Italian Elections and Politics 2022 - Our Time to Schlein

[palandio]

A new poll by Scenari Politici - Winpoll commissioned by Il Sole 24 Ore puts Italia Viva at the relatively high mark of 6.4%. Interesting are the voter flows:
1.8% from More Europe
1.6% from the PD
1.5% from Forza Italia
0.4% from non-voters
0.3% from Brothers of Italy
0.3% from Five Stars
0.3% from the Lega
0.2% from others.

I'm always wary of the classical Italian polls that pretend accuracy up to the first digit behind the dot, only to fail miserably on election day, but interesting, if true.

[palandio]

The M5S tried to enter a stealth electoral path with PD by just not running candidates at all (with the assumption that whatever voter base they have left at this point will back PD over Lega), but M5S activists, in their infinite wisdom, decided otherwise. So now the M5S is running token lists in a race they know to be beyond hopeless, and probably just hurting their coalition partner in the process.

Also significant, but not surprising, that among the seven coalitions running there are center-left, center-right, M5S, some anti-vaxers and three far-left coalitions (Power to the People, The Other Emilia-Romagna and the Communist Party). Three out of seven. No competition on the right, if you don't count the anti-vaxers.

[palandio]

One thing that's true and should be noted is that Bonaccini seems to have benefited from a significant personal vote. The lists that supported him only got about 48% of the vote, but he got almost 52%. Some of it is tactical voting by M5Sers (the M5S candidate got 1.5 points less than the M5S list) but many right-wingers also seem to have crossed over to support him. He has a solid record as ER President, with the region having one of the highest growth and lowest unemployment of the country. So in the end local factors probably mattered just as much as the anti-Salvini mobilization. It's kind of a John Bel Edwards situation where a party tried to nationalize the race to beat a popular incumbent and failed.

I don't completely agree because I think that you missed an important detail: The total number of governor votes was much higher than the total number of list votes.

Total votes (lists): 2.162.216
Total votes (governor): 2.325.497 (+163.281)

Total votes (left-wing lists): 1.040.482
Total votes (Bonaccini): 1.195.742 (+155.260)

Total votes (right-wing lists): 981.787
Total votes (Borgonzini): 1.014.672 (+32.885)

Total votes (M5S list): 102.595
Total votes (Benini): 80.823 (-21.772)

You can see that the number of governor votes was significantly higher than the number of list votes, a phenomenon that is widely seen in Italian regional and local elections and often forgotten. Borgonzoni actually got 33k votes more than her supporting lists. Bonaccini on the other hand got 155k more. I think that many of these 155k voters are not convinced left-wingers, but voted against the right (and thought that Bonaccini was an ok guy). I doubt that many of them would vote for a right-wing list. On the other hand mobilization for Borgonzoni and the right was much more via the national Lega brand. If you are a Lega supporter and want to through out the left it does not make sense to vote for Bonaccini.
Hence I do not think that there was any significant right-wing-Bonaccini crossover vote

[palandio]

Yes, I'm aware that many people only vote for the presidency ballot and not for the lists. And sure, a significant part of it is low-propensity left-leaning voters bothering to cross out Bonaccini's name but not a list logo, but I don't think it's crazy to suggest that a few of them are people who are nationally right-wing but thought Bonaccini was the better candidate locally. It makes perfect sense given everything else we know about how this campaign went down (Borgonzoni being such a weak candidate that Salvini basically locked her away to run her campaign himself). If we want to measure the "real" national political mood in Emilia, the truth is probably somewhere in between the list and personal vote.

It's not crazy to suggest this, in fact I like that you are questioning some popular beliefs. But in this case a quantitative analysis of the election data doesn't seem to back it up. And on top keep in mind that Borgonzoni outperformed the right-wing lists by 33k votes. For every voter that split his ticket between a right-wing list and Bonaccini there must be an additional voter who voted for weak candidate Borgonzoni without voting for a right-wing list.

A quantitative analysis on the provincial level shows that the Bonaccini surplus over the left-wing lists (ER overall: 3.3%) correlates positively with the left-wing results themselves. (The exception being traditionally left-wing Reggio Emilia province where the surplus was relatively low.) In the provincial capitals (which nowadays mostly lean to the left of their provinces) the Bonaccini surplus was higher than in the rest of the respective provinces.
Another parameter that can be used is the turnout difference from the 2019 EU elections to the 2020 ER regional elections. Overall turnout went from 67.3% to 67.7%. But there were significant local differences. Rise in turnout was correlated to left-wing strength and a Bonaccini surplus. (Turnout in Reggio Emilia province fell from 69.2% to 68.0% which might explain the weak Bonaccini surplus.)
The most extreme example in all of this is Bologna city, a left-wing stronghold where turnout went from 63.3% to 69.8% and where the Bonaccini surplus was 4.3%.

Measuring the "real" national mood is a different and very complex question.

Anyway, here's some hard data to draw your own conclusion from. Election results of the 2018 parliamentary, 2019 European, and 2020 regional elections in both regions (president and list for ER; in Calabria there's only one vote). I included LeU on the "left" side but not the more extreme far-left lists.

[...]

Left-right swing from the 2018 elections:
- ER (president) +5.6
- ER (list) +0.5
- Calabria -12.9

Left-right swing from the 2019 elections:
- ER (president) +12.5
- ER (list) +7.4
- Calabria -2.8

The Calabria results are scary, but they can probably be safely dismissed given the low turnout and the peculiarity of local issues. Calabria has kicked out every incumbent government since the Second Republic and seems to largely just be disaffected from politics in general (see also M5S's results in 2018). The ER numbers are more meaningful, although again the question is which numbers do you use. Either way, there's definitely been an upswing from the EU elections, but it's unclear whether there has been one from the 2018 election (where, as a reminder, the right was 11 points ahead of the left).

It's difficult to talk about upswing or not compared to the 2018 election because of the complete Five Stars meltdown. The (simplified) common theory is that the right-leaning part of the former Five Stars voters mostly voted Lega already in the 2019 EU elections and that at least in ER the remaining left-leaning part mostly voted for the left. The problem for the left being that even if they can build the same voter coalition that voted for Bonaccini on the national level, it probably won't be enough, as long as the right-wing coalition sits at a comfortable >50%.

[palandio]

So how do those results compare with previously?

Italy during the Second Republic has been leaning slightly to the right.

In the most recent past Italy has leaned strongly to the right (e.g. 2019 EP election). (And the Right has become more right-wing than in the past.)

It's difficult to extrapolate from yesterday's results to the national level, but it seems to me that the Right still is clearly favored on the national level, not as much as during the last two years or so, but clearly favored. On the other hand the left has been able to consolidate, mostly at the expense of M5S (and to a minor degree FI?), but outside of Emilia-Romagna and Toscana it seems to still trail the right. Large parts of the South have been swingy during the Second Republic. The results in Campania and Apulia show that under the right circumstances the left-wing potential is still there in the South, but it won't necessarily translate to the national level.

When talking about where regions are going to lean, we should take into account not only one side (Lega+FdI), but all sides. The Left is deeply rooted in parts of the Center-North and socio-economic trends are favorable to it in parts of the Center-North. Therefore the Lega won't magically take it all over with no resistance. On the other hand the South will not just vote for the Left because there isn't anyone else left, the Left will have to actively win back the South.

[palandio]

I went check some precincts and the main culprit is that muh city centre liberals* voted big for Toti.

The success of Zaia's and Toti's personal lists, the Lega's relative(!) weakness and the assumption that FdI can't completely fill out the center-right space in the South make me think:

Is a tandem really a safe option for Lega and FdI or would they gain a lot by including FI or some "moderate" successor in their alliance?

[palandio]

[...]
assumption that FdI can't completely fill out the center-right space in the South
[...]

I don't know if that assumption is valid to be honest.
People forget that AN took 16% of the votes in 1996. And Fini had identified as a neofascist until two years and a half before. FdI can do even better in my opinion.

And in parts of the South AN was even stronger than this. I know that a part of the Southern FI and CCD-CDU vote was not ideological, that today is a different time and that maybe FdI is more palatable than AN was in 1996. But still, in 1996 FI took over 20% in most of the South (in Sicily 32%), not even accounting for the CCD-CDU. I doubt that a party that is basically the successor of post-fascist AN without the moderate wing, can inherit (Southern) AN, FI and CCD-CDU at the same time.

[palandio]

The election results were not "good" for the left.

Tuscany Right Increase 10.1%
Apulian  Right Increase 6.2%

Whom are you even replying to?

[palandio]

LIGURIA
Liguria I think will come back to its status as Red Region Lite. It's actually pretty urban. Genoa = 1/3 of Liguria.
Toti won by so much because his list slamdunked in the city of Genoa for I don't know what reason. It was unexpected.
La Spezia 2019: Right 45 - Left 35
Savona 2019: Right 41.5 - Left 38
Genoa 2019: Left 39.5 - Right 39

La Spezia 2020: Toti +4
Savona 2020: Toti +1
At which point you would expect Genoa to be say Sansa +5 or so and instead it was Toti +8. Meh.

I think Genoa will snap back to pretty left-leaning at the next national election.

Regarding the comparison of results in Genoa (city) 2019 and 2020:
Toti overperformed the 2019 Lega+FdI+FI numbers by 12.6 points. Overperformance is actually relatively evenly distributed throughout the city, but looking the results by unità urbanistica you can still see some patterns.

The strongest overperformances:

Bavari	+18.0	village in the East
Puggia	+16.8	bourgeois quarter on the Eastern coast
Apparizione	+16.6	village in the East
Foce	+15.9	relatively bourgeois, east of center
Lido	+15.9	bourgeois quarter on the Eastern coast
S. Giuliano	+15.5	bourgeois quarter on the Eastern coast
Quartara	+15.5	bourgeois quarter on the Eastern coast
Quezzi	+15.2	high-density working class periphery around old village
Albaro	+15.2	bourgeois quarter on the Eastern coast

The weakest overperformances:

Morego	+2.5	peripheral village near the Polcevera valley
Campi	+6.0	industrial area near the Polcevera mouth
Borzoli Est	+7.4	village turned working class periphery on a hill between Western coast and Polcevera valley
Borzoli Ovest	+8.0	"
S. Quirico	+8.5	village in the Polcevera valley
Voltri	+8.5	working-class town on the far Western coast, left-wing stronghold
Pré	+9.1	immigrant-heavy quarter in the medieval/renaissance center
Palmaro	+9.2	working-class quarter on the far Western coast
Sestri	+9.5	working-class city-in-a-city on the Western coast, left-wing stronghold
Bolzaneto	+9.8	working-class town in the Polcevera valley
Multedo	+10.0	quarter on the Western coast
Molo	+10.0	immigrant-heavy quarter in the medieval/renaissance center

Compared to 2019 it seems that some of Genoa's bourgeois quarters have reverted to a very traditional pattern.

On the other hand there seems to be no evidence that many 2019 M5S voters (the M5S vote in Genoa was almost perfectly negatively correlated to income) went towards the Right in 2020. Which might come a bit as a surprise, because after all at the 2019 European elections the M5S was still governing with the Lega.

Like you I think that Genoa voted so far for the Right this one time (because of the bridge?) and that Genoa will revert to a more left-wing course at the national level. I would be cautious though to put Genoa into the same basket as Milan, Bologna, etc. Genoa is very much a post-industrial city with different issues from other big cities.

[palandio]

Hyperbolically speaking it could be argued that the governments during Italy's First Republic were very stable in the sense that it was always the same parties involved and the general outlook was always the same*. And exactly this led to a very short average survival time for each cabinet. If things will remain almost the same anyways then why not topple the government as soon as there is the opportunity for a litte gain?

*There were of course some significant developments over time like the compromesso storico, the rise of Craxi's Socialists, etc.

[palandio]

Hyperbolically speaking it could be argued that the governments during Italy's First Republic were very stable in the sense that it was always the same parties involved and the general outlook was always the same*. And exactly this led to a very short average survival time for each cabinet. If things will remain almost the same anyways then why not topple the government as soon as there is the opportunity for a litte gain?

*There were of course some significant developments over time like the compromesso storico, the rise of Craxi's Socialists, etc.

Yeah, the First Republic governments were fairly stable in terms of the policies they pursued, even if which specific people pursued them changed all the time. And this long-term policy stability allowed them to advance long-term projects in a way Italian governments have just stopped doing since the 1990s (the last one was really entry into the Euro).

What you both say is true, and, I mean, it's peak Italianness.
As Il Gattopardo says: If we want that everything stays the same, we need that everything change.

I've always interpreted that line in the context of the novel as a very clever man deluding himself about his own failure, not an actually coherent philosophy. For all the clever tactical victories achieved by The Leopard during the Risorgimento, in the last two chapters of the book (Don Fabrizio's death and his daughter elderly spinsterhood) we see that ultimately everything he's been trying to preserve has disappeared.

That's an interesting interpretational ambiguity. In fact what Tomasi di Lampedusa originally meant and how it was widely understood don't necessarily need to be the same. I never read the novel but in the film it looked like Don Fabrizio approved of Tancredi's successful adaption to the new regime. And I think that this is how many understood it.

[palandio]

I think you can explain this chart better than me

There are 321 senators, hence for a (absolute) majority you need 161 of them.

Before the crisis the government had 166, of these 18 from Renzi's Italia Viva.

Subtracting them you get 148 votes, short of the 161 you need. Three senators from the mixed group that hadn't supported the majority before, up to three senators from FI/Udc and up to three life-long senators might join the majority for a total of 151-157 senators, still short of 161. For 161 you would need another 4-10 senators to join the majority.

If the Italia Viva's senators choose to abstain, there would be at most 303 votes, if furthermore all life-long senators abstain, then there would be at most 297. Which would reduce the number of votes needed for a majority to 149 or 152 respectively, enabling Conte to reach a majority.

(I think that the makers of the chart fumbled a bit with the life-long senators in the third diagram. All of them just vanish from the Senate, but at the same time up to three of them are in the new majority. It doesn't make sense.)

[palandio]

So you only cast one vote?  And say I wanted to vote Lega on the list but didn't like the FI candidate in my constituency then tough my Lega vote counts towards that FI candidate's total?

That's naff if it's the case.  I'm all for MMP but it should be that you cast two votes - one for the list and one for the FPTP seat.

Yes, you only cast one vote, which is one of the many stupid aspects of this law. And, as diouf said, it is not an MMP system.

In Germany ideas to introduce a Grabenwahlrecht ("moat electoral law") similar to the Rosatellum existed in the 60s. "Moat" because the expectation was that voters would cast a vote for one of CDU/CSU or SPD because of the FPTP part and that smaller parties would "fall into the moat" because there were 50% proportional seats, but no second vote. It was one of several issus that over time led to an alienation of the FDP from CDU/CSU and to the social-liberal coalition in 1969.

I think that for example in Italian regional and comune-level electoral law the disjoint vote is actually very questionable. With your governor/mayor vote you determine which coalition gets 60% (or a similar number) of the seats and then you can influence the composition of the other coalition with your list vote. That's nonsensical.

But the Rosatellum abolished the disjoint vote in a situation where this would not have been a problem. Therefore I think that the motivation was similar to the one behind the Grabenwahlrecht: Make this a two-coalition fight between center-left and center-right and prevent proportional votes for parties outside of the coalitions (in particular M5S). It just didn't work at all.

[palandio]

Yes, on average during the Second Republic the Right has always been a bit stronger than the Left. When there was still near-perfect bipolarism that meant on average more than 50% for the right-of-center bloc. The Five Stars managed to attract voters from both Left and Right, but a majority of their voters from the Right has now returned to the Right. Additionally of course there have been some movements between the blocs. The Right has attracted some votes from the Left (often the movement has been Left -> Five Stars -> Right), Renzi/Calenda have attracted some voters from both Right and Left, and the Left has attracted some voters from the Right. The Five Stars vote-wise by now draw mostly from the left-of-center genepool. So the Left in a wider sense in now split between the PD-led bloc, the Five Stars and Renzi/Calenda (who are not left-wing of course, but genepool PD).

Hence 47% for the main right-of-center bloc is not insanely high by pre-Five Stars standards.

What has changed is the distribution of the right-of-center vote itself. Both FdI and Lega are arguably more right-wing than Forza Italia.

[palandio]

New projections out for the Chamber of Deputies…the CdX has a 99.4% chance of winning a majority

https://twitter.com/sbidimedia/status/1563892738793275392?s=21&t=gHXPeMUUmuQ...

At first I wanted to say that it would have made more sense to order the seats from left-leaning to right-leaning instead of from "most competitive" to "least competitive" for a variety of reasons, but then I noticed that with a few ecceptions for Scandicci, Florence and Bologna the order wouldn't change all that much.

[palandio]

So the combined total of the PD and friends, Conte and Renzi/Calenda is not too far away from the right-wing alliance, but they might hand them a two thirds majority nonetheless thanks to egos and the very questionable electoral system. How depressing.

I mean there are reasons for the split that go beyond egos.

Others have complained that the PD is treated like it has been governing alone for the last decade. And yes, it's unfair. But the PD is the only (the only stable one at least) party that has defended the state and its institutions again and again, over and over. The amount of self-identification with the institutions and the self-image as "the only adults in the house" (despite all the infighting) is maybe what keeps the main part of the PD and its remaining voter base together despite all the infighting. The PD is identified with the state to a degree that most other parties in the world are not.

On the other hand no European country has seen the same amount of stagnation and relative impoverishment as has Italy. (You can make a point for Greece during the last 13-14 years, but before that Greece was growing fast. Italy has been shrinking for a long time.) The permanent crisis has hit hard on significant parts of the traditionally left-wing voter base. And large parts of the PD have been unable to find an answer that would convince them.

Now there is a cleavage between the Five Stars on one side and the PD on the other side, with Calenda/Renzi taking the PD position on steroids to a degree that they again seem almost opposition-like. The Five Stars are the expression of impoverishment, both in the South and in working-class neighborhoods of the North and Center. Their political ineffectiveness doesn't really change that.

The problem is that the Italian center-left (in a broader sense) and its voter base don't have a common political vision that would go beyond keeping the Right out of power.

[palandio]

/snip

Good maps. There seems to be an error regarding the FdI percentage in Lombardia 4, though.

[palandio]

Okay, one final metric to try to get around the distortionary effect of the 2-weeks polling ban. I have taken the previously computed polling average and then corrected it using the average swing seen in the two illegal polls published after the ban. Admittedly that's a pretty ad-hoc process (there's a reason I'm only doing it now and not before the election) but it's the best way to actually measure the effect of late trends vs actual, genuine polling error.

Here is what we get:

FdI: 24.7%
Lega: 9.8%
FI: 7.7%
NM: 1.5%
Total Right: 43.7%

PD: 20.4%
AVS: 4.0%
+E: 2.2%
IC: 0.9
Total Center-Left: 27.5%

M5S: 15.9%

A-IV: 7.5%

Italexit: 2.9%


Based on this, the right actually performed exactly as expected, while both the left and even M5S underperformed (as well as Italexit, lol). Sad, but I guess in line with past results.

Looking party by party, the only really significant movements we see is FdI taking an additional point away from Lega, and PD just hemorrhaging 1 to 1.5 points (about half of which went to +E and probably a little to A/IV too).

You could also try to extrapolate the trend observed during the weeks before the polling ban:
https://en.wikipedia.org/wiki/File:Opinion_Polls_Italy_General_Election_2022.svg
Or you could look at exit polls.

Whatever you do, one percentage point more or less is not a lot in the polling business. Given their abyssmal track record (even regarding exit polls), I was sure that Italian pollsters would get at least one number completely wrong. They didn't, they were spot on. Either they were super lucky this time or they really improved massively. Chapeau, Italian pollsters!