That 7% number is my estimate (A.K.A. W.A.G.) but I am about to defend it based on recent polling in NY-26.
From Sienna College polling as reported on 538 we learn that Medicare is regarded as the most important issue by 21% of the voters in NY-26. Of that subset, 80% intend to vote for Hochul.
The partisan identification of that 21% subset is 50% Democratic, 26% Independent, and 24% Republicans. The distribution of voters in the poll as a whole is 35% D, 41% R, and 24% I.
In a naive analysis, you would expect all Republicans to vote for Jane Corwin, all Democrats to vote for Kathy Hochul, and Independents to divide equally. That would provide a 6% margin of victory for Corwin, (41+12=53) over Horchul (35+12=47). Coincidentally the district does have a PVI rating of R+6.
However, if you apply that naive analysis to the 21% subset who might be called "Medicare voters", you obtain (D)50% + (I)26%/2=63%. Which is obviously less than the 80% obtained in the poll. In fact, even if EVERY independent Medicare voter went to Hochul, there still must be at least some Republican voters who are switching parties over Medicare.
Unfortunately the distribution of "Medicare voters" voting for Hochul is not supplied in the cross tabs. There is only one equation in 3 unknowns, as follows. If "D" is the proportion of Democratic "Medicare voters" voting Hochul, "I" is the proportion of Independent "Medicare voters" voting Hochul, and "R" is the proportion of Republican "Medicare voters" voting Hochul we have
0.5*D + 0.26*I + 0.24*R = 0.8
Continuing the naive assumption that all Democrats were going to vote for Hochul anyway, we can subtract Democrats from both sides and obtain
0.26*I + 0.24*R = 0.3
At this point an arbitrary decision must be made as to the ratio of Independent voters to Republican voters. At one extreme is I=100%, R=16.7%. At the other extreme is R=100%, I=23%. It seems logical that Independents are more persuadable to the Democratic cause then are Republicans, and that roughly 50% of those Independents would have voted for Hochul any way.
If we take half the Independent "Medicare voters" as being people who would have voted for Hochul anyway, we are left with
0.13*I' + 0.24*R = 0.17
where I' is the percentage of Republican-leaning Independent "Medicare voters" who switched over Medicare.
Now I will arbitrarily assume that Republican leaning Independents are twice as likely as actual Republicans to be persuaded by Medicare.
This lets me add a second equation
I' = 2R
Substitution
0.13*2*R + 0.24*R = 0.17
0.5*R=0.17
R=0.34
I'=0.68
Plugging this back into the original equation I get
0.5*(D=1.00) + 0.26*(I=0.5+0.68/2=0.84) + 0.24*(R=0.34) = 0.8
That is, no Democrats changed their vote because of Medicare, but 34% of Independents and Republicans did - if and only if they choose Medicare as their most important issue.
This table summarizes the data presented thus far.
Party |
% Voters |
% Med Voters |
21% Med Voters |
%Med Voters switching vote |
%All voters switching vote
|
Democrats |
35 |
50 |
10.5 |
0 |
0
|
Republicans |
41 |
24 |
5.0 |
34 |
1.7
|
Independent |
24 |
26 |
5.5 |
34 |
1.9
|
To explain further, 41% of NY-26 voters identify as Republicans. Of those, 24% say medicare is their most important issue. Therefore, 24% of 21% of all voters, or 5.0% of the district, are Republican "Medicare Voters". If, based on my assumptions, 34% of the Republican "Medicare voters" switched their vote based on this issue, that is 34% of 5.0% or 1.7% of the electorate as a whole.
The key point here is that these are voters SWITCHING votes. So every voter gained by Hochul is lost to Corwin. Therefore, Horchul gains 1.7+1.9 = 3.6%, and Corwin loses the same, for a switch of 7.2%.
Some caveats:
- There are other candidates running so not every vote gained by Hochul necessarily came from Corwin
- Obviously "Independents are twice as persuadable as Republicans" is an arbitrary assumption
- Obviously "Independents would normally split 50-50" is an arbitrary assumption
- NY-26 Republicans may be more sane than Republicans in other districts
- There is no attempt to measure increased Democratic turnout - enthusiasm
Having covered my fragrant hindquarters with caveats, I still think that as a rough WAG, Kathy Hochul has improved her position versus Jane Corwin by roughly 7% by beating her about the head and shoulders with the Medicare issue.
Extrapolating that nationwide, I further WAG that even if the 2012 electorate looks like the 2010 electorate the Democrats might retake the House by using this issue alone - approximately 30-40 seats would switch party.