Customer Reviews

Gaming the Vote: Why Elections Aren't Fair (and What We Can Do About It)

5 star:		(11)
4 star:		(6)
3 star:		(0)
2 star:		(1)
1 star:		(1)

 

 

The book is extremely well-written, and a joy to read. It would be highly recommended, except for two fatal flaws discussed below.

Poundstone's latest book deals with an issue that is fundamental to democracy, yet almost totally ignored in the U.S. While many books focus on the role of money in elections, or voter registration, or voting machine integrity, relatively few popularly written books have tackled the more fundamental question of how votes get translated into representation. This is not a question of voting machine technology, but of logic. Most Americans are remarkably unaware of the variety of voting methods available, nor of the fact that the plurality voting method that predominates in the U.S. is not the norm among modern democracies, and, in fact, is probably the most problematic of all voting methods.

Americans generally accept as inevitable that if more than two candidates are in a race, vote splitting may cause a candidate that the majority oppose to be declared elected. Poundstone points out that it doesn't have to be that way. For hundreds of years thoughtful individuals have proposed alternative means of finding majority winners, that avoid this problem. Voting methods that allow voters to rank candidates in order of preference, for example, were first proposed over 150 years ago and have been used for government elections around the world for generations. He discusses the history of methods such as the borda count, condorcet pairwise comparisons, approval voting, and instant runoff voting.

Poundstone approaches the subject by telling stories about the key people involved (both historic and contemporary), making the history and theory of voting into a fascinating and compelling tale. His book avoids the technical formula-laden jargon of voting theory texts, but does justice to the concepts. He manages to present Kenneth Arrow's Impossibility Theorem (often summarized as "there is no such thing as a perfect voting method") in a way that makes it both understandable and interesting.

However, the book suffers from two fundamental shortcomings, that prompt me to give a poor overall rating. First is the fact that Poundstone focuses almost exclusively on the question of how to elect an executive, single seat office, as if this was the core problem we face. He gives scant attention to the single biggest issue of voting in democracies, that of how to achieve fair representation in legislative bodies. He discusses proportional representation in just a few pages, and never really tackles the problems inherent in all of the winner-take-all election methods that he spends the rest of the book discussing.

The other fundamental failing of the book is his championing the assertions of advocates of one particular reform as immune from the paradoxes and dilemmas facing all other voting methods. He simply accepts the claim that Range Voting, a theoretical method in which voters can give a score to each candidate, can avoid the dilemmas and tactical manipulation. Poundstone was either unaware, or chose to ignore the analysis of Nicolaus Tideman, in his 2006 book, "Collective Decisions and Voting," which led Tideman to place Range Voting on the list of "unacceptable" voting methods, because of how prone it is to strategic manipulation.

Unfortunately, this shortcoming misdirects people in the key second part of his subtitle -- "what to do about it." Readers truly taking Poundstone seriously are likely to end up banging their heads against the wall -- and perhaps making them more frustrated and alienated than ever.

Most books that attempt to propose new ways of carrying out elections are nothing more than sour grapes: "Since *my* guy didn't win the last election, the system is obviously flawed and should be overhauled." Therefore, most of these sort of books are a waste of time.

This one, however, is simply brilliant.

Instead of approaching the subject through party results, Poundstone instead takes a historical walk through many different voting schemes in terms of the mathematical theory behind them. Don't be scared by the word "mathematical", by the way...Poundstone not only steers clear of intense mathematics but also provides a simple glossary to help you remember something you may have forgotten from earlier pages. While I think I can determine his political leanings from a couple of different allusions, he makes such a good argument and has such an engaging style of writing that it doesn't matter. There were several times when I noticed a flaw in the argumentation and Poundstone responds to the particular question on the VERY NEXT page. Any author that can read the mind of an informed reader is doing a good job indeed. :)

In conclusion, anyone who's interested in the process of voting should read this one because it's the best of its kind.

I don't know about you but I sure am frustrated by the choices we have been presented with during the current Presidential election cycle. Over the years I have observed that regardless of political philosophy the first candidates to be eliminated during the primary season are the ones with ideas. In addition, the frequent appearance of so-called "spoiler" candidates in the both primaries and general elections very often frustrate the will of the people. Voters are frequently heard to mumble "there must be a better way." Well maybe, just maybe, there is. In "Gaming The Vote: Why Elections Aren't Fair (and What We Can Do About it) author William Poundstone considers these issues and presents for your consideration at least a half dozen possible alternatives to our current system of plurality voting. Some methods are clearly better than others but the ideas offered in "Gaming The Vote" will definitely get you thinking about the problems voters face in selecting their leaders.

The overwhelming majority of elections conducted in this nation utilize the method known as plurality voting. Plurality voting is not very complicated and works very well when there are only two candidates. For all intents and purposes, whoever gets the most votes wins. That is fine and dandy until a third or a fourth candidate enters the race. That is when a phenomenon known as "vote splitting" occurs. The end result can be what we all saw in Florida in the 2000 Presidential election. Independent candidate Ralph Nader siphoned off just enough votes from Al Gore to cost him a victory in Florida and denied him the Presidency. So just what are the alternatives? Is there really any method of voting out there that is fair and fool-proof?

William Poundstone examines several voting methods that have been developed over the years. Most of these have been deemed statistically "unacceptable" by the experts who study these things. There are simply too many ways to manipulate the results. Both "Borda Count' (introduced in France in 1784) and "Condorcet Voting" would fall into this category. You will also discover that the same appears to be true for something called "Cumulative Voting" and yet another system called "Approval Voting". It is interesting to note that there are a couple of voting alternatives that are being heavily promoted these days. "Instant Runoff Voting" ranks the candidates in order of preference while "Range Voting" offers voters the opportunity to rate all candidates based on a scale of 0 to 10. You may not even realize it but Amazon reviewers use "range voting" every time they choose to submit a review. What you will discover in "Gaming The Vote" is that reforming our elections is a very tricky proposition indeed. There are so many factors to consider and no one can be absolutely certain that any election system is fool-proof. And as another reviewer has aptly pointed out William Poundstone only discusses the election of executives in his book. Whether any of these methods could work at the state and local level in races for the legislature or city council is really unclear.

When all is said and done I really do think that "Gaming the Vote: Why Elections Aren't Fair (and What We Can Do About It)" is a book worth reading. However, if you are looking for definitive answers to the problem of conducting elections in this country then you will likely be disappointed. For most folks, "Gaming The Vote" will merely serve as an introduction to those methods that could one day make our elections fairer and the results more indicative of the voters wishes than they are today. People need to arm themselves with this kind of useful information if we are ever going to bring meaningful reform to our elections. "Gaming The Vote" is a well written book that should prove quite interesting to a wide range of readers. Recommended.

Received for the holidays, and quickly devoured, William Poundstone's Gaming the Vote. I absolutely loved this book, as its layout mirrored my own exploration of the issue of fair voting, and was filled with the sorts of great examples that grab and focus attention, written in a way that brings out the very human characters that underlie what seems, on the surface, to be a dryly academic topic. Basically, it's the book I would love to write, if I had Poundstone's experience, training, and writing ability.


Starting with an anectdote about Kurt Godel (one of my other favorite books is Douglas Hofstadter's Godel, Escher, Bach), it lays out all the problems with our current voting system, with examples scatterd across American history, from Lincoln (and his 39.8% popular-vote win) all the way up to 2006's midterm elections. Once the problem is laid out, Poundstone starts searching for solutions, again giving us great real-life examples full of all-too-human characters, including Kenneth Arrow (winner of the Nobel Prize in Economics), Lewis Carrol (author of Alice in Wonderland), and Lee Atwater (Nixon's expert slime-man.) Along the way, he points out all the downsides of IRV, Condorcet's method, Approval voting, and even his favorite horse, Range voting. Most compelling is his diagram of Baysian Regret of various voting methods (basically, the sum total of how upset people are about an election's result), which strongly suggest that he's right: Range voting scores best.


The book ends by calling me out; asking if any municipality would be willing to step up to the plate and put Range voting into law. Well Mr. Poundstone, I accept your challenge.

(crossposted from "The Least of all Evils" voting-reform blog, "leastevil" on blogspot)

I've read a number of books on voting systems, most of which are very dry and technical. This book manages to explain a lot of things in a well-written, readable form, and I recommend it highly.

The book has two main sections: in one, several elections in the past history of the United States are discussed to show how our electoral mechanism can go awry. This could be very amusing if it weren't so tragic, and sets the stage for the second part. This part describes alternatives and homes in on a method, "range voting," which solves many of the difficulties associated with our electoral system.

Range voting is familiar to Amazon customers, because it is the way books are rated here: voters give a rating (1 to 5 stars on Amazon, it could be 1-10 or 1-100, for example, in another type of election) and the ratings are averaged, with the highest rating winning. It is a system where you can sincerely rate a number of candidates and be sure that your vote will not hurt one you favor against one you dislike or help someone you disfavor against one you prefer. It never has been used in political elections, but has been used in many other contexts, and not just on Amazon.

Unfortunately, as Poundstone mentions, there has come to be a controversy where two different electoral system reforms have been set against each other: range voting and "instant runoff voting" have their advocates, each cutting down the other. Each would have advantages over the present system, but range voting has in my eyes slightly more, as IRV does generate some odd paradoxes (discussed in the book under the term "nonmonotonicity") in some situations. Whether these situations would often arise is hard to determine, because IRV has only been used for any length of time in one place: Australia. Range voting seems to be harder to implement, but avoids these paradoxes.

It is true, as another reviewer mentions, that the book concentrates on single-winner elections and does not go into the possibility of electing legislatures by proportional methods. This only means that that is another issue to deal with, and does not detract from the fact that many offices, such as mayors, governors, and chief executives in general are inherently single-winner, and this book is oriented toward such offices.

Kenneth Arrow's Impossibility Theorem (1948) states that no voting system can exist which satisfies the four minimal conditions of transitivity, unanimity, non-dictatorship, and independence of irrelevant alternatives. It's relatively well-known that good old-fashioned American one-man-one-vote, first-past-the-post plurality voting works fine when there are only two candidates running, but spoilers, clones, and other kinds of multiple candidates can screw things up so the final winner is not the candidate most preferred by most of the voters. Other voting systems get proposed to more fairly handle three-or-more-way races -- Borda, Condorcet, Single Transferable, Instant-Runoff voting -- but this book shows how each can still fail. Poundstone also looks at voting methods not covered by Arrow's premises, such as Approval and Range voting; also mentions Cumulative Voting, but spends little time on it.

I bought Poundstone's book rather than some others on the same general topic, because I was told it covered real historical examples and not just the math. It does, but not very evenly: the historical chapters are indeed interesting in places, but have a different feel from the theoretical chapters, more polemical and sometimes partisan. In mentioning the Great Figure-Skating Flip-Flop of 1995, Poundstone says "Trust me -- there wasn't [anything funny about the scoring system]. If I explained the whole voting system, you would nod your head and say, That sounds fair." Maybe so, but I'd rather you did take the time to explain it and let me nod for myself, and spend fewer pages on Lee Atwater and negative campaigning, of which I already know all I need to and more than I want to, and isn't exactly the point of the book.

Note that Poundstone is concerned almost exclusively with "voting systems" in the mathematical sense, he doesn't get into things like tampering with electronic voting machines at all. Similarly, for all the times he refers to the presidential election of 2000 it's to discuss Nader's role as a spoiler, not butterfly ballots or hanging chads, nor the disconnect between the popular vote and the electoral college. Not just OUR system, in other words, but the very theory of voting in general (albeit with virtually all examples and illustrations taken from US history)

As far as the "What We Can Do About It" part of the subtitle goes, there's not really very much about that. Poundstone has his clear favorite system (Range Voting) but admits it isn't likely to get much traction, maybe Instant-Runoff Voting is the best we can work for. He says there's not much point in writing to incumbent politicians, because they're too vested in the current system, but if you do want to he recommends writing to Senator McCain or Senator Obama -- this alone makes the book feel dated beyond its years.

Interesting, very readable, explains things I hadn't understood before; good notes, excellent bibliography. But all that said, it fails to change my life.

I consider this to be one of the most important books I have ever read. And for the price and approachable length, it is a crime of negligence not to pick it up. I say this having read many books on vote theory including Tideman's Collective Decisions. There is little more paramount than the method we choose to elect our leaders. Even more, Poundstone took what could easily have been a dry book for the typical reader and instead sculpted it into a page-turner.

Poundstone set out his purpose in this book. He focused on singe-seat voting systems while putting aside issues like election fraud and campaign financing. You can only do so much with one book, and I make an informed statement when I say that the voting system is the most important component. But Poundstone's approach does much more than detail the voting systems. He has you experience the people that were affected by the system, countless who gamed it, and those who tried to make voting fairer. How many average people would have thought that the Alice in Wonderland author Lewis Carrol (Charles Dodgson) had his hand in voting systems? Poundstone is truly skilled at the way he guides you through his stories with gripping narrative.

So what do you do after you read such a book? You learn more. You tell others. Ignorance is the obstacle to overcome. Fairer voting systems like approval and range voting need to be talked about. When others complain about the political system and vote splitting, you'll know what to tell them after you read this book. No longer can we get away with suggesting that there are no answers. And the stakes are too high to be fatalistic or apathetic. One is irresponsible to take this approach in the face of real solutions.

We advocates of voting system reform are an unhappy lot. At some point each of us had his eyes opened to how the commonly used voting systems are awful and unfair, but our warning cries are met with only quizzical looks. Few people have a clue that the voting system problem exists (in the US, you'll hear far more talk about voting machines and the Electoral College, red herrings both), and, when made aware, they don't understand the urgency. William Poundstone's "Gaming the Vote" is an excellent resource for voting system reformers, presenting the colorful history of voting systems (with numerous amusing anecdotes) and the checkered history of democracy in the US (with numerous spoiled elections). This, combined with the assessments of the various voting systems, including the opinions of various voting theorists, is convincing stuff. To boot, Poundstone is an excellent, entertaining writer. "Gaming the Vote" has joined my list of must-read eye-openers that I recommend to friends.

The book could benefit from some additions, firstly an overall model of the voting process. Voting does not usually occur suddenly and unexpectedly but in an iterated cycle of discussion, expression of preference, and vote counting, in phases: self-selection into groups (e.g. cliques, voting districts or parties), nomination, runoffs, and final selection. The early phases occur with some expectation of the later phases, and certainly with historical awareness, so the system evolves continuously (e.g. via gerrymandering).

One of those early phases is establishing a constitution, but this gets no mention. Because even the technically fairest election can produce evil results ("Democracy is when two wolves and a lamb vote on what to eat for dinner"), we typically have constitutions to set the voting rules, e.g. limiting the range of questions that can be voted upon. Behind a "veil of ignorance" as to the details of future votes, we can agree on voting rules so that votes will inflict minimal harm. The book mentions Social Choice theory often, but Public Choice only once, and merely as the journal where voting theorist Donald Saari happened to publish an article. Saari gets practically a whole chapter, but Gordon Tullock and Nobel Prize-winner James Buchanan, the authors of Public Choice theory's landmark work "The Calculus of Consent" are completely missing. CALCULUS OF CONSENT, THE (Tullock, Gordon. Selections. V. 2.)

Poundstone attributes much of the current election evil to the rise of political consultants, who consciously coordinate negative publicity, spoilers, and other unfair shenanigans. He does not mention that their rise may be not a cause but a result of the increased government spending and power that began with FDR (the book focuses instead on the other Roosevelt). As the stakes grow higher, so do the campaign budgets. "As long as there is power to be bought, there will be money to buy it."

Proportional voting is mentioned only briefly, but surely deserves more. Under what conditions it is more appropriate than single-winner elections? Some decisions are made only once, while others (e.g. for a legislature) are for representatives who themselves will vote repeatedly. Even for the seemingly one-off cases of choosing a restaurant or a movie, if the same group of friends sees many movies together, surely they would want a voting system that occasionally acknowledges the minority preference? Achieving *consensus* is important for any group that wants to stay friends. Contrast this with modern American politics, with its polarization and ongoing bitterness.

It is ironic that voting theorists still disagree vehemently on the proper voting system for a given set of circumstances. It is doubly ironic that Range Voting is a simplified version of the Comparison Matrix, which is used to evaluate candidates in just such situations when people are emotionally attached to their favorites. Comparison matrices remove some of the emotion by having evaluators (voters) assess each candidate against a weighted set of criteria. The weighted scores are averaged to producing a single score: a range vote. The book would benefit from the addition of such a matrix summarizing how various voting schemes satisfy various fairness criteria.

As books of non fiction go, Gaming the Vote begins with an amazing story. William Poundstone starts the book with a telling of the 1991 governor's race in Louisiana. I never knew how corrupt politics in America could really get till the Blagojevich scandal broke. I had no idea I had been missing out on so much fun! The story of how Edwin Edwards ended up squaring off and winning against David Duke had me hooked and immediately raised my expectations from the book. And it delivered.

The book is an examination of Kenneth Arrow's Impossibility Theorem [...] i.e The Problem followed by proposals of other voting systems that try to address the issues of vote splitting which plagues the plurality voting system. Mr. Poundstone makes this admittedly dry subject funny and engaging through anecdotes aplenty. I found the Charles Dodgson/Lewis Carroll story especially enlightening. I had no idea the writer I knew as Lewis Carroll was also a philosopher mathematician. And now I know what a Condorcet winner is!

If you decide to read the book, I suggest putting the glossary at the end of the book to use - the author has condensed all the different voting systems described in four pages that help clarify the main text when you get bogged down in the jargon of IRV, Approval Voting, Condorcet Cycle etcetera.

I'm strongly recommending Gaming the Vote.

How many times have you thought about a third-party-candidate, then dismissed the idea of actually voting for them BECAUSE YOU WOULD BE THROWING AWAY YOUR VOTE. An entertaining discussion of the subtleties of various alternative voting systems which offer solutions to the vote-splitting problem takes up the bulk of the book, leading up to this conclusive ch 16 "The Way Democracy Will Be". Folks like me who always read last-chapter-first should note that ch 17 "Blue Man Coup" is kind of an afterword - if say as someone already familiar with such things you just want the punch line, read ch 16.
Author Poundstone is *SO* good at weaving in the human interest angle. This is how voting theory is made entertaining. The kickoff sentence of the book "...Grand Wizard of the Ku Klux Klan, David Duke felt he was destined for something bigger." You now know you're going to be titillated, and you are, but the nice theory IS woven in, so you're learning along the way. "I wish I could write like that."
Wikipedia will confirm that "Peephole Optimization" is something which a compiler can do to improve a SMALL PART of the generated code, WITHOUT worrying about more global things. Consider that a compiler translates your source-code-intentions into code which runs, actually fairly analogous to politics translating "the will of the people" into political action.
You need to get to page 244 before you meet the phrase "..."tragedy of the commons", the "volunteer's dilimma", and the "free rider dilemma"...shirkers get the same benefit as the volunteers". So Poundstone knows fully well about these deep GLOBAL problems with the very foundations of voting and coalitions, but to widen the scope of the discussion to tackle them borders on hopeless. It is in other words to be viewed as a STRENGTH of this book that it stays FOCUSED on the "Peephole" issue of "Voting Systems", which is something one might actually hope to change, without breaking other things ! Accept the status quo regarding say who-gets-to-vote( should women vote ? ), just focus on the peephole-issue of the Voting Process Itself. Yum.
With Internet-search available to anyone the value-add in such a book is that the author has made N passes over the material, in this case N being perhaps 16 or even 32 ? The N passes, by analogy to the years on the label of some whiskey, provide SMOOTHNESS. Evidence for this disciplined approach is in the footnote/source linkage provided at the end. Pages 313..323 list the sources actually used in the book in a very small font for about 20 sources-per-page. Pages 293..311 provide footnotes linking pages of each chapter to the sources at about 30 footnotes-per-page. In other words if you are REALLY into this stuff it would be easy to drill-down on some issue. For me it's a waste of paper so the small font is fine<g>.