With the state government lurching from budget crisis to budget crisis, many frustrated Californians are thinking about what political reforms might make the state Legislature more functional.

Californians passed Proposition 11, but the impact of the independent redistricting commission won't be felt until the 2012 elections. With Californians still seemingly hungry for reform, what other changes might clean up the mess in Sacramento? One of the proposals being discussed in various circles is what is known as the "top two" primary. Under that method, the nominees from all political parties, including multiple candidates from the same party, compete against each other in a single primary free-for-all, reminiscent of California's short-lived use of the popular "blanket primary" back in the mid-1990s (which was done away with as a result of a U.S. Supreme Court ruling).

But unlike the blanket primary, which advanced to the November election the primary winner for every political party, resulting in a multicandidate field, the top two primary advances only the top two finishers.

And those two final candidates can even be from the same political party. The top two primary was rejected by California voters in 2004, but proponents are trying to revive it, saying that the top two primary will:

• Give voters more choice.

• Create more competition.

• Elect more moderate legislators.

• Get rid of spoiler candidates and elect majority winners.

Let's examine each of these claims. Certainly, the top two primary would give voters more choice during the primary election. But it actually would reduce voters' choices in the November election to only two candidates, which is when most voters turn out. Moreover, in a very liberal district, such as in the urban areas, the top two candidates in November very likely would be two Democrats; in a conservative district, the top two probably would be Republicans. Third-party candidates and independents almost never would appear on the November ballot.

Would the top two primary foster more competitive races? To answer that question, I examined elections from the state of Washington, which used the top two primary for its 2008 state legislative elections. Here's what I found:

Of the 98 state House races, only five races (5 percent) were won by a competitive margin (defined as a 4 percentage point difference between the top two candidates). Sixty-five races (66 percent) were won by landslide margins of 20 points or higher, with 17 of those races uncontested.

The results in the 26 state Senate races were very similar, with 62 percent of races won by landslides and only two races competitive. That's a level of competition that is no better than what we have now in California.

How about electing moderates? How did the Washington elections do in that regard? The term "moderate" is a relative one, with different definitions from state to state, so a better way to examine this is to look for how many opportunities are available for moderates to get elected. In theory, when you have two Democrats running against each other in November, or two Republicans, the voters from the other party could cross over and act as a moderating influence against either the most conservative Republican or the most liberal Democrat winning.

In Washington's House races, only six out of 98 (6 percent) had two candidates from the same party, and in the Senate, two out of 26 races (8 percent) did.So that's not a lot of races in which moderates could have an opportunity to get elected. With Washington's elections being so noncompetitive generally, that greatly limited electoral opportunities for moderates.

One positive from the Washington elections is that for the handful of races that were decided by competitive margins, they did not have to worry about spoiler candidacies coming from third-party candidates. But is essentially banning third parties from participating in the November election really the best way to handle this? A far better way would be to use instant runoff voting, where voters could rank a first, second and third choice, and the runoff rankings would be used to elect majority winners in a single election. Third-party candidates would not be spoilers, and this would preserve voters' choices. As discussion of the top two primary proceeds in California, it seems important that the discussion be factbased.

And the facts from Washington state's elections at least show that the top two primary this past year did not result in more competition or many opportunities for moderate candidates to get elected. It gave voters more choice in the primary but at the cost of reducing their choice in the November election. It elected majority winners and got rid of the spoiler problem but at the price of greatly restricting third parties from the November ballot.

All in all, a cautionary tale about the consequences of the top two primary as political reform.