aberrant, case– at least at the national level.

What about the district level? A national outcome is obviously somehow an aggregation of all those separate district-level outcomes. The SPM, however, sees it differently. It says that the districts are just arenas in which the nationwide election plays out. That is, we have a logical grounding that says, given a national electoral system with some seat product, we know what the nationwide party system should look like. From that we can further deduce what the average district should look like, given that each district is “embedded” in the very same national electoral system. (The logic behind this is spelled out in Votes from Seats, Chapter 10).

The crazy claim of the SPM, district-level extension, is that under FPTP, assembly size alone shapes the effective number of votes-earning parties in the average district (N’V, where the prime mark reminds us that we are talking about the district-level quantity rather than the nationwide one). (Note that for FPTP, it must be the case that N’S=1, always and in every district).

The formula for expected N’V under FPTP is: N’V=1.59S1/12 (Equation 2). It has a strictly logical basis, but I am not going to take the space to spell it out here; I will come back to that “1.59” below, however. It is verified empirically on a wide set of elections, including those from large-assembly FPTP cases like Canada, India, and the UK. So what I want to do now is see how the elections of 2019 in Canada and UK compare to this expectation. (Some day I will do this for India’s 2019 election, too.)

If the effective number of seat-winning parties at the national level (NS) is off, relative to the SPM, then it should be expected that the average district-level effective number of vote-earning parties (N’V) would be off as well. They are, after all, derived from the same underlying factor–the number of single-seat districts, i.e., the assembly size (S). We already know that NS was close to expectation in Canada, but well off in the UK in 2019. So how about the districts? In addition to checking this against the expectation from S alone, we can also check one other way: from actual national NS. We can derive an expected connection of N’V to NS via basic algebra. We just substitute the value from one equation into the other (using Equations 1 and 2). If we have NS=S1/6 then it must be that S= NS6. So we can substitute:

N’V=1.59(NS6)1/12= 1.59√NS (Equation 3).

In a forthcoming book chapter, Cory L. Struthers and I show that this works not only algebraically, but also empirically. We also suggest a logical foundation to it, which would require further analysis before we would know if it is really on target. The short version suggested by the equation is that the voting in any given district tends to be some function of (1) the basic tendency of M=1 to yield two-candidate competition (yes, Duverger!) in isolation and (2) the extra-district viability of competing parties due to the district’s not being isolated, but rather embedded in the national system. The 1.59, which we already saw in Equation 2, is just 22/3; it is the expected N’V if there were exactly two vote-earning parties, because it is already established–by Taagepera (2007)–that the effective number tends to be the actual number, raised to the power, two thirds. And the square root of NS suggests that parties that win some share of seats (i.e., can contribute more or less to the value of NS) tend to attract votes even though they may have no chance of winning in any given district. By having some tendency to attract votes based on their overall parliamentary representation, they contribute to N’V because voters tend to vote based on the national (expected, given it is the same election) outcome rather than what is going on in their district (about which they may have poor information or simply not actually care about). If the parliamentary party system were fully replicated in each district, the exponent on NS would be 1. If it were not replicated at all, the exponent would be zero. On average, and in absence of any other information, it can be expected to be 0.5, i.e., the square root.

How does this hold up in the two elections we are looking at in 2019? Spoiler alert: quite well in the UK, and quite badly in Canada. Here are graphs, which are kernel density plots (basically, smoothed histograms). These plots show how actual districts in each election were distributed across the range of observed values of N’V, which in both elections ranged from around 1.35 to just short of 4.5. The curve peaks near the median, and I have marked the arithmetic mean with a thin gray line. The line of most interest, given the question of how the actual parliamentary outcome played out in each district is the long-dash line–the expected value of N’V based on actual NS. This corresponds to Equation 3. I also show the expectation based solely on assembly size (light dashed line); we already have no reason to expect this to be close in the UK, but maybe it would be in Canada, given that the actual nationwide NS was close to the SPM expectation, based on S (Equation 2).

Here is the UK, then Canada, 2019.

What we see here is interesting (OK, to me) and also a little unexpected. It is the UK in which the actual mean N’V is almost the same as the expectation from nationwide NS (i.e., Equation 3). We have actual mean N’V=2.485 compared to expected N’V from actual NS of 2.45; the ratio of actual to expected is 1.014. We can hardly ask for better than that! So, the nationwide party system (as measured by NS) itself may be well off the SPM expectation, but the vote fragmentation of the average district (N’V) closely tracks the logic that seems to stand behind Equation 3. Voters in the UK 2019 election tended to vote in the average district as if parties’ national viability mattered in their choice.

In Canada, on the other hand, even though national NS was very close to SPM expectation, the actual average district’s N’V (2.97) was really nowhere near either the expectation solely from S (the light dashed line, at 2.58) or the expectation from the actual NS (2.66). The average district was just so much more fragmented than it “should be” by either definition of how things ought to be! (The ratio of actual to that expected from Equation 3 is 1.116; the Equation 3 expectation is almost exactly the 25th percentile of the distribution.)

The Canadian outcome looks as if the exponent on actual NS in Equation 3 were around 0.64 instead of 0.5. Why? Who knows, but one implication is that the NDP (the third national party) performed far better in votes than the party’s contribution to NS implies that it should have. Such an overvaluing of a party’s “viability” would result if voters expected the party to do much better in terms of seats than it did. This is probably a good description of what happened, given that pre-election seat extrapolations implied the NDP would win many more seats than it did (and the Liberals fewer). The NDP also underperformed its polling aggregate in votes (while Liberals over-performed), but it held on to many more voters than it “should have” given its final seat-winning ability would imply. That is, the actual result in votes suggests a failure to update fully as the parties’ seat prospects shifted downward at the very end of the campaign. In fact, if we compare the final CBC poll tracker and seat projections to the ultimate result, we find that their actual votes dropped by 13.6% but their seats dropped by 31.7% (percent change, not percentage points!). In other words, this was just an unusually difficult context for voters to calibrate the expectations that Equation 3 implies they tend to make. (I am assuming the polls were “correct” at the time they were produced; however, if we assume they were wrong and the voters believed them anyway, I think the implications would be the same.)

It should be understood that the divergence from expectation is not caused by certain provinces, like Quebec, having a different party system due to a regional party, as some conventional expectations might point towards. While Quebec’s size is sufficient to exert a significant impact on the overall mean, it is not capable of shifting it from an expected 2.6 or 2.7 towards an observed 3.0! In fact, if we drop the Quebec observations, we still have a mean N’V=2.876 for the rest of Canada. The high fragmentation of the average district in the 2019 Canadian election is thus due to a Canada-wide phenomenon of voters voting for smaller parties at a greater rate than their actual viability would suggest they “should”. In other words, voters seem to have acted as if Trudeau’s promise that 2015 would be the last election under FPTP had actually come true! It did not, and the electoral system did its SPM-induced duty as it should, even if the voters were not playing along.

On the other hand, in the UK, voters played along just as they should. Their behavior produced a district-level mean vote fragmentation that logically fits the actual nationwide seat balance resulting from how their votes translated into seats under FPTP. There’s some solace in that, I suppose.

The Brexit Party

Just a quick add-on to my previous remarks on the UK 2019 election. Via @kiwiting on Twitter comes this example of a Brexit Party local leaflet.

Look closely and you might actually see the local candidate’s name! As I stress in the preceding post, I expect parties under FPTP (at least in parliamentary systems) to require a national presence in the party system in order normally to do well at the constituency level. That is a key insight of the Seat Product Model, and how it stands apart from “bottom-up” approaches that stress local district-level “coordination” as what drives a party system. But this is pretty extreme: the Brexit Party is not only a single-issue party, it is also a one-man band!

Even though this party at one point was polling above 20% (and won a plurality of the UK vote in the European Parliament elections), it was always hard for me to take the Brexit Party seriously. On the one hand, it certainly is a nationally focused party. On the other hand, the leader Nigel Farage made a decision not to contest any constituency, or to target even one seat somewhere that some candidate of the party might win. The process behind the SPM implies that voters respond to the “viability” of a smaller party, and tend to vote for it without too much regard for the viability of its candidate in their own district. But for that to work, it has to be viable–and preferably winning–somewhere. Not only did the Brexit Party not even try this, it pulled its candidates out of seats the Conservatives hold, while retaining candidates only in districts held by other parties. It is a bizarre strategy if the party was serious, and it is no wonder the party is on life support. Of course, they are going to get their one policy issue enacted (even if not as “hard” as they would like), precisely by not posing too big a risk to the incumbent government’s pursuit of a (manufactured) majority.

UK election 2019

The UK general election is almost here. At this point, it seems quite unlikely that the result will be anything other than a good old fashioned FPTP manufactured majority. Boris Johnson and his Conservatives will win a majority of seats, barring a surprise, despite under 45% of the votes, and will be able to pass their Brexit deal.

If one looks at the polling aggregate graph by the Economist, one might be tempted to conclude it was also a good old fashioned “Duvergerian” pattern at work. As recently as early October, before the election was legislated, the Conservatives were leading on about 33% of the votes, and three other parties ranged from 12% to 25%. Go back further, to June, and all for were in the 18–25% range (with Labour then on top, and the Brexit Party ahead of the Conservatives). Since the latter part of October, and especially since the campaign formally got underway, Conservatives and Labour have both taken off, at the expense of the LibDem and Brexit parties. Notably, the gap between the top two has been quite steady, at 8-10 percentage points. Unlike 2017, there is no evidence at all that Labour is closing the gap. Labour simply are hoovering up the non-Tory (and Remain or second-referendum) votes at the same time as Leave voters have realized there’s no point in voting for a single-issue Brexit Party when the Tories have a pretty “hard” Brexit deal already to go, if only they win a majority of seats.

So, on the one hand, a far more “normal” election for a FPTP-parliamentary system than seemed possible during the long parliamentary deadlock of the past year or more. Just like Duverger’s “law” predicts, right? Desertion of the third and fourth parties for the top two.

Only sort of. Let’s take the current polling estimates for the parties (and not forgetting to include the current 5% “other”, which I will treat as one party, given most of it is one party–the Scottish National Party). It results in an effective number of vote-earning parties of 3.05. That’s a little high for a supposedly classic two-party system! It is, however, lower than seen at any election from 1997 through 2015. In 2017, however, it was 2.89, which was the lowest since 1979. The top two would be combining for 78% of the votes, which is a little higher than most elections from 1974 (February, in a two-election year) through 2001. Even in 2017, hailed by many at the time as the return to two-party politics–albeit dubiously–had a combined top-two of just 82.4%. (It looks like a high figure only compared to 2005-2015, when it ranged from 65.1% to 67.6%.)

Of course, it is the seats that really matter. Seat projections based on election polls under FPTP are never easy. There are various ones out there, but I will go with YouGov‘s.* It has the Conservatives with a projected 359 seats, which is 55.2%, with Labour on 211 (32.5%). Taking all the parties (and here breaking the “Northern Ireland” bloc down a bit, as we know it will consist of more than one such party), we get an effective number of seat-winning parties around 2.4. That is even lower than 2015, driven mainly by the presence of an expected single-party majority.

[*Note: just after I posted this, YouGov posted an update of their projections. I am not going to revise the numbers here. The differences are small, though potentially politically significant. See my first comment below this post.]

The problem with the standard Duvergerian claims about FPTP is that they ignore assembly size: In a larger assembly, we should expect more parties, other things (like district magnitude and formula) equal. While we could argue over how much the expected results of the 2019 election correspond to the so-called law, I’d rather not. What is of interest to me is that the UK case continues its long-term defiance of the Seat Product Model (SPM), and that’s something that I can’t take lying down.

While the conventional wisdom would see 2017 and 2019 as some sort of return to normalcy, it’s actually a challenging case for me. From the SPM (which explains over 60% of the variation in party-system outcomes worldwide, including FPTP systems), we should expect:

Effective number of seat-winning parties: 2.95.

Seat share of the largest party: 0.445.

Effective number of vote-earning parties: 3.33.

The seat outcomes actually never have come very close to the expectations. As for votes, the 1987 election got it right, but was a terrible performer in terms of seats (effective N=2.17!). Taking all the indicators together, the 2010 election is about the closest to what should be “normal” for a FPTP system with such a large assembly: effective N on votes 3.72, seats 2.57, and largest seat share of 0.47. So why was that not finally the start of the kind of party system the country “should” have? I guess we need to blame Nick Clegg. Or David Cameron. (I’d rather blame the latter; he was the one, after all, who thought a Brexit referendum was a good enough idea to go ahead with it.) More to the point, voters’ reaction to Clegg and the LibDems entering a coalition and–gasp–making policy compromises. After which, voters reverted to supporting the big two in greater shares than they are supposed to. In other words, contingency and path dependency overcome the SPM in this case. I hate to admit it, but it’s the best I’ve got!

Speaking of the LibDems, they should have had an opportunity here. Labour has the most unpopular opposition leader in decades. (Deservedly so, but I digress.) And the best hope for stopping Brexit would be tactical voting to increase their chances to win seats where Labour is not best positioned to defeat a Tory. Yet, despite lots of constituency-level tactical voting advice being offered in this campaign, there’s little evidence the message is getting though.

There is tactical voting happening, but as Rob Johns points out in a short video, it is happening based on the national outcome and not on district level. Under the Duvergerian conventional wisdom, voters are alleged to think of their constituency, and vote tactically (strategically) to effect the local outcome. Yet in real life, only a relatively small minority of voters behave that way. That voters use a strategy based on who is best placed to defeat a party they do not like on the national level, instead of at the constituency level, is a point made forcefully by Richard Johnston in his book, The Canadian Party System. It is also the underlying logic of the SPM itself.

So from the standpoint of the SPM, what is surprising is not that there isn’t more tactical voting at the constituency level. It is that there does not remain (so to speak) a strong enough third party, such as the Liberal Democrats, to appear viable nationally so that voters would be willing to vote for its district candidates. Quite apart from the legacy of the coalition that I referred to above, the case for the LibDems as a viable counterweight probably was not helped by a tactical decision it made in this campaign. Its leader, Jo Swinson, declared that a LibDem government would revoke the Article 50 notification and cancel Brexit. Put aside the ridiculous idea that there would have been a LibDem government. If one had resulted from this election, it would have been on far less than 50% of the votes. So you have a government resting on a minority promising to go back on the majority voice of the 2016 referendum without even bothering with a second referendum. That seemed at the time like a dumb position for the party to take. Only recently has Swinson offered the message of what the LibDems could accomplish in a no-majority parliament. But it’s too late. There almost certainly won’t be such a parliament.

The UK really needs a national third party (and fourth…). Contrary to the Duvergerian conventional wisdom, the electoral system actually could sustain it; we would expect the party system to look more like Canada’s (which conforms to the SPM very well, both over time and, in terms of seats, in 2019). Given the larger assembly, the British party system should be even less two-party dominated than Canada’s actually is. It is by now rather apparent that the LibDems are not the third party the system needs to realize its full potential. Will one emerge? Alas, not soon enough to stop a hard Brexit from being implemented by a manufactured majority (for a leader who is pretty unpopular himself) while Labour gobbles up most of the opposition, but falls well short.

Attorneys General–institutions matter

Now that indictments have been announced against the (outgoing–dare I say?) Prime Minister of Israel, it is worth reviewing the institutional basis of the office of Attorney General in Israel.

I am seeing some casual takes on Twitter about why the US doesn’t have an Attorney General who takes a tougher line against law-breaking at the top of government. But the offices could hardly be more different. The US Attorney General is a cabinet appointee. The President picks who holds that position, subject only to Senate majority confirmation. Of course, Trump has had a highly compliant Senate majority throughout his presidency.

Trump could not have had occupants of the office that have been as awful for the rule of law as they have been, if the office were structured like Israel’s. So it is worth sketching how the process of appointing the Israeli Attorney General works. My source for this is Aviad Bakshi, Legal Advisers and the Government: Analysis and Recommendations, Kohelet Policy Forum, Policy Paper No. 10, February 2016.

a. There shall be formed a permanent selection committee that shall screen suitable candidates, one of which shall be appointed to the position by the government. The term of each committee shall be four years. 

b. The chairman shall be a retired justice of the Supreme Court who shall be appointed by the President (Chief Justice) of the Supreme Court upon the approval of the Minister of Justice, and the other members shall be: a retired Minister of Justice or retired Attorney General appointed by the government; a Knesset Member elected by the Constitution, Law and Justice Committee of the Knesset; a scholar elected by a forum comprising deans of law schools; an attorney elected by the Israel Bar Association. 

c. The AGI term duration shall be six years, with no extension, irrespective of the term of the government. 

d. The government may remove the AGI from his position due to specific reasons.… These reasons include, in addition to personal circumstances of the AGI, disagreements between the AGI and the government that prevent efficient cooperation. In such an event the selection committee shall convene to discuss the subject and shall submit its opinion to the government, in writing. However, the opinion of the committee is not binding, and the government may decide to remove the AGI contrary to the recommendation of the committee. The AGI shall have the right to a hearing before the government and before the committee. 

All of this makes for a reasonably independent office. Even if appointment and dismissal are still in the hands of the government, the screening and term provisions make it an arms-length relationship. The occupant of the post is obviously not a cabinet minister, as in the US, and is not a direct appointee of the head of government or the cabinet.

Worlds apart, institutionally.

And this is even before we get into the parliamentary vs. presidential distinction. A president is–for better or worse–meant to be hard to indict, let alone remove. That’s why the main tool against a potentially criminal executive in the US and many other presidential systems is lodged in the congress, through impeachment, and not in a state attorney. A prime minister in a parliamentary system, on the other hand, by definition has no presumption of a fixed term.

The normal way to get rid of a PM is, of course, a vote of no-confidence or the PM’s own party or coalition partners withdrawing support. But that’s the point–they are constitutionally not protected when the political winds, let alone the legals ones, turn against them.

In the broader institutional context of a parliamentary system, it is presumably much easier to take the step of also designing an independent Attorney General’s office that has the ability to indict a sitting head of government.

On the other hand, there is still no obvious way to remove Netanyahu from office any time soon, unless his own party rebels against him. Even though Trump’s own party will probably block the super-majority in the Senate needed to remove him from office*, the resolution of the case against Trump might happen considerably sooner than any resolution of Netanyahu’s case. Barring a rebellion by his current allies, Netanyahu may remain PM fore another 4-5 months, through a now-likely third election (since last April) and the post-election coalition bargaining process.

* Assuming the House majority impeaches him, which now looks all but inevitable.

Canada 2019: Results and a good night for the Seat Product Model

Add Canada 2019 to the set of plurality reversals. As anticipated before the election, the two largest parties each ended up with around one third of the vote. This is the lowest vote percentage for a governing party in Canada ever, I believe. The seats are somewhat less close than the CBC’s Poll Tracker estimated they would be. Instead of 133 seats to 123, the seats split 157 to 121. The Liberals are indeed that largest seat-winner, despite trailing the Conservatives in votes percentage, 34.4 – 33.1.

The NDP was either overestimated by polls or, more likely, suffered some late strategic defection. Instead of the near 19% of the vote in the final Poll Tracker, the party ended up with only 15.9%. More importantly, its seats stand at only 24, well below where estimates late in the campaign had them (per the CBC Poll Tracker).

As excepted the BQ had a good night, with 32 seats. The Greens picked up one new seat to augment the two they already held. The new seat is Fredricton, New Brunswick, whereas the other two are both on Vancouver Island.

In what I will call the two best pieces of news form the night (other than there being no single-party majority), the People’s Party crashed and burned, winning only 1.6% and seeing its leader lose his seat. That and the fact that Jody Wilson-Raybould, the former Attorney General who was kicked out of the Liberal caucus, retained her seat, Vancouver-Granville, as an independent.

 

Anomalous FPTP

I will certainly use this result often as a demonstration of how the first-past-the-post (FPTP) system can produce strange results.

Not only the plurality reversal for the top two, but the differential treatment of the next three parties, show anomalies of the sort that are inherent to FPTP. The BQ is only somewhat larger in votes than the Green Party, but will have more than ten times the number of seats. Under FPTP, it is good to have efficient regional distribution of support, and getting all your votes in one province, where you perform exceptionally well, is really efficient. The Greens, on the other hand, gained in almost all provinces, but it was good enough to add only one seat.

The NDP’s situation is one of a quite strong third party, but also inefficient regional distribution: 7.1% of the seats on 16% of the votes is a punishing result, but nothing at all unexpected, given the electoral system.

For that matter, the plurality reversal is itself a signal of the problem of inefficient vote distribution. The Conservative Party mostly gained votes where they could not help the party win seats, whereas the Liberals were much more successful winning close contests.

In his victory speech, PM Justin Trudeau was bold enough to use the M-word (mandate), but this most certainly is not one. For the moment, he can be pretty happy he broke that promise on 2015 being the last FPTP election. His party remains in position to form the government, and has a substantial seat bonus. The advantage ratio (%seats/%seats) is 1.40. (How does that compare with past elections? Click to see.)

Canada would be well served by at least some degree of proportionality. In fact, so would the Conservatives, given their tendency to run up margins where they are already strong. (Note that they are only barely over-represented in seats, with 35.8%.) However, this result is unlikely to advance the cause of reform, as the Liberals’ position–46% of the seats and a 36-seat (more than ten percentage point) edge over the runner-up–looks quite solid.

The other reason the country could really use electoral reform is the map. There is no Liberal red to be seen from central Ontario westward, except around Vancouver (and two northern territories). The party lost some of its ministers’ reelection bids in Alberta and Saskatchewan. With even a minimally proportional system, the situation of a governing party without members of its caucus in nearly every province would not happen.

While a PR system would be beneficial, the country is stuck with FPTP at least for now. So how did this result compare to what we should expect from the electoral system actually in use?

 

The Seat Product Model and the outcome

The Seat Product Model (SPM) performed better than the CBC Poll Tracker’s seat estimator. For an assembly of 338 and districts with magnitude of 1, we should expect the largest party to have, on average, 48.3% of the seats, which would be 163 seats. So the actual result (46.4%) misses the expectation by 6 seats, or 1.78 percentage points (compared to the a 20-plus, or 6 percentage point, miss by the Poll Tracker).

Of course, the SPM has one advantage in its favor: it does not “know” that the seat-winning party would have under 33.3% of the vote, whereas the Poll Tracker must work with this expectation (and, as it turned out, reality). In fact, when a party wins 48.3% of the seats, the formulas of SPM (collected in Table 9.2 of Votes from Seats) expect it to have won 43.3% of the votes. (Theoretically, we do not expect the SPM to perform as well with votes as with the seats that are at its core; but in Votes from Seats, we show that, on average, it performs about equally as well with both.) The Liberals underperformed this expectation by more than ten percentage points! The voters genuinely voted for something their electoral system could not deliver, even if the system indeed delivered what should be expected solely on institutional grounds.

In terms of the effective number of seat-winning parties (NS), the actual result was 2.79. This is slightly higher than the SPM expectation, which is 2.64. The miss is minor, with a result only 1.057 times expectation.

On the other hand, the effective number of vote-earning parties (NV) was 3.79. The SPM expects 3.04. Let me pause and emphasize that point. Because Canada uses FPTP in a 338-seat assembly, we should expect the votes to resemble a “three-party system” and not the two-party system that all the conventional “Duvergerian” wisdom claims. If we calculated expected Nbased on the known NS=2.79, we would expect NV=3.17. However, neither the SPM nor Duverger’s “law” expects that the largest party nationwide should have only around a third of the votes. That is the really remarkable thing about this outcome.

 

The district level

At the district level, there were numerous non-Duvergerian outcomes, as would be expected with the known distribution of nationwide votes among parties. According to an extension of the SPM (in a forthcoming book chapter), we should expect the effective number of vote-earning parties at the average district (N’V) to be 1.59 times the square root of the nationwide NS. That would be 2.66. It will be a while before I am able to calculate what it actually was, but it would not surprise me if it was a fair bit higher than that. But, again, let me pause and say that a Duvergerian two-party competition at the district level is NOT to be expected, given both the nationwide electoral system and the actual aggregate seat outcome. (If we went off expected nationwide NS, instead of the known outcome, the district-level mean still would be predicted to be 2.58; see Chapter 10 of Votes from Seats.) Canadian elections of the past several decades have tended to conform closely to this expectation for district-level N’V.

The country does not tend to have two-party contests at district level, nor should it (when we have the Seat Product Model to guide our expectations). In other words, voters do not tend to vote in order to “coordinate” their district outcome around the two most viable candidates. They tend to vote more towards their expectation (or desire) about what the nationwide parliamentary outcome will be. This is so even in Quebec where, in this election, many Francophone voters returned to the regional party, the Bloc Québécois. Quebec has numerous district contests that feature three or four viable parties.

So if your image of Canada’s party system is that in Quebec districts it is BQ vs. Liberal, with other parties barely registering, while elsewhere it is Liberal vs. Conservative, except where it is one of those vs. NDP, it is well past time to update. Canada does not have nationwide multiparty politics because it has separate regional two-party systems (as many folks, even political scientists, seem to believe). Canada has district-level multipartism because it has nationwide multipartism. (See Richard Johnston’s outstanding book for a rich “analytic history” that supports this point.) And this may be even more true in the one province in which there is (again) a strong regional party. Consider the aggregate provincial outcome in terms of vote percentages in Quebec: Liberal 34.2% (slightly higher than nationwide), BQ 32.5%, Conservative 16.0%, NDP 10.7%, Green 4.5%. This gives a provincial-level NV of 3.82, a bit higher than nationwide.

I will offer a few striking examples of multiparty contests at district level, just to illustrate the point. The new Green Party MP from Fredericton, Jenica Atwin, won 32.8% of the vote. The Conservative had 31.1%, the Liberal 27.3%, and the NDP 6.0%. There may indeed have been strategic voting happening here, with some NDP voters–the party had 9.9% in 2015–switching to Atwin to stop the Conservative (and perhaps some who don’t like the Greens boosting the Liberal). But the outcome here is N’V=3.53!

The change from 2015 in Fredericton is really striking, as the Liberal candidate was an incumbent who had won 49.3% in 2015 (against 28.4% for the Conservative, meaning this party gained a little here in 2019). Clearly many Liberals defected from their party to the Green following that party’s success, including a local win, in the recent provincial election. In doing so they only narrowly avoided the serious “coordination failure” that would have been a Conservative win.

Another Green MP, the reelected Paul Manly in Nanaimo-Ladysmith, won 34.5%. This was actually a pretty clear victory despite being barely over a third of the vote; Manly had been elected in a by-election this past May with 37.3%. The runner-up Conservative had only 25.9% in the general election contest, the NDP 23.7%, Liberal 13.6%. N’V=3.83!

Wilson-Raybould’s win in Vancouver-Granville as an independent was also with under a third of the vote. She had 32.3%, beating the Liberal’s candidate (26.6%) and the Conservatives’ (22.1%). The NDP candidate had 13.1%. The Greens, who tried to recruit Wilson-Raybould to be their candidate, put up their own against her, who got 5.0%. It should be noted that the NDP candidate in this riding last time won 26.9%, so it would appear there was ample strategic voting here in Wilson-Raybould’s favor. (She won 43.9% as the Liberal candidate in 2015.) The Green voters, on the other hand, did not seem to warm to their near-candidate; the party’s actual candidate did better in this district in 2019 than in 2015 (when the party got 3.1%).

One of my favorite cases is Sherbrooke, in Quebec. The winner was Liberal Elisabeth Briere with 29.3%, edging out an NDP incumbent who won 28.3% in this election. He had won the seat with 37.3% in 2015. Close behind in this year’s contest was the BQ candidate who had 25.8%. Following behind them was a Conservative (10.7%), and Green (4.5%). N’V=4.06!! The Liberals won this by basically standing still in vote share, having lost this district by a wide margin in 2015 when their candidate had 29.8%.

A few interesting tidbits from candidate backgrounds. Bernier’s defeat in his own riding of Beauce was at the hands of a dairy farmer, Richard Lehoux. The Conservatives recruited him because of Bernier’s opposition to supply management policies in the dairy sector. (Info found in the CBC’s Live Blog.) Lehoux won only 38.6% of the vote, but it was sufficient to beat Bernier rather badly, as the latter (elected as a Conservative in 2015 and previously) had just 28.4%.

There were several mayors recruited to run, including a case in Quebec where the Conservatives hoped the candidate’s local popularity would overcome the party leader’s unpopularity. (The specific case was Trois-Rivières; the Conservative finished a close third in a riding the BQ candidate won with 28.5%.) There was also an Olympic medal-winning kayaker, Adam van Koeverden, whom the Liberals recruited in Milton (in Toronto, Ontario) to run against the Conservative Deputy Leader, Lisa Raitt. He defeated her–easily, winning 51.4% to her 36.5%. Presumably his celebrity (and perhaps his local roots, which he made a point to emphasize in an interview after his victory was confirmed) helped him win despite a nationwide swing against the Liberals and in favor of the Conservatives. (She had won 54.4% in 2015.) In other words, while I may emphasize that district politics under FPTP in a parliamentary system is mostly national politics, there is still plenty of room for local and personal factors to matter.

 

What it means for the near term

As to the shape of the government to result, it should be a reasonably stable minority government, although it may not last full term. It can form legislative majorities with either the BQ or the NDP, and thus need not be tied to either one in a coalition. And the NDP certainly is not strong enough to demand a coalition (even if it wanted to try). Nor is it likely strong enough to demand action on electoral reform, even if an election in which two thirds of the voters voted against the governing party, and various other aspects of the outcome can be seen as anomalous, suggests that reform is needed more than ever.

Rediscovering an old publication

Believe it or not, I just noticed an article by me, published in an academic journal, has been missing from my CV for over twenty years! In fact, I had to search on the web to find it.

“The Jenkins Paradox: A complex system, yet only a timid step towards PR,” Representation 36:2 (1999).

I thought of it when wanting to link to it in my previous note about the Quebec proposal. And then I could not find the link because it was not on my CV (or website)!

My personal favorite passage from my forgotten article, after commenting on the Jenkins Commission proposal for the UK and its flaws:

It would seem, therefore, preferable to use MMP with a small percentage of PR seats, or MMP with multiple regional PR compensation regions, or straightforward alternative vote, but not some combination of all three!

The other thing I realized in searching for this is just how dreadfully bad the interface of the Taylor and Francis journals website is.

Presidentialization

[As long as I made a tweet storm in response to, first, a tweet by Ezra Klein, and then a question by Nicholas Smith, I might as well turn it into a blog post. Process made easy by the Spooler app.]

First the preliminaries and context that got it all going…

Great question! A short thread on “personalization” and “presidentialization” of political parties…

Matthew Shugart

Yes, existing to promote and protect the leader who was separately elected to the country’s top office is the very definition of a presidentialized party (Samuels and Shugart, 2010).

The US, during this presidency, finally has become a more normal presidential democracy. https://twitter.com/ezraklein/status/1177244610931781634 

Ezra Klein

@ezraklein

There is nothing “conservative” about the Republican Party we’re seeing in these hearings. It’s a party that exists to promote and protect Donald Trump.

Nicholas Smith @_SmithNicholas_

What step in the process is it when parties are being created and collapsing solely around a party leader (e.g. IL, UA, FR)?

_____________________ Now for the thread (lightly edited)…

Personalization can happen under any type of democratic political system. In brief, it means that the election turns on the assessment of the leader, rather than on a platform, issues, ideology, or long-run party ID.

Presidentialization is something more. It is the leader becoming the de-facto principal over the party once elected (or even once nominated). It is the selection of an executive candidate who may not share the values of the party because the party needs someone who can win.

This can happen in parliamentary systems, but it is much less likely. It can be avoided in presidential systems, but it is harder to avoid. Why? Parl party leader, including PM (normally) remains accountable to the party (in the legislature). A president, by definition, does not. Presidents and legislative parties, by definition, can have distinct electoral coalitions. But the more they approach being identical, the more likely it is that the party falls in line behind the president, even if the pres. is taking the party places it otherwise would not go. That is, the party legislators’ fates become tied to the fate of the executive. In principle, it remains the reverse in a parliamentary system.

HOWEVER, there are exceptions. Corbyn, maybe Johnson, look pretty “presidentialized” as party leaders (and the latter as PM). There is obviously some degree of separation that has developed in these parties lately, due in part to unusual leadership selection rules and circumstances. However, of course, the voting remains fused–separation can’t extend to how voters vote for party & executive. The original tweet I am responding to mentioned cases of newly formed parties, whereas above I have referred mostly to established parties. Let’s take the mentioned examples

Israel. Blue & White can be seen as a personal vehicle for Gantz to be PM. But it is an alliance. The internal partners are also personal vehicles (Lapid, etc.). A key point is Gantz needed pre-electoral and now post-electoral allies if he is to head the government. Gantz, or any head of a parliamentary party/alliance, can’t present himself to the electorate separately from the party system, as is possible in a presidential or semi-presidential system. He has to get the nomination of an existing party, or form a new one, and the party must win seats.

(Semi-presidential = a popularly elected presidency AND a premier who depends on the confidence of a majority of the assembly. They vary a lot in the constitutional powers of the presidency.)

France. Semi-presidential. Macron formed an entirely new party, totally beholden to him. And benefited from the fact that assembly elections come AFTER presidential. It is sort of presidentialization on steroids!

See (and my earlier posts linked within): (fruitsandvotes.wordpress.com/2017/06/18/fra…)

(Because it’s semi-presidential and not “pure” presidential, he did need his party to do well in assembly elections, in order to be able to choose an ally as PM. In a pure presidential system, one can control the executive without a party, though one would rather have allies in congress, obviously.)

Ukraine. Also semi-presidential. I also wrote about Servant of the President. Oops, I mean to say Servant of the People, here. (fruitsandvotes.wordpress.com/2019/07/21/ukr…)

Israel 2019b, compared to 2019a

Here, following up on the earlier discussion of post-election bargaining scenarios, I want to compare Israel’s two elections of 2019 on several statistical measures. The 2019b (September) results are not quite official yet, but are very unlikely to change other than in the smallest of voting detail.

The table below compares the votes for Netanyahu’s “Bibi bloc” of right-wing and Haredi parties, by various definitions, as well as the indicators of fragmentation: effective number of parties by seats and votes, total number of lists with seats, and the seats won by the largest list. For each measure, there is a comparison of change from April to September. The final three columns refer to output of the Seat Product Model (SPM) for the indicators of fragmentation–what is expected from the model (given an assembly size of 120 and district magnitude also of 120), and ratios of the actual indicators to the expectation.

Measure April Sept change SPM expected Ratio, April Ratio, Sept.
Bibi bloc (percent votes) 48.7 44.5 -4.2
… plus YB 52.7 51.5 -1.2
… plus Otzma 46.4
… plus YB & Otzma 52.7 53.3 0.6
Effective N, seats 5.24 5.67 0.43 4.93 1.06 1.15
Effective N, votes 6.33 6.11 -0.22 5.23 1.21 1.17
No. of lists with 1 or more seats 11 9 -2 11 1.00 0.82
Seats for largest list 35 33 -2 36 0.97 0.92

The scale of the defeat for the core Bibi bloc is clear. Already in April, these parties had less than 50% of the votes, at 48.7%, which is why they won only 60 seats under Israel’s proportional system. If we include Yisrael Beiteinu in the total Bibi bloc, we get 52.7% (which is why this larger definition of the bloc had 65 seats). As I have explained already–both before and after the most recent election–we should not count YB in the bloc, particularly since it was this party’s actions that precipitated the early elections of 2019–yes, both of them.

In the second election of 2019, this Bibi bloc fell to 44.5% of the vote, a drop of 4.2 percentage points. If we include YB, they do have a narrow majority of votes (51.2%), but we should not include them. However, we probably should include Otzma Yehudit, given that it was part of the Union of Right Wing Parties in April, and probably would have been invited to join a coalition had it cleared the threshold in the September election. But still this is short of a voting majority without YB, at 46.4% (which would mean a loss of 2.3 percentage points off the April showing of 48.7).

For a baseline, consider that the Bibi bloc had 48.4% in 2015, or 53.5% including YB (which was without doubt part of the bloc at that time–their staying out of the coalition initially in 2015 was a surprise). Note that, leaving out YB, they were already below majority voter support in 2015, but had managed 61 seats. The reason they gained ever so slightly in votes in April, yet got only 60 seats, was all the wasted votes for New Right (3.22%), which did not clear the threshold in the April, 2019, election.*

If we include both Otzma and YB in the 2019b election, it looks like a very small gain for the wider bloc. But we should not do this because some of YB’s increased votes probably came from Blue and White or other parties not in the right, due to YB’s promise not to return to a Likud-led government unless it was a “unity” government with Blue and White.

On the fragmentation indicators, the effective number of seat-winning parties went up, from 5.24 to 5.67, despite the drop of the total number of lists winning seats, from 11 to 9. The increase in the effective number is due to the smaller size of the largest party in the more recent election, 33 seats (Blue and White) vs. 35 (tie between Blue & White and Likud).

The effective number of vote-earning parties came down somewhat, from 6.33 to 6.11. None of these measures is much different than what we should expect under the SPM, although the raw number of represented lists this time is actually smaller than expected, while the effective number of seat winning parties was closer to the expectation in April than now.

We should expect the largest party, given this electoral system, to have 30.2% of the seats, which out of 120 works out to 36 (rounded down). The election pretty much nailed that in April, but this election saw a return to a smaller than expected plurality party.

So, strictly from the SPM, this was a slightly less “normal” election than 2019a, although not too far off. From the standpoint of the usual pattern with a “b” election (a second one within a year), it was, as I anticipated, a little unusual. Typically, the effective numbers go down and the size of the largest up. Israel went the opposite way between April and September, and thus government formation still will not be easy.


* We could go back and include Yachad (of which Otzma Yehudit was a part) in the 2015 count, which would bring it to 51.3%, but at the time I do not recall their being taken seriously as part of the bloc. Doing so, of course, increases the scale of the loss of voter support already as of the first election of 2019.

Israel is about to have a very unusual ‘b’ election

Israel is about to hold its second election of 2019, and it will be unusual, relative to other cases of a second election within a year elsewhere. While the number of lists winning seats is likely to go down, other indicators of fragmentation are likely to go up.

Using the National Level Party Systems Dataset (Struthers, Li, and Shugart, 2018), I performed calculations to find out how the standard indicators of party-system fragmentation change from a first election that fails to produce a “stable” government or any government at all, leading to a second election. I looked at all cases in the dataset in which two elections were held in the same Gregorian calendar year, plus all cases where an election is in the second half of a year and followed by another in the first half of the next year. The first table below gives the full list, including the first and second election in each sequence. In one case in the dataset (Greece, 1989-1990) the second election was followed by yet another within a year, indicated by a “3” in the final column. Note that a country’s data sequence begins in the early post-WWII era or when a country democratized and ends in 2016, so any cases outside that timeframe are not included.

country year date mo within_yr_seq
Denmark 1953 4/21/53 4 1
Denmark 1953 9/22/53 9 2
Denmark 1987 9/8/87 9 1
Denmark 1988 5/10/88 5 2
Greece 1989 6/18/89 6 1
Greece 1989 11/5/89 11 2
Greece 1990 4/8/90 4 3
Greece 2012 5/6/12 5 1
Greece 2012 6/17/12 6 2
Greece 2015 1/25/15 1 1
Greece 2015 9/20/15 9 2
Iceland 1959 6/28/59 6 1
Iceland 1959 10/25/59 10 2
Ireland 1982 2/18/82 2 1
Ireland 1982 11/24/82 11 2
Japan 1952 10/1/52 10 1
Japan 1953 4/19/53 4 2
Japan 1979 10/7/79 10 1
Japan 1980 6/22/80 6 2
Moldova 2009 4/5/09 4 1
Moldova 2009 7/29/09 7 2
Spain 2015 12/20/15 12 1
Spain 2016 6/26/16 6 2
Sri Lanka 1960 3/19/60 3 1
Sri Lanka 1960 7/20/60 7 2
St. Lucia 1987 4/6/87 4 1
St. Lucia 1987 4/30/87 4 2
Thailand 1992 3/22/92 3 1
Thailand 1992 9/13/92 9 2
Turkey 2015 6/7/15 6 1
Turkey 2015 11/1/15 11 2
UK 1974 2/28/74 2 1
UK 1974 10/10/74 10 2

The list contains 17 cases of an election within twelve months of the preceding one. Not a large sample; fortunately, this sort of thing does not happen very often. (There are 1,025 elections in the sample.)

If elites and/or voters “learn” from the experience of bargaining failure or lack of stability from the first election in such a sequence, we would expect the second to be less fragmented. We can test this by looking at mean differences between the second election and the first. The indicators I have are the number of parties (or lists, more precisely, counting an independent as a “list” of one) that win at least one seat (NS0), the effective number of seat-winning lists (NS), the effective number of vote-earning lists (NV), the seat share of the largest party (s1), and the vote share of the largest party (v1). The first three should go down if there’s an adaptation occurring, while the second two should go up (i.e., the largest party gets bigger).

Here is what we see from the results, reporting the mean differences:

NS0: –0.215

NS: –0.098

NV: –0.469

s1: +0.010

v1: +0.0035

In terms of raw direction, all are as expected. On the other hand, the number of lists winning seats hardly budges (recall that the first number is the actual number, not “effective”), and the effective number on seats changes much less than the one on votes. The implication is that fewer votes are wasted in the second election, as we would expect. On the other hand, the seat share of the largest party–the single most important quantity because it determines whether there is a single-party majority and if not, how far from majority it is–rises by a very small amount, on average. That is partly due to most of these systems being proportional, so large shifts should be unusual. The complete list of elections and their indicators is provided in an appendix below.

As far as statistical significance is concerned, only in NV and v1 is the difference significant (NV at p<0.03; v1 at p<0.10), when comparing these “second” elections to all others. (This is not meant to be a sophisticated test; I am not comparing to a country baseline as I really should.)

We might expect that the first election in such a sequence is anomalously fragmented, hence the need for a second election to calm things down once again. That is also supported, for NV and v1 again, but also, crucially, for s1.

Now, how might the Israeli second election of 2019 compare? We can use the polling average from Knesset Jeremy (using the poll of polls from three weeks before the actual election), and compare to the actual results of 2019a (the first election in the sequence) and the previous election (2015). Also included in the Seat Product Model expectation.

measure 2019b (poll avg) 2019a actual diff 2015 diff SPM expected
NS0 9 11 –2 10 1 11
NS 6.04 5.24 0.801 6.94 –1.70 4.93
NV ? 6.33 ? 7.71 –1.38 5.24
s1 0.258 0.292 -0.034 0.25 0.042 0.3
v1 ? 0.2646 ? 0.234 0.031 0.289

For the number of lists that look likely to clear the threshold, we have the direction expected: currently there are 9 likely to win seats, compared to 11 in April. In turn, the April figure was one seat-winning list higher than in 2015. However, in terms of both NS and s1, the case is anomalous. All indications are that the largest party will be smaller than it was in April, which also will drive up the effective number. Moreover, these measures in April were less fragmented than they had been in 2015; that is, the first election of the 2019 sequence was not unusually fragmented. Quite the contrary; I called it a “normal” election at the time for a reason.

So the Israeli sequence of two elections in 2019 is unusual indeed.


Appendix

Below are two tables. One has all the “second” elections, and changes in the various measures. The second has all “first” elections. In each case, the comparison is just to the immediately preceding election (not to all other elections), so we can see how much short-term fluctuations were affecting the process in each sequence.

Elections ocurring within one year of previous, compared to previous results
country year mo diff_Ns0 diff_Ns diff_Nv diff_s1 diff_v1
Denmark 1953 9 1 -0.2199998 -0.1000001 0.014 0.009
Denmark 1988 5 -1 0.0100002 0 0.005 0.005
Greece 1989 11 1 -0.0800002 -0.1700001 0 0
Greece 1990 4 5 0.05 0.0700002 0.005 0.017
Greece 2012 6 0 -1.07 -3.75 0.07 0.108
Greece 2015 9 1 0.1490002 -1.19 -0.014 -0.008
Iceland 1959 10 0 0.24 . 0 .
Ireland 1982 11 -1 -0.01 0.03 0 0
Japan 1953 4 . 0.8099999 0.8999999 -0.088 -0.091
Japan 1980 6 -8 -0.3999999 -0.24 0.074 0.033
Moldova 2009 7 1 0.8699999 0.27 0 -0.048
Spain 2016 6 -1 -0.3700004 -0.7999997 0.04 0.043
Sri Lanka 1960 7 . -1.22 -2.52 0.166 0.032
St. Lucia 1987 4 0 0 -0.1099999 0 0.007
Thailand 1992 9 0 -0.0999999 0.0999999 0 0.017
Turkey 2015 11 . -0.322 0.03 -0.126 -0.089
UK 1974 10 -1 -0.01 -0.02 0.028 0.021
Election that is the first in a series of two within a year, compared to preceding election
country year mo diff_Ns0 diff_Ns diff_Nv diff_s1 diff_v1
Denmark 1953 4 0 -0.1300001 -0.0900002 0.013 0.008
Denmark 1987 9 0 0.27 0.5799999 -0.009 -0.023
Greece 1989 6 1 0.26 0.1400001 -0.044 -0.006
Greece 2012 5 2 2.24 5.79 -0.173 -0.25
Greece 2015 1 0 -0.6700001 -0.77 0.067 0.066
Iceland 1959 6 0 -0.28 . 0.035 .
Ireland 1982 2 -2 -0.05 -0.1699998 -0.039 0.009
Japan 1952 10 . . . . .
Japan 1979 10 -1 0.1199999 -0.2199998 -0.002 0.027
Moldova 2009 4 1 0.1400001 0.1600001 -0.079 0.035
Spain 2015 12 -3 1.93 3.23 -0.18 -0.159
Sri Lanka 1960 3 . 1.456 2.26 -0.206 -0.043
St. Lucia 1987 4 -1 0.55 -0.0800002 -0.295 -0.049
Thailand 1992 3 . . . . .
Turkey 2015 6 . 0.4320002 0 0.002 0.005
UK 1974 2 2 0.1900001 0.6900001 -0.05 -0.077