very sensitive to performance: few votes hinge on whether he has
done a good or a bad job. So not only do people lack the information
on which to judge performance, but relatively few are going to base
their votes on this judgment.
Perhaps also, the scope for the government to produce a good
performance is really quite modest, maybe due to its own limita-
tions. Especially after years of poor performance, a government may
simply lose faith in its own ability to make a decisive difference to
economic events.
Finally, suppose that if the government does choose to be good
it has to forgo behavior that is decidedly lucrative. Messing about
with the economy may be detrimental to ordinary citizens, but it
opens up many little niches and crannies for personal enrichment,
and for rewarding loyalty among followers. If all these opportuni-
ties are closed off, the leader has no means of maintaining loyalty.
So how does this stack up? As the quality of voter information
is made weaker, as identity politics freezes more and more votes, as
the government’s confidence in its own ability to shape events di-
Votes and Violence
27
minishes, and as the costs of forgoing bad governance are increased,
a point is reached at which facing an election simply does not dis-
cipline an incumbent politician into trying to perform well. And
if politicians can still face a reasonable chance of winning without
bothering to deliver good performance, then—and this is Tim’s
killer point—the sort of people who seek to become politicians will
change. If being honest and competent does not give you an elec-
toral advantage, then the honest and competent will be discouraged.
Crooks will replace the honest as candidates.
One depressing indicator of such a process is that democratic
politics in the countries of the bottom billion tends to attract candi-
dates with criminal records. You might reasonably expect that having
a criminal record would make running in an election a nonstarter. I
think it would in America or Britain, and indeed across most of the
rich world. But in the societies of the bottom billion it simply isn’t so.
Electors just don’t have enough information to sort out the accusa-
tions from reality: either the press is muzzled or it is too free—there is
so much mud being slung without recourse to verification that voters
discount whatever they are told. Or electors are frozen in ethnic loyal-
ties and so support their own politicians even if they are criminals.
Evidently, one reason elected office is more attractive to crimi-
nals than to the honest is that only the criminals will take advan-
tage of the opportunities for corruption. But there is sometimes a
further reason: elected office provides immunity from prosecution.
Ask yourself for whom this is particularly valuable. For the honest,
it merely protects from mischievous attacks that, in the end, they
could probably resist anyway. But for the criminal, immunity from
prosecution is likely to mean the difference between freedom and
jail. Sometimes this turns to farce. Following the Nigerian guberna-
torial elections of 2007 there was a race between the police and a vic-
torious deputy governor as to whether he could get himself sworn
in before they could reach him to arrest him. It was touch and go
whether home would be jail or the deputy governor’s villa.
28
WARS, GUNS, AND VOTES
If honest people realize that they are unlikely to win and so do
not come forward as candidates, then voters lack even the choice of
a decent leader. There is really not much point in finding out about
the candidates, and this adds a further twist to the vicious circle.
Tim’s analysis is about at the frontier of serious work on democ-
racy. But even Tim’s world is thoroughly sedate when compared to the
election campaigns familiar to the bottom billion. Basically, in Tim’s
world politicians still play by the rules; it is just that they face badly informed electors. Again, I put myself in the situation of an old autocrat
now having to retain power in a democracy. What options do I face?
Hard as it is to bear, I have to be honest with myself that my people
do not love me. Far from being grateful for the wonders that I have
achieved, they may increasingly be aware that under my long rule our
country has stagnated, whereas elsewhere initially similar countries
have transformed themselves. There are even a few cogent voices out
there explaining why this is my fault. I shake my head in disbelief that
it has come to this, seize my gold pen, and start listing the options. I
decide to be systematic, in each case putting down the pros and cons.
Option 1: Turn over a new leaf and become a good
government
Pros: This is probably what most people want. It would make
a change, I might start feeling better about myself, and I might
even leave a legacy that my children could be proud of.
Cons: I haven’t much of an idea how to do it. The skills I have
developed over the years are quite different, essentially how
to retain power through shuffling a huge number of people
around a patronage trough. My God, I might have to read those
damned donor reports. And even if I worked out what needed
to change, the civil service isn’t up to implementing it. After all,
Votes and Violence
29
I’ve spent years making sure that anyone who was exceptional
or even honest was squeezed out: honest people cannot easily
be controlled. Yes, I too read Herodotus. Even worse, reform
might be dangerous. My friends, the parasitic sycophants with
whom I have surrounded myself, might not put up with it: they
might decide to replace me in a palace coup. They would proba-
bly dress it up to the outside world as reform! But suppose I did
it; suppose I actually delivered good government. Would I get
reelected? I start to think through all those rich-country politi-
cal leaders who, over the years, have met me, often lecturing me
on the need for good governance. What became of them? What
was their record of electoral success? I do a rough tally—they
seemed to win their own elections only around 45 percent of the
time. So, if I pull it off, I have a 45 percent chance of winning.
Option 1 does not seem that attractive, whatever the foreign
ambassadors might imply with their incessant homilies about good
governance. The evident difficulties of governing well make your
electoral task daunting relative to that of your fortunate rich-coun-
try counterparts. You contemplate having a comforting sulk about
the inequities of life, but put self-indulgence behind you: you have
to make the best of what you have. And then it strikes you that com-
pared with your rich-country counterparts you have one potential
advantage. Although you are going to have to win an election, you
are not subject to much effective scrutiny as to how you go about it.
Does this open up any strategies that might enable you to win de-
spite continuing to be a bad government?
Option 2: Lie to electors
Pros: You control most of the media, so it is relatively easy. What is more, your citizens have neither education nor good refer-
30
WARS, GUNS, AND VOTES
ence points by which to tell how bad things really are. So you
can tell them how fortunate they are to have you as president.
Cons: You have been doing this for years and so people heavily
discount anything you say.
On balance, although lying seems to be worth doing, you sim-
ply cannot rely on it to deliver victory.
Option 3: Scapegoat a minority
Pros: This one works! You can blame either minorities within
your country or foreign governments for all your problems:
that President Mugabe of Zimbabwe is a role model. The
politics of hatred has a long and electorally pretty successful
pedigree. Most of the societies of the bottom billion have un-
popular ethnic minorities to pillory, and failing all else you
can always blame America. You can also promise favoritism
for your own group.
Cons: Some of your best friends are from ethnic minorities. In
fact, they have been funding you for years in return for favors.
You prefer business people from ethnic minorities because
however rich they become, they cannot challenge you politi-
cally. It is the core ethnic groups that you want to keep out of
business. If you scare the minorities too badly they will move
their money out.
So, although scapegoating works, beyond a certain point it gets
rather costly.
Votes and Violence
31
Option 4: Bribery
Pros: Bribery plays to one of your key advantages over the op-
position—you have more money.
Cons: Can you trust people to honor the deal? If you pay them
money will they actually vote for you? After all, there are some
pretty unscrupulous people out there.
On balance you are not sure. If only there was some reliable
research evidence! You search the Net and stumble on something
by someone called Pedro Vicente, of the Centre for the Study of Af-
rican Economies at Oxford. You start to skim it and rapidly become
riveted, as well you might. Pedro has conducted a randomized, con-
trolled experiment on electoral bribery in São Tomé and Principe,
which is just off the coast from your own state.
Tiresomely, you find that the main thrust of his research is to
investigate whether bribery can be countered. Then, however, you
find the pertinent gem. In some districts bribery was restrained by
external scrutiny, whereas in others it was not. Systematically, the
candidate who was bribing gathered more votes in those districts
where bribery was not restrained. Bribery works!
In fact, bribery comes in two modes: retail and wholesale. Re-
tail bribery is expensive and difficult but may still be worthwhile. Its
advantage is that you can target pockets of voters who are critical for
success. For example, President Moi of Kenya managed by astute
attention to key votes to win an election with only 37 percent sup-
port. Why doesn’t bribery backfire? If the British Labour Party was
caught offering money to individual voters in exchange for their
support the electoral damage would be massive. But in many so-
cieties elections are viewed differently. Politicians deliver nothing
during their period in office, and so people expect that during the
one brief moment when they exert some power politicians should
32
WARS, GUNS, AND VOTES
dispense patronage, and hard cash in the pocket is better than prom-
ises. But even if politicians can offer bribes without provoking criti-
cism, how can they enforce the deal? After all, the vote is secret.
What is to stop voters from accepting the money and then voting
for the opposition?
In Kenya the opposition recognized that telling people that tak-
ing bribes was wrong would be a vote loser and so did not even
attempt it. Instead they proposed that people should take the bribe
from the government but vote for the opposition. Why is such an
opposition message not a very effective counter? The government
has two points of discipline. One, paradoxically, is morality: often,
ordinary decent people feel bad if they take someone’s money but
then renege on their undertaking. The opposition argument that
one wrong neutralizes the other is smart, but it is morally a little
tortured. The other is fear of detection: how secret is the ballot? In
Zimbabwe President Mugabe’s street boys spread the word that the
government would know how votes were cast, and in the prevail-
ing conditions of misgovernance this could not be treated as an idle
threat.
It is not as if one individual vote will determine the choice of
government: realistically, it will have no effect whatsoever on the
outcome. And so even if there is only a small risk that a vote against
the government may be detected, it may not be worth taking. It
might land the voter in trouble and so be irresponsible for an adult
struggling to bring up a family in conditions that are already dire.
Having got this far in his train of thought, the president will
perhaps be counting his fortune. How much does it cost to bribe
the typical voter, how many votes does he need to buy, and how
much can he afford? In some societies he will sit back contentedly:
this strategy is within his budget. In others he may be pondering
whether there is a cheaper way of buying votes. There is: it is time
for wholesale bribery.
Wholesale bribery works by paying for votes delivered in blocs
Votes and Violence
33
rather than individually. Bloc voting is very common in impover-
ished traditional rural societies: the local big shot gives the lead and
his advice is not seriously questioned. When votes are counted it is
common for many villages to have voted 100 percent for one candi-
date. If the big shot determines the voting, it is obviously cheaper to
buy his support directly rather than try to attract individual votes.
Overall, you conclude that bribery is your kind of strategy. The
only problem is whether you have enough money to win with it.
This inspires you to carry on thinking.
Option 5: Intimidation
Most politicians try to ingratiate themselves with voters, but a
radically different technique is to intimidate them.
Pros: Most people are not particularly brave, and when con-
fronted by thugs threatening personal violence, they back
down rather than stand up for themselves. One big advantage
of intimidation is that even if you cannot observe how people
vote, you can observe whether they vote. Given that you are
playing in identity politics, you know perfectly well the iden-
tity of those who intend to vote for your opponent. So you can
threaten them that if they vote they will suffer. Does it work?
In Kenya President Moi used it to force a mass of Kikuyu living
in
the Rift Valley who were likely to vote against him to move.
In moving they went to areas where they were not registered
to vote, so he no longer had to worry about them. He claimed
the violence was just a local dispute about land rights, but a
careful statistical study by two Kenyan researchers, Mwangi
Kimenyi and Njuguna Ndung’u, gave the lie to that one. They
show that “the central rationale of the violence appears to have
been to maintain the political and economic status quo in the
34
WARS, GUNS, AND VOTES
region during the run-up to the general elections.”* Indeed, the
bows and arrows ostensibly used by irate and untamed tribes-
men turned out to have been manufactured in East Asia and
presumably planted by the government. You also recall that
President Mugabe has not been reticent in using intimidation
against opposition voters.
Cons: If politics turns violent there is no knowing where it might
stop. The other side might turn violent. After all, the other side
has the advantage of numbers: if they didn’t, you would not
have to worry about winning the election. You don’t want to
risk losing a contest in violence.
Overall, violence might turn out to be a can of worms. The oppo-
sition might be even more violent than you are. This is not reason for
not doing it: you may well need to do it simply to counter the violence
that is coming from the opposition, who are, after all, making the same
calculation. But violence may not be enough to ensure that you win.
Option 6: Restrict the field to exclude the strongest
candidates
Pros: This is particularly appealing because not only do you in-
crease your chances of winning but you hit directly at the people
you most hate: your personal opponents. You have to find some
reason for excluding them, but that is not particularly difficult.
You can accuse them of corruption—after all, it is quite likely
to be true. A delicious nuance is that since the donors are always
* Mwangi Kimenyi and Njuguna Ndung’u, “Sporadic Ethnic Violence: Why Has Ke-
nya Not Experienced a Full-Blown Civil War?” in Understanding Civil War (Volume 1: Africa), ed. Paul Collier and Nicholas Sambanis (Washington, D.C.: World Bank, 2005).
Votes and Violence
35
urging you to be tougher on corruption, they can scarcely object
to this option. Even the challengers to those international role
Paul Collier - Wars, Guns, and Votes Page 4
