Making Winner Not Take All: The Case for a New Electoral System

Making Winner Not Take All: The Case for a New Electoral
System in Uganda
By Xavier Ogena
Introduction
During what seems relatively safe to think of now as the height of the Walk-to-Work
demonstrations, not to mention the-now-legendary, police brutality that attended it,
several, impassioned calls were made for the Constitution to be amended, to void to
the current ‘winner-takes-all’ nature of our political system, which was seen in some
quarters as being at least one of the causes of the demonstrations.1 It will be argued
here that the only way in which Uganda can avoid a ‘winner-takes-all’ political
system, is by having a multi-member-constituency, proportional-representation
electoral system for local-government, parliamentary and presidential elections.
The ‘winner-takes-all’ political system
The modern, democratic, political system is founded on four principles.2 These have
all been, explicitly or implicitly, recognised in the Constitution of the Republic of
Uganda (1995)3:
First, popular control of decision-makers or representatives through regular, free and
fair elections4;
Second, ‘effective choice’ through the existence of ‘procedural political freedoms’,
such as secret ballot5 ; freedom to form and join political parties, and stand for
election to a political office6; freedom of speech and expression7;
Third, political equality of voters, through universal, adult suffrage; every vote cast
counting equally; and each person having one vote, that is, no plural voting8;
Fourth, when voters or representatives are divided on a candidate or question, the
position of the majority will be adopted.9 This is the so-called majority principle.
However, these four democratic principles have been combined in the Uganda
Constitution with an electoral system that is a mixture of single-memberconstituency, plurality, or ‘first-past-the-post’ system for parliamentary and localcouncil (LC) elections, under which seats are awarded to the candidate with most
votes in the constituency; and a plurality and proportional-representation (PR)
electoral systems for presidential elections, under which
a candidate shall not be declared elected as President unless the number of votes cast in
favour of that candidate [...] is more than fifty percent of valid votes cast at the election [...]
Where at a presidential election no candidate obtains the percentage of votes specified in
clause (4) on this article [...] a second election shall be held [...] in which the two candidates
who obtained the highest number of votes shall be the only candidates.10
The use of single-member-constituency, plurality and PR electoral systems was
intended to exploit the advantage of each system where it is most needed, that is, the
production of a majority, stable government of the plurality system at localgovernment level, and an accurate reflection of public opinion of proportional
representation at the presidential level.
While the risk associated with a ‘winner-takes-all’ electoral system is significantly
reduced at local-government level ─ a fact recognized in the Constitution through its
allowing of the exclusive use there of the plurality electoral system, the ‘winnertakes-all’ electoral system par excellence ─ where service delivery is considered by
voters to be much more important than the ‘politics’ of service deliverers, where the
right to decide the sort of service to be delivered has already been granted to the
governing party by the electorate, Local Council (LC)5 and Local Council (LC)3
chairpersons often appoint as secretaries, councillors belonging to other political
parties, even where the majority of councillors belong to their party, because they are
aware that voters would approve of this, desiring as their secretaries the people who
can get the job done regardless of party colour. But it is at the presidential level that
the winner has been, as it were, increasingly, disproportionately taking all. The
constitutional provision for a ‘run-off’ between the two candidates with the highest
votes, was intended to ensure that the votes of the other candidates would be shared
out among the two candidates with the highest votes, resulting in a higher percentage
of votes counting, than would have been the case had, for example, a presidential
candidate who achieved 46 percent of popular votes cast been elected in the absence
of a run-off, leaving 54 percent of the votes having not counted in his election. Now,
as long as the president was elected by a relatively large proportion of the popular
votes cast ─ assuming, admittedly, far too generously, it could be argued, that the
Supreme Court has been right, after all, to dismiss the election petitions that sought
to overturn the results of the 1996 and 2001 presidential elections11 ─ only a small
proportion of votes cast did not count equally. But with the decline, on the whole, in
the proportion of the total votes cast achieved by the winning presidential candidate
─ 75.5% (1996), 69.40% (2001), 59.26% (2006), 68.38% (2011) ─ as well as voter
turnout ─ 72.60% (1996), 69.70% (2001), 72% (2006), 59.29% (2011)12 ─ not only an
increasingly greater proportion of the valid votes cast have not been counting
equally, but also an increasingly greater proportion of the electorate have not been
electing the president. The country needs, therefore, a system that will ensure that at
least the vast proportion of the valid votes cast count equally at every election, and
not only when Uganda ‘gets lucky’ and a presidential candidate wins by a veritable
landslide, or there is a second ballot or ‘run-off’.
A proportional-representation electoral system
Now, while the second ballot or ‘run-off’ currently provided for by the Constitution is
some form of proportional representation (PR), PR is normally used to refer to an
electoral system, in which a constituency ─ which may be as large as a country ─ is
allowed at least two representatives. The representatives are elected by each party
providing a list of its candidates, either with the order in which they should be
elected already indicated (‘closed list), or left to be decided by the voter (‘open list’),
by ranking the candidates in order of preferences, using numbers such as 1, 2, 3, etc.
The number of a party’s candidates elected by achieving the formula-determined,
quota necessary for election will depend on its share of the valid votes cast in the
constituency. The vote is counted by transferring the surplus vote of candidates
reaching the quota needed for election to the voters’ second choices, and also from
the eliminated, lowest candidate to the voters’ second choices, and so forth until all
the seats are filled. There are, of course, many, clever variations of this system in use
around the world. For example, the vote in a multi-member constituency could be
counted without transferring votes from one candidate to another; a constituency
may have a single member, with candidates being elected by being scored by voters
on a scale of o-5, 0-9, etc., and the candidate with the highest score being elected, etc.
Now, in deciding what particular electoral system, or indeed combination of electoral
systems, to adopt, the country would do well to bear in mind Rousseau’s observation
in another context:
the wise legislator does not begin by laying down laws good in themselves, but by
investigating the fitness of the people, for which they are destined, to receive them.13
It could be argued that PR electoral systems, especially one involving multi-member
constituency or constituencies, with transferrable or non-transferrable votes, might
prove to be just a bit too complicated for the significant section of Uganda’s
electorate, that can still neither read nor write, making the adoption of such a system
tantamount not to ensuring that ‘all citizens qualified to vote [...] exercise their right
to vote’.14 But, if adult literacy, defined by UNESCO as
the ability to identify, understand, interpret, create, communicate and compute, use printed
and written materials associated with varying contexts15
is anything to go by, then Uganda, with an adult literacy rate of 73.3 percent16, should
be able to adopt some form of multi-member, party-list, PR electoral system, what
with, for example, the Democratic Republic of Congo (DRC), with an adult literacy
rate of 67.2 percent17, already having a multi-member, open-list, PR electoral system.
It could also be argued that, since a multi-member-constituency, PR electoral system
does not, more often than not, only produce a party with an absolute majority in
Parliament, but the increased choice of candidates also produces a divided
legislature, with smaller parties having disproportionate influence as ‘king-makers’
in any coalition government, the parties must be not only mature enough, but also
willing to work together in coalition governments, something which the current
parties are arguably not yet able to do. But going by the recent, ‘cross-bench’ support
for opposition to the Government’s proposal to de-gazette parts of Mabira forest; the
setting of an ad hoc parliamentary committee to investigate allegations of bribetaking from foreign oil companies by certain ministers18; the re-introduction of
presidential term limits, etc., it is arguable that, despite appearances, things that
unite Uganda’s political parties far outnumber things that divide them. Put another
way, the current ideological soil is fertile enough for coalition governments. And,
since, as has already been argued, most voters are in favour of coalition governments
at local level, it would be reasonably safe to assume that they would also be in favour
of a coalition government at national level, of course, given the choice of electing one,
as is being argued here they should be. In fact, even if the soil and climate had not
been that good for a coalition government at national level, it would not be a
sufficient reason for the country not to carry out some ‘electoral engineering’ and
adopt an electoral system, that would regularly produce coalition governments, as
the long-term advantages would far outweigh any short-term disadvantages.
In fact, it is arguable that, if Uganda is to adopt a PR system, then it should use it for
all elections. While, as already argued, the ‘winner-takes-all’ nature of the current
electoral system is less of a problem at local level, it is so by ‘accident, and not by
constitutional design, as it should ideally be. In fact, there is strong, albeit anecdotal,
evidence to suggest that a LC5 or LC3 ‘cabinet’ that is made up of people affiliated to
various political organisations, not only tend to have a healthy competition among its
members, resulting in better service delivery to the electorate, but also collude to
misappropriate public funds significantly less than when it is made up of politicians
from just one party. It seems reasonable to assume that a national coalition
government would have similar end-results. Granted that a PR electoral system
would, not only make all popular votes cast count equally, but also help significantly
reduce the level of corruption, which is without question, the single, greatest threat
to Uganda’s survival as a viable state19, then any objection to such an electoral system
would seem to be dangerous idle talk.
Plus an Upper House of Parliament?
While the country is still in the market, as it were, for electoral systems, it might as
well consider acquiring for its political system a second, wholly-appointed House of
Parliament, made up, at least mostly, of distinguished senior citizens, the George
Kanyeihambas20 of the day. The members of the Senate or Upper House, would be
nominated by an independent authority, and subsequently approved by a crossparty, parliamentary committee, ensuring that the decision to appoint them is made
by elected representatives. The chamber would arguably bring several advantages to
our political system, such as it would be reasonably impartial, as they would not be
motivated by the desire to win the next election; it would also give the country
regular access to the vast experience and wisdom of people who have served the
country in various, high-level capacities as civil servants, judges, advocates,
academics, religious leaders, politicians, etc. To avoid violating the democratic
principle that policy should be made by elected representatives, the Upper House
would be only allowed to suggest changes to a bill, and delay its becoming law for a
limited period, say, not more than a year, after which it would become law, if a
majority of members of the would-be Lower House of Parliament are in favour of it,
with or without any amendments to it suggested by the Upper House. What is more,
the people would reserve the right to vote the Upper House out of existence in a
referendum, if they so wish.
Conclusion
There is undeniable evidence that, as the margin between votes cast for the winning
candidates and votes cast for the losing candidates in Uganda’s elections narrows, an
increasing proportion of popular votes cast have actually not been counting in
electing members of parliament or the president. The fact that more votes cast count
at local than at national level, is so due to ‘accidents’ of LC5 and LC3 chairpersons,
being mindful of voters’ wish and the need for self-preservation, allocating secretarial
positions at least not entirely on the basis of political affiliation. Unfortunately, that
has not been the case at national level. The provision for a run-off for presidential
election, while being a very dilute form of PR, does not by any measure go far
enough, because it means that a presidential candidate can win with, say, 51 percent
of the popular votes cast for as long as possible without triggering off a ‘run-off’,
leaving 49 percent of the votes cast not counting, and those who cast them ‘without a
president’.
It would not do to argue that an elected president, and indeed any representative, is
not only expected to serve the interests of all, but is, by definition, also personally
interested in serving the interests, not only of the majority of constituents who voted
for him or her at the last election, but also the minority, whose support he or she
would need to win, come the next general elections. This is, of course, a version of the
so-called ‘virtual representation’ theory21, which says that it does not really matter if
the votes of the minority or majority did not count at an election, as long as the
elected representative(s) take into account the interests of the majority or minority
that did not vote for the representative(s). But, if virtual representation is indeed
practically the same thing as actual representation, then why not just grant actual
representation and have the argument over and done with, if there is a demand for
it? That was, in fact, T. B. Macaulay’s response when the ‘virtual representation’
theory was used in Britain in 1770s against the demand by her 13 American colonies
for direct representation in the British Parliament.22
It is undeniable that the vast majority of Ugandans want actual representation, and
not ‘virtual’ representation, if the Walk-to-Work demonstrations and the desire by
the electorate for councillors to be allocated secretarial positions, whether they are of
the same party as the LC5 or LC3 chairperson or not, are anything to go by.
Therefore, it is high time the country adopted an electoral system that always ensures
that every vote counts equally, which is the multi-member-constituency, PR system.
And while Uganda is still at it, let Uganda also amend her Constitution to provide for
an Upper House of Parliament, giving the country’s legislators regular access to the
minds of the country’s most distinguished citizens.
Notes and References
1
New Vision, May 19, 2011, Comment by R. Kiboijana; May 20, 2011, Letter by Dr J. Serwadda; May 21,
2011, Letter by C. Lokwiya. Also available at http://www.newvision.co.ug
2
Mayo, H. B., An Introduction to Democratic Theory, New York, Oxford University Press, 1960.
3
Uganda Government, The Constitution of the Republic of Uganda, Entebbe, 1995.
4
Ibid., Article 61(a).
5
Ibid., Article 68(1).
6
Ibid., Articles 29(1)(d)-(e), 70(1)(c), 72(1).
7
Ibid., Article 29(1)(a).
8
Ibid., Articles 59(1), 68(2).
9
Ibid., Articles 89(1), 103(6).
10
Uganda Government, ibid., Article 103(4)-(5).
11
Democracy Monitoring Group, Money in Politics: Pervasive Vote Buying in Uganda Elections, Kampala,
2011, pp. 5, 11; also available at http://www.demgroup.org/node/69
See also Sekaggya, M., Uganda: Management of Elections: A Study by AfriMAP and The Open Society
Initiative for East Africa (OSIEA), Nairobi, 2010, pp. 12-13; also available at
http://www.reliefweb.int/node/369964
12
http://www.ec.or.ug/eresults
13
Rousseau, J. J., The Social Contract (1762), in Rousseau, J. J., The Social Contract and Discourses, trans.
Cole, G. D. H., London and New York, Everyman, 1953, p. 35.
14
Uganda Government, ibid., Article 59(3).
15
http://www.unesdoc.unesco.org/images/0016/0016370e.pdf
16
http://www.uis.unesco.org/literacy
17
Ibid.
18
Sunday Monitor, October 16, 2011, pp. 3-4. Also available at http://www.monitor.co.ug
19
See the present writer’s ‘Fighting Corruption Using Morality’, available on this Website.
20
Professor George Kanyeihamba is an academic and retired judge of the Supreme Court of Uganda.
21
Mayo, H. B., ibid., p.136.
22
Quoted in ibid., p. 135-136.