Improvements to the Democratic Primaries

Given the crap that has happened this year, the DNC should think about overhauling its Primaries process.

Some possible areas of improvement/change
[li]Eliminate superdelegates (since the system seems undemocratic)[/li][li]Eliminate caucuses (due to a host of issues, e.g. they do not provide voter-choice privacy, since others see who you vote for)[/li][li]Shorten the primary season so that it doesn’t last this long (since a long fight is more likely to harden the warring camps’ supporters)[/li][li]Rotate which states get to be first in the primaries (so that other states don’t feel left out and push their primaries so early that they violate the party rules and lose their delegates)[/li][/ol]

I personally would like to see the above implemented.

What do you guys think of the above set of changes? Do you agree with any of them? Do you have others that you propose?

I like your suggestions. I’d add make them winner take all, as is the case in the General Election. If you want to treat Maine and Nebraska differently, that’s OK-- use their system of allocating electors during the primaries, too.

I’m conflicted on open primaries. Lots of people don’t register, and if you close off the primaries, then you don’t get input from the people who are critical in every elections-- those unaffiliated with a party.

Better idea (strictly WRT scheduling): Five Super Tuesdays, spaced two weeks apart. No rotation. Instead, the ten least populous states go first, the ten most populous states go last. That means the small states get the early attention they want; each Super Tuesday will be progressively more expensive to campaign for, so only the most serious candidates will stay in after the second or third; nothing is finally decided until the ten biggest states vote, so the contest remains exciting and engaging right up to the end.

Also, the Super Tuesdays should be in May, June and July. Shortens the campaigning season (I hope we can all appreciate that at this point!), while still leaving plenty of time to gear up for the convention and the general election.

  1. Eliminate superdelegates (since the system seems undemocratic)
    OPPOSE (partially). I think the US reps and Senators, governors, and former nominees should get their votes without regard to the primaries.

    1. Eliminate caucuses (due to a host of issues, e.g. they do not provide voter-choice privacy, since others see who you vote for)
      OPPOSE. If the state party wants a caucus, I think it’s their right.

    2. Shorten the primary season so that it doesn’t last this long (since a long fight is more likely to harden the warring camps’ supporters)
      FAVOR. I would have no primary before Feb 1 without permission from the DNC and no primary after March 31. That six week lull before Pennsylvania was nuts this year.

    3. Rotate which states get to be first in the primaries (so that other states don’t feel left out and push their primaries so early that they violate the party rules and lose their delegates)
      OPPOSE. IA and NH are good at what they do. The candidates should have to perform in the “retail” states and let these experienced voters go over them.

I am opposed to winner take all as I think that exaggerates the importance of the early states. For all its faults, this year’s Dem primary season got a lot of people involved and the candidates campaigned in nearly all the states. Except for the length of the season, which can be fixed easily, the system worked.

Shouldn’t you make the case that the primaries are broken and need fixing?

Just because this one had a close fought race does not mean the system is broken because roughly half the people are pissed off about losing.

  • I agree the Superdelegates do seem somehow undemocratic and the thought of them possibly overriding a candidate that has the popular vote would be ugly.

  • Why are caucuses worse? Because people see you vote? They have been done for many, many elections and no one seemed to find them overly problematic till now (and frankly they were Clinton’s to lose at the outset).

  • I agree a shorter primary season would be good. I suppose the party’s figure that means more press coverage. I do not know it is “bad” per se.

  • I like the rotating states idea. Perhaps it would stop the jockeying to be first that nailed FL & MI.

I like this idea.


Why should candidates have to perform in “retail” states? What is it exactly that IA and NH are good at? And isn’t it very undemocratic to have a tiny fraction of the population with so much influence on the process?

Our current system is like a car that gets you to your destination with a lot of audible grinding and sputtering and at least three stops to pour water on the overheated engine. It’s not strictly broken, but a trip to Pep Boys is something you’d better not put off another week.

I don’t understand.

What was wrong with this year’s primary season?

I found the Democratic horse race to be very exciting. I got to see a lot more of the Democratic candidates than I would have if Hillary had been crowned “presumptive” like McCain was in early March. It keeps the party base pumped up until the convention.

What are you trying to fix?

I’d divide reforms into two categories: calendar and noncalendar.

Calendar, of course, is about which states hold their primaries when: all at once? Regional primaries? A few early ones, then a free-for-all? Who gets to go first? All that stuff.

The range of options for calendar ‘reforms’ is so extensive, and intelligent people have such widely varying views on the subject, that I despair of getting any sort of consensus on this, ever. For instance, to me the idea of a one-day national primary is crazy, but every time the idea of how to reform the primaries comes up, people who have a goodly number of brain cells to rub together propose it.

The noncalendar reforms, OTOH, are a more fertile ground. Even though there’s still plenty of room for disagreement, it’s not quite as extreme.

1) Proportional representation v. assorted forms of winner-take-all: I happen to like the idea of proportional representation, just as I like the idea of the Presidential election being decided by popular vote. But there’s not much to debate: you either like it or you don’t.

The one argument I’ll make for it is that, at least IMHO, it should help a candidate just as much if his/her support in State X goes from 55% to 65% as when it goes from 45% to 55%. Why should that next 10% count less than the middle 10%, when it’s the same number of people involved?

*1a) How to do proportional representation. * Not the way the Dems did it this year - proportional by district and statewide, with two different pots of delegates statewide. I never did figure out why the PLEOs were treated as a separate pool with respect to proportionality.

Just have one big pool of delegates in each primary state that are apportioned to the candidates proportionally. If Dems are worried that somebody or another won’t get represented as a result, they should find another way to deal with that. They could lower the 15% threshold in larger states, for instance. But take Ohio: Hillary won that state, 54%-44%. She won the delegate competition by a distinctly closer 74-67 margin, and stuff like that made it harder than it should have been for her to get back in the race. If she’d won by 78-63, in keeping with her and Obama’s relative proportions of the vote, it would have still been very hard for her to catch up, but at least she would have had a chance.

Even though Hillary’s people should have been able to play the delegates-by-district game as well as Obama’s did, the fact is, the voters are the ones being gamed, either way. (You know me: I’m a small-d democrat. I don’t like it when the players can play games to maximize the effect of some voters, and reduce the effect of other voters, for their own ends.)

*2) Superdelegates: * I’d keep some, but I’d limit it to elected Congresscritters, Governors, Presidents, Veeps, and ex-Presidents and ex-Veeps. That would cut the size of the superdelegate pool by more than half, putting them in a position to only be able to decide really close races. And the ‘deciders’ would be people that had gotten to that position by being nominated to high office by their party, and elected by the voters, rather than being dominated by a bunch of state Democratic Party officials you’d never heard of.
3) Caucuses: *I’m personally partial to having a handful of caucuses in small-to-midsized states that don’t have NFL/MLB-sized cities. (It’s just too hard to do caucuses when you get too many people in one place involved.) I think caucuses measure the enthusiasm of supporters, and I think a system that puts some weight on that, even if most of the weight goes to the sheer numbers of people voting for a candidate, isn’t a bad idea. Some candidates have broad support within the party, but few enthusiastic backers. They look like Dukakis in 1988, or Dole in 1996.

I’ll give my idea of a good alternative primary calendar later; I’ve got to go do some work now.

Why is it crazy?

Why should the primary vote be different from the general election?

Would you be OK if the general election for president were done like the primaries? i.e. Iowa and NH go first, then a few other states, then we have a “Super Tuesday”. etc. And in many elections, the president will be decided before many states have even voted.

If this sounds bad for the general election, why is it OK for primaries?

We need some superdelegates as a safety valve in the event of a close election. I think if you’re in the US Congress or a governor you’ve earned the right to have your voice in the mix.

Iowa and New Hampshire do their jobs well because they get and take enormous amounts of time going over the candidates. That early test often separates the contenders from the pretenders.

A national primary would be a national disaster. If we had one this year, the nominees would be Hillary and Guiliani. 'Nuff said.

Some see this primary as a giant cluster f***. I see it as working just as it was intended two with two popular and closely matched candidates. I see Sen. Clinton’s open and public courting of Superdelegates as the main problem. This is a candidate problem, not a problem with the system.

I like the proportional representation AND caucuses. One problem with simple voting is it measures the depth of a candidate’s support, but not the depth. That is to say it makes no distinction between a voter that passionately supports a candidate, and another voter that is only choosing the slightly lesser evil (in their mind). Caucuses (imperfectly) measure the voter’s level of commitment in a way that a simple counting of noses never can.

I also have no problem with the Superdelegates. They would need a VERY good reason to overturn a mandate by the electorate, or would be certain to lose the GE. Imagine if Rush Limbaugh’s “operation chaos” had been responsible for Hillary getting the nomination…

I’m all for a rotating primary order and shortened primary season.

Glad you asked, even if that was simply meant to be illustrative of how intelligent, well-meaning people come to diametrically opposite positions about what the ideal primary calendar would look like.

The general election and the primaries are fundamentally different kinds of contests. A major-party candidate goes into the general election campaign with the imprimatur of a major party, and having already been introduced to the country by winning the lion’s share of his party’s primaries, plus the more formal introduction of his party’s convention. Anything that each candidate and party can’t communicate to the voters about themselves and the other party and candidate in the two months or so between conventions and election probably isn’t worth communicating. There’s no reason why everyone can’t vote at once, and all other things being equal, it’s the fairest way to do things.

The contenders in primaries, though, are individuals, with greatly varying levels of backing, money, name recognition, and all that - but still largely known superficially, to the extent they’re known at all. And FWIW (a great deal, I think), there are often more than two candidates running. In many years, a national primary would almost surely give the win to the candidate who entered the primary season with the most money and name recognition (1984, 2000, 2008). But in many other years (1972, 1976, 1988, 1992, 2004), chances would be good that nobody would win: the vote would be divided several different ways, the primary would be over and done with, and the voters would now be entirely outside the real process of choosing a nominee. The voters would have gotten a brief look at a bunch of people they barely knew, and would have cast their votes on that basis, as nothing more than a stopover on the way to the smoke-filled room.

It just doesn’t seem to be a very satisfactory approach for either the voters or the candidates, except for the occasional candidate that comes in with all the marbles.

I would make one change to your plan: the first flight of ten states will be the ten states where the statewide results most closely matched the national popular vote results from the previous general election. This means that the early states which tend to generate momentum are also the states whose demographics are most representative of the most recent snapshot of the voting public. If we followed that example, this year’s first ten would have been:

New Hampshire
New Mexico

…and Obama would have won six of them and tied one (if you assume this year’s results are representative). Take a look at how many of those turn out to be Must-win and Swing states according to Five-Thirty-Eight, and I think you’ll at least see why I believe the idea has merit.

I agree that a single-day national primary is a bad idea because it rewards fundraising at the expense of useful debates on the issues. As far as how primaries are decided, I prefer proportional representation. Currently a candidate can win the general (via the electoral college) by only carrying the fourteen most popular states with 51% each, and that seems wrong to me. I would keep superdelegates, but restrict the list to politicians whose down-ticket performance will be affected by the candidate’s popularity (that is: facing election or reelection in 2008-2011 for this cycle). This helps the party choose a candidate with coat-tails, which is an important factor and probably contributed to Obama’s success among the supers this year.

Why not just have one national primary decided on nothing but the popular vote? Why dick around? They could have the Primary season over in a day.

I came back to amend my list of the first ten states to vote. In order, the states whose vote totals had the smallest root-sum-square difference from the 2004 national popular vote were:

Ohio (0.4%)
New Mexico
New Hampshire
Pennsylvania (3.55%)

…which means that Clinton wins five and ties one, probably giving her a narrow lead after the first Tuesday but leaving Obama with plenty of momentum to run on.

Honestly, did you even read any of the other messages in the thread? Here are a few good reasons, taken from three of the last four posts:

How about we let anyone who wants to be the Democratic nominee write their name down on a slip of paper. Then we put all the slips into a big hat. Then on January 1, we pull a slip out and that person gets to be the Democratic nominee.

Because pretty much any other system is going to have problems. It’s not the procedures of the system that causes discord, it’s that a number of people are all competing for a single nomination. No matter what you change, you’re going to have one winner who thinks the system worked right and several losers who think the system needs to be changed.

Being in a semi-hybrid caucus/primary state with confusing enough rules, I haven’t really been paying attention to how other states allocate their delegates. I had assumed that states with primaries divided their delegates proportionally - is that not so? Are there winner take all primaries?

This is going to be a hugely unpopular opinion, I think, but I rather like the Texas primacaucus system. It allows most delegates to be apportioned based on a secret ballot, high-participation primary, while still allowing other delegates to be apportioned based on depth and passion of support in a caucus*. I don’t know if the 2:1 split that Texas does is the right one, but I like the fact that there is a split.

  • So why do I like caucuses? For much the same reason that Kos over at DailyKos does: they foster party building and organization. You get a (potentially) huge number of people to show up, provide information about themselves to the party, get signed up to be door-knockers and phone-bankers for the general election, and in general get more firmly tied to the party.