An Anonymous Response to a Lobster's Thoughts on a Leftward Lurch

My new-found blogging opponent, Lobster Thermidor, has posted a lengthy analysis about how the Liberal Party ought to move to the left, not the right. While I disagree with these thoughts in general, his attack is targetted particularly at me, and he has called for a cordial flame war. Well, I always enjoy a cordial flame war, so here goes.

LT's first objection is to my assumption that a one percentage point move to the right in the Liberal party necessitates a one percentage loss of support on the left. He writes that, if this were the case, the Liberal Party would be forever condemned to 30% support.

I very much object to that objection. The methodology used above was one that implicitly said ceteris paribus. This methodology never assumed that Liberal Party support would remain constant. It assumed, however, that in the short term, it would remain constant, and that movements along the political spectrum would maintain constant support for the Liberals.

There is, within this framework, a modification that can be made to allow for not only shifts in the Liberal party along this spectrum, but also a widening of the Liberal support base. In fact, I have updated the model so that it can include variations of this type.

Thus, I do not presume that the Liberal Party is condemned to 30% forever, as I believe that it can expand its support base by attracting both Conservative and NDP voters without major shifts in policy doctrine along the political spectrum.

Now, as LT has rightly pointed out, it may be the case that the a shift in policy of a certain direction may not net neutral for the Liberals. Maybe right-wing voters are unlikely to jump ship if the Liberal party moves to the left. Maybe. I don't think this is the case now, especially as a lot of Red Tories--those fiscal conservatives but social liberals who have continued to vote Liberal because of continued fears of the Tory social agenda--have partially had their fears quelled by limited Tory movement on the social front. Thus, I don't agree with that analysis, but I'm willing to engage it. So, let's use that same model to do so.

Below is presented the the table from my first blog, for reference. For a full explanation of the methodology, read the original post.

Scenario 1: Uniform Distribution

Liberal Deviation	Liberal Seats	Conservative Seats	NDP Seats
-10	66	156	11
-9	67	152	14
-8	72	147	14
-7	75	140	18
-6	77	137	19
-5	80	131	22
-4	81	128	24
-3	85	124	24
-2	86	121	26
-1	89	117	27
0	90	114	29
1	95	109	29
2	94	106	31
3	98	101	34
4	97	97	38
5	97	90	45
6	99	85	48
7	98	82	52
8	99	81	53
9	99	78	54
10	98	75	59
15	96	58	78
20	82	36	115

Now, we'll specifically analyze what LT calls for. Let's assume that when the Liberals move to the left, its right-wing voters are still relatively reluctant to jump ship. Specifically, let's assume that for every 1% shift to the left, the Liberals gain 1% of the vote from the NDP, but they only lose 0.5% to the Conservatives. Thus, they gain a total of 0.5% of voters for every shift to the left.

Scenario 2: Leftward Gain Twice as Fast as Rightward Loss
Liberal Deviation	Liberal Seats	Conservative Seats	NDP Seats
-10	102	117	14
-9	101	116	16
-8	101	116	16
-7	100	116	17
-6	100	116	17
-5	98	116	19
-4	97	116	20
-3	93	115	24
-2	93	115	25
-1	91	115	27
0	90	114	29
1	90	114	29
2	89	114	30
3	87	114	32
4	83	112	38
5	82	111	40
6	82	106	45
7	79	102	52
8	79	100	53
9	78	100	55
10	77	96	56

Under these circumstances, I've made assumptions extremely generous to LT's proposition, assumptions which I believe to be false. Even in these circumstances, the Liberals don't begin to come close to forming a government. Even compared with a rightward shift under this hypothetical situation--it should be noted that using these assumptions, the Liberals have 10% fewer of total voters under the +10 scenario compared to the -10% scenario--the Liberals are almost no closer to being the dominant party. Under a 10% to the right scenario, the Liberals are 19 seats shy of being the largest party, while under the 10% to the left scenario, the Liberals are still 15 seats shy of being the largest party. I grant that these are all restrictive assumptions and such analysis should of course be taken with a grain of salt, but even using LT assumptions, his theory is not confirmed.

Let's be even more generous to the Lobster and assume that for every 1% that the Liberals move to the left, they only lose 0.33% of rightist Liberals--i.e. they gain 2/3 of voters with every 1% move leftward. I find these assumptions to be crazy, but I'll present the results anyways.

Liberal Deviation	Liberal Seats	Conservative Seats	NDP Seats
-10	115	105	13
-9	113	105	15
-8	109	108	16
-7	109	108	16
-6	106	108	17
-5	104	111	18
-4	102	110	21
-3	100	110	23
-2	97	111	25
-1	93	113	27
0	90	114	29

So, with these assumptions, and a large enough movement to the left, the Liberals will form the minority government. But this has taken some pretty dramatic assumptions, which I think are completely untenable.

LT has suggested that there a lot of fronts where the Liberals could gain votes from the left without losing votes to the right. I think this is probably true, and wherever this can be done, it ought to be. I also think there are certain fronts where the Liberals could gain votes from the right without losing votes to the left. That should be done too. But what we're talking about isn't that. It's what direction we ought to move in general.

And as for LT's note that the elections in which the NDP have done best are those elections where Conservatives triumph, sure. That's not because of Liberal shifts, but rather because of the size of the Liberal base (i.e. the other type of change I was talking about earlier which LT completely ignored).

Finally, LT writes:

"A vast majority of NDP votes are spent in ridings they lose, but if all 17.5% of NDP support were to shift to the Liberal Party (absurd, I know, but hypothetically), then making use of my shmancy election predictor, the Liberals would win 184 seats, the Conservatives 81 and the Bloc 43. In fact, the NDP penalises us least in the ridings they win; the real sting of the NDP is in those 17 ridings in Ontario which the Liberals lost to the Conservatives by less than 1000 votes, while the NDP received more than the difference."

Sure, I grant that that's where the NDP penalized the Liberals most, and if the Liberals had those NDP votes, they might have won a bunch of ridings. I also think they could have done that by just trying to get back those Conservative voters. The Liberals were the party of fiscal responsibility in the mid 90s, but as long as there was a Red Tory party (i.e. the PCs), Red Tories were split between the Liberals and the PCs; even then, that moderate right-wing fiscal policy worked out well. Now, there exists no Red Tory party, but there are a lot of Red Tories out there. Those are all there for the Liberals' taking without much work.

Anyways, this all comes down to the question of where party support is the most flexible. I probably won't believe LT's assertion that right-wing Liberals won't ditch the Liberals if they move to the left, and he probably won't believe my counter assertion. But I have showed that, even if we make assumptiosn which are supportive to LT's position, the Liberals still won't be able to win back power.

If we made a similar assumptions favourable to my position, the Liberals would win a landslide with a shift to the right. If we stay neutral as before, the Liberals only chance of winning is with a shift to the right. If we make assumptions amenable to LT's position, a shift to the right won't help, but neither will a shift to the left.