[EM] Oops! I said it backeards. LR's overall-counted s/q are more equal in Raph's example.

Michael Ossipoff email9648742 at gmail.com
Tue Jul 10 08:13:32 PDT 2012

Contrary to what I said before, LR, not SL, makes the overall s/q more
equal in Raph's SL bad-example.

When the small parties are considered as a whole, their overall s/q is more
nearly equal to that of the big party in LR, as compared to in SL. And for
both sets of voters (small and big party voters), the s/q are closer to
what they should ideally be (when the small voters and their seats are
looked at overall, instead of individually by party).

So it could be argued that LR looks considerably better in that example.

But, if you're a voter of one of the small parties, you might not think so.
When a small party gets 0 seats instead of 1 seat, its s/q differs from the
ideal common s/q by more than it would if it got one 1 seat.

You might not feel any affinity to the other small parties and their
voters, and you might wonder why we'd think that it makes sense to lump you
together with them.

So, looking at it in terms of fairness to individual parties and their
votes, SL is still fairer in this example--if splitting strategy isn't
being used.

So, still, LR's value remains only as the backup for when
splitting-resistance is needed.

LR stays within quota, and that tends to look better. For SL/Webster to
look better in that regard, maybe it would be better to go back to the way
Webster and Jefferson were done initially.

When Jefferson's method (d'Hondt) was used for apportionment, and later,
when Webster's method was first used, they were used with variable
house-size. When the states' populations were divided by a pre-chosen
divisor, and rounded off (according to the particular method's round-off
rule), that was the number of seats. In other words, the divisor, and not
the total number of seats in the House, was what was specified and
constant. Maybe we should go back to that.

I'd suggest choosing a divisor that is chosen so that the smallest state
rounds to one seat. Then every state could have a seat without violating
the Webster allocation rule.

If that makes the House quite large, then maybe it would be best to make it
a unicameral Congress, consisting of the House only. But that would violate
the Great Compromise, under which condition the small states were willing
to join the union, and so it couldn't be done without their agreement.

Mike Ossipoff
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20120710/b8baeee9/attachment-0003.htm>

More information about the Election-Methods mailing list