Capturing Government Policy on the Left-Right Scale: Evidence from the United Kingdom, 1956-2006
| Published date | 01 December 2009 |
| Author | Armèn Hakhverdian |
| Date | 01 December 2009 |
| DOI | 10.1111/j.1467-9248.2008.00763.x |
| Subject Matter | Article |
P O L I T I C A L S T U D I E S : 2 0 0 9 VO L 5 7 , 7 2 0 – 7 4 5
doi: 10.1111/j.1467-9248.2008.00763.x
Capturing Government Policy on the
Left–Right Scale: Evidence from the United
Kingdom, 1956–2006post_763720..745
Armèn Hakhverdian
University of Oxford
The left–right scheme is the most widely used and parsimonious representation of political competition.
Yet, long time series of the left–right position of governments are sparse. Existing methods are of limited
use in dynamic settings due to insufficient time points which hinders the proper specification of
time-series regressions. This article analyses legislative speeches in order to construct an annual left–right
policy variable for Britain from 1956 to 2006. Using a recently developed content analysis tool, known
as Wordscores, it is shown that speeches yield valid and reliable estimates for the left–right position of
British government policy. Long time series such as the one proposed in this article are vital to building
dynamic macro-level models of politics. This measure is cross-validated with four independent sources:
(1) it compares well to expert surveys; (2) a rightward trend is found in post-war British government
policy; (3) Conservative governments are found to be more right wing in their policy outputs than
Labour governments; (4) conventional accounts of British post-war politics support the pattern of
government policy movement on the left–right scale.
The measurement of the policy positions of actors has always ranked high on the
political research agenda. Valid and reliable policy estimates of parties, govern-
ments and voters are indispensable for testing theories of political competition,
representation, coalition bargaining, policy responsiveness and cleavage politics, to
name but a few. Of all these actors, the most powerful is, arguably, the govern-
ment. This certainly holds for majoritarian democracies such as the United
Kingdom, where the Cabinet ‘wields vast amounts of political power to rule as the
representative of and in the interest of a majority’ (Lijphart, 1999, pp. 10–1). Yet
long-term variables of government policy in Britain are sparse.1 The construction
of such variables is important, since it enables researchers to use time-series
methods to study statistically the movement of political phenomena over time.
The power of dynamic models has long been acknowledged and its use among
researchers has risen steeply in recent years (DeBoef and Keele, 2008).
The spatial representation of competition goes back to the seminal contributions
of Duncan Black (1958), Anthony Downs (1957), Harold Hotelling (1929) and
Arthur Smithies (1941). Ever since, it has become common to refer to where
actors stand on substantive matters such as the economy, immigration, the envi-
ronment and so on. But, as Kenneth Benoit and Michael Laver (2006, p. 11) point
out,‘the very notion of position implies distance’ and ‘the very notion of distance
© 2008 The Author. Journal compilation © 2008 Political Studies Association
G OV E R N M E N T P O L I C Y O N T H E L E F T – R I G H T S C A L E
721
implies movement’. In other words, if we are to get the most out of theories of
political competition and the like, we have to move from static to dynamic
models, and accordingly move from cross-sectional to longitudinal policy vari-
ables. For example, some static models calculate the distance between elites and
voters on a predetermined axis of competition (Norris and Lovenduski, 2004;
Powell, 2000). But do elites then adjust to voter movements or vice versa? Static
models are ill equipped to deal with such reciprocal relationships. Time-series
models on the other hand allow for a temporal ordering of variables in order to
test whether policy at time t is related to independent variables at time t-1.
Longitudinal policy variables thus constitute a stepping stone to building dynamic
macro-level models of politics.
The benchmark against which movement is to be captured is the left–right scale,
the use of which is traditionally traced back to the 1789 French Constituent
Assembly. In those days, representatives on the right-hand side of the president
(Côté Droit) were labelled conservative, while those on the president’s left (Côté
Gauche) were labelled ‘destructive’ (Carlyle, 1871, p. 192). Nowadays, the substan-
tive meanings of the political left and right are linked to ‘whether one supports
or opposes social change in an egalitarian direction’ (Inglehart, 1990, p. 293). The
one-dimensional left–right scale serves as a summary of one’s stance on important
political issues and has proven highly suitable to describe party competition in
Britain ( Benoit and Laver, 2006; Budge and Klingemann, 2001; Webb, 2000). Yet
the only available longitudinal left–right measure of British government policy
(Klingemann et al., 2006) is of little use in dynamic models due to its limited
number of time points.
The main contribution of this article is methodological, as it elaborates on a
method of measuring the left–right score of government policy over a long
period of time (around 50 annual observations).While the focus is on Britain, the
method can easily be replicated for other countries. The first section reviews
other existing long-term variables of British government policy – one based on
policy positions, the other based on policy emphasis – and discusses where these
studies leave room for improvement. The second section elaborates on a recently
developed computerised content analysis tool, known as Wordscores (Laver et al.,
2003). Moving beyond the conventional political text, i.e. the party election
manifesto, the third and fourth sections show how Wordscores can be applied to
legislative speeches to yield equally reliable and valid results. The last section
cross-validates the new left–right policy variable with four independent sources
and concludes that it broadly conforms to established accounts of post-war British
politics.
Time Series of Government Policy
There are basically two different approaches to constructing long government
policy time series in the UK. The first approach sets out to calculate left–right
© 2008 The Author. Journal compilation © 2008 Political Studies Association
POLITICAL STUDIES: 2009, 57(4)
722
A R M È N H A K H V E R D I A N
policy positions of parties and governments, while the second approach measures
policy emphasis and prioritisation of issues. Scholars in the Downsian tradition
(Downs, 1957) place parties, governments and voters on a one-dimensional
policy axis – Downs’ own example employs an economic left–right scale that
runs from laissez-faire capitalism on the right to total state control of the
economy on the left. In this model, political elites compete by adopting dif-
ferent policy positions along the same dimension of conflict. In contrast, issue
saliency theory ( Budge and Farlie, 1983) argues that differences between parties
consist of contrasting emphases placed on different policy areas. Ian Budge
(Budge et al., 2001, p. 85) writes that ‘saliency theory gives a picture of party
competition that corresponds more closely to the intentions and strategies of
the parties themselves than an approach based on a confrontation between
specific party positions’. Party competition then is not so much about shifting
policy positions as it is about prioritisation of issues. While issue competition
and positional competition may in fact be related (Riker, 1996), they lead to
different operationalisations of government policy.
Thus far, the Comparative Manifesto Project (CMP) has composed the only
available long-term time series of the policy position of governments in Britain
(Klingemann et al., 2006).2 Their key assumption is that in one-party govern-
ments the position of the government is close if not identical to the position of
the governing party as set forth in its election manifesto (see Powell, 2000).
Others have shown that this assumption is largely met in Britain: ‘British parties
can be assumed to carry through most of their election priorities into govern-
ment and to impose them, at least in a broad sense, upon the state apparatus’
(Hofferbert and Budge, 1992, p. 181). The CMP uses human coders to classify
sentences in election manifestos into more than 50 policy categories. The
left–right score of each manifesto, and hence government, is then calculated by
subtracting the percentage of right-wing categories from the percentage of
left-wing categories. The CMP scale thus has a range of –100 (for a party
devoting its entire manifesto to left-wing issues) to 100 (for a party devoting its
complete programme to right-wing issues). Substantively, the scale ‘opposes
emphases on peaceful internationalism, welfare and government intervention on
the Left, to emphases on strong defence, free enterprise and traditional morality
on the Right’ ( Budge and Klingemann, 2001, p. 21).
Figure 1 shows the movement of British governments over time on the left–right
scale as coded by the CMP. Budge (1999, p. 6) states that the scale’s ‘broad
conformity to accepted historical interpretations helps confirm its validity’. The
Thatcher and Major governments are the most right wing, the Wilson govern-
ments are the most left wing, while the Heath and Blair governments more or less
occupy a middle ground. Furthermore, we also see that the Conservative gov-
ernments of the 1950s and early 1960s are left of centre as part of the so-called
post-war consensus in which the Conservative party ‘sought to convince electors
that they could safely be entrusted...
To continue reading
Request your trial