interactive chart on the history of parties in Congress

[Dan Eckam]

Hi y'all, I thought you might be interested in a new interactive chart I just published, showing parties in Congress by % of seats, from the 1st through the current 114th. I think it's a bit more complicated than the commonly understood orthodoxy that we've always had a two-party system.

Here it is: http://www.daneckam.com/?p=314

-Dan

[Warren D Smith]

that's a great chart! I'd like to steal it
(with credit to you of course) and
make it also be a page on rangevoting.org ?

> --
> You received this message because you are subscribed to the Google Groups
> "The Center for Election Science" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to electionscien...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
>


--
Warren D. Smith
http://RangeVoting.org <-- add your endorsement (by clicking
"endorse" as 1st step)

[Dan Eckam]

Sure, that's fine with me! You'll probably want to use an iframe, as I did; I'll presume you can figure out what you need from the page source, but just let me know if I can help.

[Warren D Smith]

I think this chart pretty clearly illustrates how 2-party domination
took over the USA, much more fun than a numerical table.
It might also be interesting if you could redo it for a number of
other countries, e.g. Australia (house), Canada (parliament),
India, UK, Ireland (Dial), and some longstanding PR countries...

[Dan Eckam]

I don't think I have time to do all those, but maybe Australia, if I could find a comprehensive source for the data. (Aside from the custom-placed labels, the chart is driven by a spreadsheet.) I've been curious about the Australian situation... I guess I'll go ahead and elaborate on that here.

I know that Australia is generally considered to have a 2-party system, but if so, it seems not to be the duopolistic version that we have in this country. Looks like about 20 out of 76 of their Senate seats are occupied by people who are not part of the main two coalitions.

Their House is more dominated by the two coalitions, and I know they use IRV for those elections. And I've seen the claim here that this is evidence that IRV leads to two-party domination. But (a) there are other factors, particular to Australia (such as mandatory voting and party-recommended rankings); how can we be sure that IRV is the crucial factor? And (b) is there any model for *how* this dynamic works? I mean, a solid theory should include a mechanism, right?  

Thinking about how the "spoiler effect" works (ie. Duverger's law), it seems to me that if plurality voting leads to two-partyism (which seems clear), and a top-two runoff system leads to more than two (note Duverger said "the double ballot majority system ... tend[s] to favor multipartism"), then I don't see *why* IRV (per se) should lead to two-partyism. 

I've been curious about that and would be interested in hearing thoughts. If this isn't the right thread for it, I'll be happy to copy this to another thread.

-Dan

[Warren D Smith]

Australia senate is elected by a PR voting system
and therefore it avoids 2-party domination both in theory and in practice.

Australia house is elected via IRV and if has developed
massive 2-party domination described here:
rangevoting.org/AustralianPol.html

----------

Sources of data on party composition thru history of various bodies:
Australia:
http://parlinfo.aph.gov.au/parlInfo/download/library/prspub/3283353/upload_binary/3283353.pdf;fileType=application/pdf
and
https://en.wikipedia.org/wiki/List_of_Australian_federal_elections

Ireland:
their dial is elected via a PR system similar to australian senate but
fewer winners
per election so that the proportionality is poorer. They indeed do achieve more
than 2 parties. Data here:
https://en.wikipedia.org/wiki/D%C3%A1il_%C3%89ireann_election_summary

Canada house of commons (plurality system, but so far has not yet developed
massive 2-party domination):
https://en.wikipedia.org/wiki/List_of_Canadian_federal_general_elections

India: do not have a data source, they are similar to Canada but
even less 2-partified

Sweden would be a good party-list-PR country to use

---------------

Re stealing Dan's page...
CRV pages intentionally avoided using javascript, php, etc and
basically tried to avoid
anything going beyond bare bones HTML... this was to make
it more maintainable and to avoid issues that some system became
outmoded... so that means I may have some difficulty stealing it
or making my own versions of his work...

[Dan Eckam]

OK thanks for the links; I accept the mission to make this chart for Australia's House. I'll probably be able to get to it in the next 1-2 weeks. Along the way, I'll probably modularize the code a little bit, to separate the "engine" from the country-specific part (so if you're going to delve into it to that level, it might be best to wait). 

If you don't want to rely on Javascript, or use an Iframe, one option might be to just use a static screenshot linking to my page. I'm not saying this is a great (or even good) option, just one possibility. That would mean my data notes would go with the chart, without having to reprint them, or leaving them out. (I'm not sure leaving them out would be desirable -- but you may disagree; feel free to do whatever you want.)

The only non-bare-bones things it uses are: JS, D3,js, 2 CSV files for the data, some Google fonts, and 2 small CSS files. But if you include it as an Iframe all that will be handled on my server. (Actually, I just realized I'm using an Object/Embed tag, not Iframe, but I think both could work and for cross-domain, Iframe might be the way to go.) 

Regarding two-party domination, I think your page makes a lot of strong points, but it all seems to be empirical. I have been hoping to find something more theoretical. The thing is, in idealized terms, IRV is basically an "N-ballot majority system" -- following Duverger's meaning when he spoke of the "double ballot majority system". So there is a pattern that comes into play with what I think of as the spoiler-effect mechanism:

1-round (plurality) ---> system tends towards a max of 2 parties

2-round (top-two) ---> system tends towards a max of 3 parties (*)

N-round (IRV) ---> no max imposed, at least on acct. of the spoiler effect mechanism (N is however many rounds are needed)

* I suppose this is debatable. But in terms of the spoiler effect, it makes sense to me. I would propose that France does have this tendency. But of course, a tendency can easily be overridden by other factors. France has several small but significant parties. In 2002, the top-two system backfired when Le Pen made it to the 2nd round, where Chirac wiped him out. That backfire is the form that the pressure of Duverger's law takes, it seems to me, in driving them towards no more than 3 parties. But that pressure is outweighed by other pressures for more parties. 

I did say "idealized" above, and that's because I'm eliding the difference between IRV and the exhaustive ballot, which would be the true N-ballot system. But if that difference makes a real difference to the outcome, it's unclear to me why it should. 

-Dan

[Warren D Smith]

well, re desire for "more theoretical" approach to 2-party domination, see

http://rangevoting.org/NESD.html

which could use improvement, i.e. rewriting... especially if you do it :)


re your chart, I know how to make such a chart with plain HTML although it
would not be as good as your chart (would miss your "magnify" feature)

[Warren D Smith]

also I suppose we could just use an image (such as png)
e.g. your page produces an image in the end, get it, save it,
and then I could just use that with zero code complexity.
(Could even say "click here for magnified image")

[Dan Eckam]

UPDATE: I'm way behind my original schedule for the Australia chart, but I'm now within a few days of completing it.  I decided to include data for both the House and Senate, viewable either separately or together.  It's gonna be great!!  Link soon.

-Dan