The basic provisions of Nevada's recall law are standard -- you get 90 days to collect the signatures, and you need 25% of the total turnout when the position was last up for a vote. That means it is all of the voters who came out to vote, not just those who voted for the Assembly position. This distinction does matter, as there are real drop-offs for down ballot elections (this sites quotes a paper noting a 15% drop-off). So the 25% is harder than many states, like Wisconsin, that require 25% of turnout for that specific race. However, other states (NJ) require the signatures of 25% of registered voters. This is a higher standard than Nevada.
What Nevada does have is a stricter provision -- the 25% of voters had to cast a ballot in that last election for that position. This was the source of a lawsuit in 2010, where the Nevada Supreme Court unanimously ruled that the signatures had to be from people who actually voted in that last election. Being a registered voter is not enough. The voter actually had to be shown that he or she actually cast a ballot in the previous race (the original baseline was actually for a Supreme Court race). This decision overturned a 2009 Nevada law that opened up the signature line to be any registered voter.
The Supreme Court could conceivably have gone further and demanded that the voter attest to actually voting in the race for the specific position facing a recall. Other jurisdictions do have that exact provision or harsher ones. A City Council recall in Lubbock was tossed out because the petitioners needed 10% of the signers to attest by affidavit that they actually voted for the candidate facing the recall. Petitioners in that case handed in well over the required amount, but were left crying, waiting and hoping in vain for decision that would allow them to proceed. Here's two other examples, one in New Mexico and the other in Idaho.
Now, in addition to this harder provision, we have to note that signature failure rates -- meaning signatures tossed out for not conforming to the rules for whatever reason -- vary greatly from state to state. But there does seem to be some intrastate conformity. So in California, 15% failure is not unusual, same with Michigan. But in Arizona and Colorado, we frequently see 40% reject rates. Why this happens is not clear -- could be the law, could be attempts to protect incumbents, or it could be that the county clerks take a much harsher line on signature issues. But it does seem to be a pattern. It's possible to overcome it, but it is clearly a problem for petitioners.
Nevada seems to be a state that has a high signature failure rate. Presumably, this may be caused by the requirement that voters actually have voted in the previous election. It also may be due to other reasons. The reality though is that in four years, this blog has seen only one recall get on the ballot in the state.
I mentioned a recent matter last week in Ely that shines some additional light on the problem. The Ely recall is itself worth reading about (the city council broke into a railroad's office to obtain records). But what is useful for us is that petitioners handed in signatures against the Mayor and 5 councilmembers. Enough signatures were tossed out against 3 of them to prevent the recall from getting to the ballot. The only one that I have numbers for saw a 38% failure rate.
The other 3 council members had the Secretary of State and county clerk unable to find the voting records for the 2011 elections. Therefore, they couldn't verify that the recall signers actually voted in the 2011 election. All of the petitions were thrown out and it is clear under this interpretation, no recall can be launched against those three officials.
So you can actually do everything right, but still have an administrative error pull the rug out from under you and kill the recall.
Now, none of these facts should be proof that a recall couldn't get on the ballot against Assembly Speaker John Hambrick or the other two assembly members, Chris Edwards or Stephen Silberkraus. In fact, the signature requirement to get a Hambrick recall on the ballot is low enough (4116) that it would seem to be doable. But as you can see, Nevada law can make it very tough to get a recall on the ballot.
Jeff Singer at the Daily Kos asked me about how a replacement would be chosen, and that also has an odd provision worth looking at. Generally, the recall is basically a new election. All the candidates run, and whoever gets the most votes wins (the First Past the Post system). Arizona and Michigan have this type of recall law (Wisconsin has a similar one, but also includes a party primary).
However, if no one enters the race, there is a "for" or "against" recall vote with just the elected official on the ballot. If he loses, then the replacement is chosen in the "manner provided by law." For the Nevada state legislature, that would be a special election, In this case, the elected official who was just removed in the recall vote would probably be able to run again in the replacement special election. That's not so unusual, but worth noting.
Additional Post Script: Ballot Access News' Richard Winger noted that the Nevada's Supreme Court decision arguably violated federal law. Richard (who really knows this area) argues that the Supreme Court in Bush v. Gore and elsewhere have held that past votes cannot be used to bar signing petitions or from voting. Since the Nevada decision was not appealed to the federal courts, only a future decision would decide this point.