New Zealand’s electoral system gives it a parliament which represents voters. Its winner-takes-all executive government, however, remains unrepresentative.* (This is a follow on from the earlier column on coalitions.)
This paper tries to evaluate various coalitions on the basis of their political ideologies. It uses the scores given to parties by the TVNZ website Vote-Compass, which identifies two dimensions: Right-Left and Social Conservative-Social Progressive. The party scores are derived from coding the party positions over the thirty attitudinal questions by a combination of interpreting the party statements and after dialogue with the parties.
(Note that I have carried out some simple transformations of their scores which means the average New Zealand voter is 0 on the right-left dimension and 0 on the conservative-progressive dimension and that the standard deviation across all scores is 1 in each case. These are trivial changes and do not affect the results.)
Table I: The Ideological Positions of the Parties
|
+Right/Left- |
+Conservative/ Progressive- |
Seats |
National Party |
1.05 |
0.90 |
58 |
Labour Party |
-1.00 |
-0.82 |
45 |
New Zealand First Party |
-0.37 |
0.83 |
9 |
Green Party |
-1.26 |
-1.99 |
7 |
ACT New Zealand |
2.00 |
1.29 |
1 |
The Opportunities Party |
-0.26 |
-1.68 |
|
Maori Party |
0.16 |
-0.11 |
|
Mana |
-0.89 |
-1.76 |
|
United Future |
-0.05 |
-0.04 |
|
Conservative |
0.90 |
1.93 |
|
All Voters |
0.00 |
0.00 |
120 |
Table I gives the Vote-Compass (transformed) scores for ten parties together with the score for All Voters. **
It shows National at 1.05 on the right-left spectrum which means it is 1.05 standard deviations to the right of the average voter. National’s score on the conservative-progressive dimension represents 0.90 standard deviations more conservative than the average voter. In contrast Labour is -1.00 and -0.82 which makes it to the left of average voters and also more socially progressive. Intriguingly it is almost exactly as extreme as National but in the opposite direction.
On the right-left dimension, Vote-Compass thinks the right-wing parties are ACT (the most extreme), National, Conservative, and the Maori Party. The other six are to the average voters’ left: UF (very slightly), TOP, NZF, Mana, Labour and Greens (the farthest left).
It thinks that socially conservative parties are Conservative, ACT, National, and NZF. United Future is again near the middle. The socially progressive parties are, the Maori Party, Labour, TOP, Mana, and the Greens (the most socially progressive).
The numbers are calibrated so that the average voter is (0,0) but that need not be true for parliament. In fact it is almost, at (0.02, 0.08). The small differences are because parties which did not win seats were, on the whole, slightly more left and slightly more socially progressive.
(If parliament had no list seats and had the same electorate seat allocation as in the recent election it would be 0.23 standard deviations to the right and 0.21 standard deviations conservative. So it would be unrepresentative of the population. Because voters change their choices in a Front-Runner election, this is hypothetical, but it illustrates how unrepresentative a parliament can be under the FR system.)
Coalitions
TABLE II: Governments
|
+Right/Left- |
+Conservative/ Progressive- |
Seats |
All Voters |
0.00 |
0.00 |
|
Parliament |
0.03 |
0.08 |
120 |
National-Labour |
0.18 |
0.18 |
102 |
National-NZF |
0.78 |
0.82 |
65 |
National-Green |
0.69 |
0.49 |
64 |
Lab-NZF-Green |
-0.84 |
-0.65 |
63 |
WTA in a FR election |
1.05 |
0.90 |
|
However, Table II also shows that while the next parliament may be representative, any of the four coalitions will not be.*** One, L/NZF/G, would be strongly to the left and socially progressive; the other three, dominated by National, would be strongly to the right and socially conservative. (The closest there is to the population average would be the Grand Coalition of National and Labour.)
Perhaps, ironically, the coalition outcomes look about as extreme as a Front Runner parliament. However if there had been such an FR parliament, it is probable that National would have been elected to govern alone as winner-takes-all. In which case the executive government would have been even less representative of voters.
In summary, while New Zealand democrats may be satisfied with the MMP electoral system which gives the public a parliament which reflects their values, they must be uncomfortable about the resulting executive government. It may not be as extreme as the one the Front-Runner system would give them but it still does not reflect the populace.
Perhaps the logic of an MMP parliament is a minority government. (It might be buttressed by minor parties which provide supply and confidence but no guarantees on individual policy.) Such minority governments have occurred more often in parliamentary systems in other countries than you might expect, but as far as I know they have never been dominant in any one of them.
In principle when there is a minority government, parliament, which is more representative of what voters want, would be driving policy more. That is a bit like some governments we have had in the post-MMP recent past. They seemed to be anomalous, but they may only seem so from a Front-Runner perspective.****
* The original paper, publihed with election night outcomes has been updated to the final outcome. There are small numerical differences but no substantive changes. For the records the specials votes were 0.20 standard deviations to the left of election night votes, and 0.24 standard deviations more socially liberal.
** Vote Compass does not give scores for very minor parties which together obtained about 0.6 percent of the vote.
***The calculations assume that the elected MPs reflect their voters. In fact they may not.
**** The same calculations using the Political Compass scores give very similar conclusions.