Frequently Asked Questions

1. What is Approval Voting?
2. Doesn't Approval Voting violate the one-man-one-vote ideal?
3. Won't voters just continue to vote for only one candidate anyway?
4. Is it true that a candidate who is the favorite of a majority could lose under Approval Voting?
5. Doesn't IRV capture more information about a voter's preferences than Approval Voting?
6. What's the best way to decide how to vote in Approval?  When should I approve of more than just my favorite candidate?
7. Will Approval Voting elections be more complicated to count?
8. Why is there so much more support for IRV than for Approval Voting?
9. Why don't you just join the IRV movement instead of starting a counterproductive movement for a system no one's heard of?

What is Approval Voting?

Approval Voting is a single-winner election system.  The current system, plurality voting, presents a ballot listing the candidates for a given race and allows only one vote for one candidate.  Approval Voting uses the same ballot format but allows a voter to approve any number of candidates, effectively allowing a vote for or against each candidate.  Approval Voting is a simple reform that gets rid of the spoiler and lesser-of-two-evils problems and would result in a level playing field for all parties and candidates.

Doesn't Approval Voting violate the one-man-one-vote ideal?

There are two good answers to this question.  First, "one man, one vote" originally meant simply that no voter have more power than any other, which is just as true for Approval as it is for any other fair system.  Second, if "one man, one vote" is taken as a blanket objection to any change to plurality voting, Approval replaces that ideal with the more reasonable "one candidate, one vote"; each voter can contribute only one vote to any one candidate's total.  Why should voting for one candidate restrict your right to vote for another?  It could be argued that Approval Voting actually gives more equal power to voters than plurality or IRV, since those systems tip the scales in favor of voters who support major-party candidates.

Won't voters just continue to vote for only one candidate anyway?

Some will, but only when their favorite candidate is leading the polls.  For example, say polls leading up to an Approval election show Bush in the lead with 52% approval, McCain second with 49%, Gore third with 44% and Nader trailing with 23%.  A voter whose favorite is Bush will likely approve only Bush, but a voter whose favorite is Nader would be free to vote for both Nader and whichever candidate(s) he sees as a good compromise.  And it's not only supporters of third-party candidates who'd vote for more than one—a voter who likes Gore first and McCain more than Bush will probably vote for both Gore and McCain.  Under plurality or IRV that Gore voter would have to decide whether to vote Gore first sincerely or to vote McCain first as the lesser evil; Approval Voting allows a voter to help both a favorite and a needed compromise without "burying" the favorite.

Is it true that a candidate who is the favorite of a majority could lose under Approval Voting?

It is possible, but only when another candidate has wider-ranging support.  Which candidate deserves to win: one who is the favorite of 51% of the voters or one who is acceptable to 75%?  (Which result is likely to disappoint more voters?)  But it's practically a moot point, since a majority that shares the same favorite candidate will almost certainly know that the candidate is strong and will likely vote for only that candidate, so such majority candidates overwhelmingly tend to win under Approval anyway.  Approval tends to elect candidates with wide appeal no matter how smart and informed the voters are.

IRV advocates like to claim that IRV guarantees a majority winner.  But IRV only guarantees that the winner will have a majority over one other candidate.  When there are several or more candidates, IRV often eliminates compromise candidates early on and ends up electing a candidate that would have lost in the last round to other candidates—in other words, other candidates can have voted majorities over the IRV winner!  But even in the final round, the IRV winner might not have a majority of the votes since many voters might have voted for neither of the two finalists.  IRV's majority guarantee is weak at best, and that's assuming voters are sincere.

Another reason to abandon this particular ideal of majority rule is that every voting system that guarantees a majority winner sometimes rewards voting a compromise candidate over a favorite candidate.  Plurality and IRV certainly do.  Approval never does.

Doesn't IRV capture more information about a voter's preferences than Approval Voting?

It is true that IRV allows a voter to express more preferences among candidates.  For example, a voter who votes Browne over Nader over Gore over Bush in IRV is expressing six pairwise preferences (Browne over Nader, Browne over Gore, Browne over Bush, Nader over Gore, Nader over Bush and Gore over Bush), while a voter who approves Browne and Nader is expressing four (Browne over Gore, Browne over Bush, Nader over Gore and Nader over Bush).  However, IRV throws most of that information away during the count!  IRV looks only at first-place votes at any given time, which is why compromise candidates with wide appeal are often eliminated.  A sincere voter has no control over which preferences IRV will count, and IRV often eliminates a voter's compromise candidate before a favorite candidate, which can result in that voter's least favorite being elected.  Also, smart IRV voters will often vote contrary to their true preferences, so voted IRV preferences can be a poor guide to sincere preferences.  Approval Voting lets the voter decide which preferences are most important and reliably counts every single voted preference.

What's the best way to decide how to vote in Approval?  When should I approve of more than just my favorite candidate?

The short answer is that instinctive Approval Voting strategy tends to be very effective.  Voters generally know whether they approve or disapprove of a given candidate, and they usually can tell how much compromise is necessary without deep analysis.  But whether strategic thinking is conscious or not, some voters find an explicit strategy useful.  There are several good ones.

Perhaps the simplest guideline to follow is to approve the candidate you would have voted for under plurality, plus every candidate you like more.  So if Nader is your favorite, but Gore and Bush are the two clear frontrunners, you might have voted for Gore under plurality to keep from "helping" Bush.  In that case you'd approve both Nader and Gore under Approval.  But if the Gore/Bush race weren't remotely close, you might have voted for Nader under plurality, so you might be more comfortable approving only Nader under Approval.

Other similar strategies involve following polls leading up to an election, which would certainly be approval polls.  It's generally smart to differentiate between the two poll leaders; the simplest way to do that would be to vote for your favorite of the top two candidates in the latest poll, plus every candidate you like more.  An even better strategy, only slightly more complicated, would be to vote for every candidate you prefer to the candidate leading the latest poll, plus that top candidate if you prefer him to the current second-place candidate.  Computer simulations show that this last strategy is especially effective, and electorates that follow it would tend to have a dynamic party system rather than few parties with entrenched power.

Will Approval Voting elections be more complicated to count?

Absolutely not.  In fact, if anything, counting Approval elections will be simpler to count than plurality is now, since the vote-for-only-one restriction is removed and there will be fewer spoiled ballots.  On the other hand, a close IRV election would be an absolute nightmare.  Hand recounts would be almost impossible in elections with several candidates.  Adding newly-found ballots would require beginning the elimination process all over.  Ties or near-ties could occur at each elimination stage of IRV.  Considering what a mess Florida's 2000 election was, it's hard to imagine a similarly close IRV election.  If anything can be learned from the Florida debacle, it's that ballots must be as simple as possible to vote and to count.

Why is there so much more support for IRV than for Approval Voting?

The Center for Voting and Democracy must get much of the credit.  They have been helped by members of the Green Party who doubtless have regrets about the outcome of the 2000 U.S. presidential election.  Together they worked hard to spread awareness about the problems of plurality voting and the benefits that they expect from IRV.  The only problem with this is that IRV breaks down in practice and often does not deliver the benefits that are promised.  That is why many of the leaders of the Approval Voting movement are former IRV supporters.  They want the benefits that IRV purports to offer, but they have learned that Approval Voting does a better job of delivering those same benefits!

Why don't you just join the IRV movement instead of starting a counterproductive movement for a system no one's heard of?

Some Approval advocates do consider IRV an improvement over plurality.  Others point to many IRV disadvantages that plurality doesn't have, such as relatively complicated ballots and difficulty of hand recounts, not to mention nonmonotonicity, inconsistency and non-reverse-symmetry.  But the most important reason not to join IRV supporters is that IRV retains the spoiler and lesser-of-two-evils problems of plurality and would not level the political playing field to allow a true multi-party system (Australia's lower house is elected using IRV and has only two powerful parties).  In fact, very few voting theorists take IRV seriously due to these and other flaws.  Since changing voting systems is a "big deal" requiring a lot of effort, it makes sense to insist that a new voting system have a firm theoretical basis.  Since Approval Voting does and is highly practical too, it seems more likely to be ultimately adopted than IRV once more people know about it—and we are determined to make sure that people do know about Approval Voting!