| Hacker defined "divisive" primaries as being those where the winner received less than 65% of the vote. He looked at 220 elections for Senator and Governor between 1956 and 1964 and found that only 27 cases did both candidates in the general have divisive primaries and in 99 one candidate did and the other did not.
In those 99 cases, the candidate who had the divisive primary lost 68 times. While this is a significant majority, the minority where the winner of a divisive primary won the general election is large enough to deserve some attention. It would be helpful to know what electoral characteristics allowed such a large minority to exist. If those characteristics can be identified and encouraged; then there would be no fear of primary challenges other than someone's primary candidate would have to lose.
Of those who withstood primary challenges and then won the general, 18 were gubernatorial races and 13 were senatorial races. What is more interesting than that is that in eight of these cases, the successful gubernatorial candidate beat an incumbent Governor. This was true only three times for the Senate.
The next factor Hacker looked at was partisan friendliness. Utah, for example, is friendly to Republicans. So a Republican who survived a heated primary but won the general is not really anything to write home about. You'd expect it. The real race, in that case, is the primary (happens here in Hudson all the time, but with the Dems, obviously).
Or maybe not. Hacker found that 61 of the 89 candidates running in partisan friendly states actually lost in the general. In those 61 cases, the primary appears to have had enough power to negate the natural partisan friendliness of the territory.
What stands out in this analysis is the power of incumbency to overcome divisive primaries. In friendly states, incumbents won seven and lost two. In unfriendly states, incumbents won two and lost four. Incumbency in friendly states appears to allow any wounds from a divisive primary to heal quickly. In an unfriendly state, they are more likely to open wounds that will actually kill the incumbent.
Interestingly, this appears to be true no matter if only one candidate has a divisive primary or if both have them. Turn-out had a very low effect - only a six percentage-point difference between high and low turnout elections.
Obviously, to the extent that this can be applied to New Jersey, it's good news. Even if the current primary gets nasty and divisive, it is more than likely that the winner will be able to retain the seat in the general. But it doesn't mean it's a sure thing.
In 1974, Donald Bruce Johnson and James R. Gibson looked at the effects of a divisive primary on partisan activists. For their case study, they looked at the First Congressional District in Iowa in 1970, when both parties had contested primaries followed by a divisive general election. The names aren't that important, but what matters is that nearly half of party activists were so upset with the primary that they did not participate (as activists) in the general. This effect was exaggerated in the incumbent's party (and it should be noted that the incumbent party retained the seat).
It was also true that "grassroots" activists, rather than "institutional" activists (my designations) were more likely to drop out. This only makes sense. If a person is a county committeeman, he is more likely to give up the challenger and support the victor in the general than to give the whole party the finger to the bitter end. For some people, this defines the difference between being a grassroots person and an institutional person.
Our final stop of the night is with two articles written by Patrick Kenney and James R. Gibson in 1984 and 1987. They had a much more intricate model, and I won't go into the details. The important part is the findings - and they found that, in gubernatorial contests, a divisive primary costs about 3% support in the general election. In a Senate election, it rises to 7.5%. The second study found that this effect is exaggerated when one party contests the primary and the other does not.
There are several more studies that have been published in the last thirty years. I haven't read them all (yet), but I haven't found anything that directly contradicts the findings of these studies. There are some squibblings over the degree to which the effect exists, but little debate over whether or not it does. One key voice of dissent was that of Alan Ware, who, in 1979, broached the difference between a "contested primary" and a "divisive one".
Narrowly decided elections tend to get nasty towards the end - blowouts tend to be less so. But a narrow result does not necessarily mean that the race was, in fact, divisive. There are also states where "mudslinging" is just business as usual and tends to be forgotten a week after it lands. It is a case where variation may be as important as the overall trend.
So where does this leave New Jersey and our current contested primaries (the Senate race, believe it or not, is far from the only one)? Well, it means that truly divisive primaries will have their largest long-term effect among the grassroots activists. That means that divisive primaries would tend to breed future divisive primaries - such as we are seeing in Bergen County. But it also means that in contested primaries where the divisiveness (calling each other "real Democrats", maybe) is absent, that there shouldn't be enough of an impact on the general election for Democrats to end up on the losing side of things. This is particularly true with the NJGOP in such disarray (Murray Sabrin actually is a Democrat's best friend right now).
Even at the Congressional District level, there should be no real danger to any of our incumbents from a primary challenge. It would be the state legislative districts, the county offices, and the municipal seats where the results consistently come in under a ten percent final vote margin that contested primaries, if they become divisive, could do the most damage. That would happen by eroding grassroot legitimacy and setting the grassroots Democrats against the institutional Democrats (again - Bergen).
I'll close by saying this: Even if a truly divisive (not just a challenged) primary does occur, there is no indication that it helps anyone, other than in the eyes of their own hard-hearted faction. There is absolutely no indication that it helps the party or that it helps the long-term advance of progressive policies and politics.
But it is also true that challenged primaries are a necessary part of our party's growth and health. So long as the involved parties can remain focused on positive proposals and aim at unifying themes and values, there is no reason (well, other than the actual political considerations) not to have a challenged primary. When you have your real opponents on the ropes is no time to start slugging at the guys in your own corner.
So let's keep it above the belt. |
