Published online by Cambridge University Press: 07 December 2012
This article sets out and tests a theory of public policy investment – how democratic governments seek to enhance their chances of re-election by managing a portfolio of policy priorities for the public, analogous to the relationship between investment manager and client. Governments choose policies that yield returns the public values; and rebalance their policy priorities later to adjust risk and stabilize return. Do the public reward returns to policy capital or punish risky policy investments? The article investigates whether returns to policy investment guide political management and statecraft. Time-series analyses of risk and return in Britain 1971–2000 reveal that risk and return on government policy portfolios predict election outcomes, and that returns, risk profiles and the uncertainty in public signals influence the prioritization of policies.
USC Price School of Public Policy, USC Gould School of Law, University of Southern California (email: email@example.com); and School of Public Policy, University College London, respectively (email: firstname.lastname@example.org). The authors have greatly benefited from the comments of David Barker, Laura Chaqués, Scott Desposato, Keith Dowding, Patrick Dunleavy, Francesca Gains, Christoffer Green-Pedersen, Stephen Greasley, Will Jennings, Bryan Jones, Christopher Kam, George Krause, John Matsusaka, David Nixon, Oguzhan Ozbas, Anna Palau, Kaare Strøm, Andrew Whitford, Jonathan Woon and those of seminar participants at the University of California, Los Angeles; University College London; the London School of Economics; the University of Pittsburgh; the University of Hawaii-Manoa; the University of Manchester; the Montesquieu Institute, Den Haag and panel participants at the European Political Science Association 1st Annual Conference, Dublin, 2011. They also wish to thank three anonymous referees and the Editor, Hugh Ward, for their attention to and engagement with the manuscript. They acknowledge generous support from the UK Economic and Social Research Council (grant reference R105938). An appendix containing Annex 1 can be viewed at http://www.journals.cambridge.org/jps.
2 Knight, Frank, Risk, Uncertainty and Profit (Chicago: Houghton Mifflin Company, 1921)Google Scholar
Erikson, Robert, Wright, Gerald and McIver, John, Statehouse Democracy: Public Opinion and Policy in the American States (New York: Cambridge University Press, 1993)Google Scholar
Jacobs, Lawrence, The Health of Nations: Health Policy and Public Opinion in the US and Britain (Ithaca, N.Y.: Cornell University Press, 1993)Google Scholar
Erikson, Robert, MacKuen, Michael and Stimson, James, The Macro Polity (Cambridge: Cambridge University Press, 2002)Google Scholar
Soroka, Stuart and Wlezien, Christopher, Degrees of Democracy Politics, Public Opinion, and Policy (Cambridge: Cambridge University Press, 2010)Google Scholar
Jennings, William and John, Peter, ‘The Dynamics of Political Attention: Public Opinion and the Queen's Speech in the United Kingdom’, American Journal of Political Science, 53 (2010), 838–854CrossRefGoogle Scholar
Hakhverdian, Armen, ‘Political Representation and its Mechanisms: a Dynamic Left–Right Approach for the United Kingdom, 1976–2006’, British Journal of Political Science, 40 (2010), 835–856CrossRefGoogle Scholar
7 Wouter van der Brug, C. van der Eijk and Mark Franklin, The Economy and the Vote (Cambridge: Cambridge University Press, 2007)Google Scholar
8 Alt, James and Lassen, David Dreyer, ‘Transparency, Political Polarization, and Political Budget Cycles in OECD Countries’, American Journal of Political Science, 50 (2006), 530–550CrossRefGoogle Scholar
Brender, Adi and Drazen, Allan, ‘How Do Budget Deficits and Economic Growth Affect Reelection Prospects? Evidence from a Large Panel of Countries’, American Economic Review, 98 (2008), 2203–2220CrossRefGoogle Scholar
10 Frank Baumgartner and Bryan Jones, Agendas and Instability in American Politics, 2nd edn (Chicago: University of Chicago Press, 2009 )Google Scholar
Jones, Bryan and Baumgartner, Frank, The Politics of Attention: How Government Prioritizes Problems (Chicago: University of Chicago Press, 2005)Google Scholar
Baumgartner, Frank, Breunig, Christian, Green-Pedersen, Christopher, Jones, Bryan, Mortensen, Peter, Nuytemans, Michiel and Walgrave, Stefaan, ‘Punctuated Equilibrium and Institutional Friction in Comparative Perspective’, American Journal of Political Science, 53 (2009), 603–620CrossRefGoogle Scholar
11 Mortensen, Peter, Green-Pedersen, Christoffer and Breeman, Gerard, Jennings, Will, John, Peter, Timmermans, Arco, Chaques, Laura and Palau, Anna, ‘Comparing Government Agendas Executive Speeches in the Netherlands, United Kingdom, and Denmark’, Comparative Political Studies, 44 (2011), 973–1000CrossRefGoogle Scholar
Hofferbert, Richard and Budge, Ian, ‘The Party Mandate and the Westminster Model: Election Programmes and Government Spending in Britain, 1948–85’, British Journal of Political Science, 22 (1992), 151–182CrossRefGoogle Scholar
14 Budge, Ian and Fairlie, Dennis, Explaining and Predicting Elections: Issue Effects and Party Strategies in Twenty-Three Democracies (London: George Allen & Unwin, 1983)Google Scholar
16 Adams, James, Ezrow, Lawrence and Somer-Topcu, Zeynep, ‘Is Anybody Listening? Evidence that Voters Do Not Respond to European Parties’ Policy Programmes’, American Journal of Political Science, 55 (2011), 370–382CrossRefGoogle Scholar
17 Strøm, Kaare, Minority Government and Majority Rule (Cambridge: Cambridge University Press, 1990)Google Scholar
Rose, Richard and Mackie, Thomas, ‘Incumbency in Government: Asset or Liability’, in Hans Daalder and Peter Mair, eds, Western European Party Systems: Continuity and Change (Beverly Hills, Calif.: Sage, 1983)Google Scholar
18 Nanestad, Peter and Paldam, Martin, ‘The Cost of Ruling’, in Han Dorussen and Michael Taylor, eds, Economic Voting (London: Routledge, 2002)Google Scholar
19 Paul C. Light, The President's Agenda, 2nd edn (Baltimore, Md.: Johns Hopkins University Press, 1991), p. 32Google Scholar
22 Neustadt, Richard, Presidential Power and the Modern Presidents (New York: Free Press, 1990 )Google Scholar
Kernell, Samuel, Going Public: New Strategies of Presidential Leadership (Washington, D.C.: Congressional Quarterly, 1986)Google Scholar
III, George C. Edwards, At the Margins (New Haven, Conn.: Yale University Press, 1989)Google Scholar
23 Canes-Wrone, Brandice, Who Leads Whom? Presidents, Policy, and the Public (Chicago: University of Chicago Press, 2006)Google Scholar
24 Butler, David, Adonis, Andrew and Travers, Tony, Failure in British Government: The Politics of the Poll Tax (Oxford: Oxford University Press, 1994)Google Scholar
Hood, Christopher, Explaining Economic Policy Reversals (Buckingham: Open University Press, 1994)Google Scholar
Finer, Samuel, ‘Introduction: Adversary Politics and Electoral Reform’, in Samuel Finer, ed., Adversary Politics and Electoral Reform (London: Anthony Wigram, 1975), pp. 3–32Google Scholar
27 Bulpitt, Jim, Territory and Power in the United Kingdom (Manchester: Manchester University Press, 1983)Google Scholar
30 There is, of course, no political market where voters go continuously to the polls with high turnout and assess individual policies. This is arguably the central difference between studies of politics and of economics. However, we claim that citizens can express opinion that we measure through frequent polls. Governments endeavour to learn such information and take it as a price signal for its policy assets. With that information, sometimes noisier than at other times, government makes investment choices. We hypothesize that the resulting risk in the portfolio and returns to policy capital investment impacts election results.
31 Myerson, Roger, Probability Models for Economic Decisions (Belmont, Calif.: Duxbury, 2005)Google Scholar
33 Our assumption here is that the characteristic voter can participate in the political process without cost and that he or she is presented with competing views in that process such that a prioritization of policies is possible. Relaxing this assumption would mean that some voters would be systematically excluded from the political marketplace, as in the case of African-Americans before the Voting Rights Act of 1965 in the United States or when secrecy in foreign crises is maintained by governments as a matter of national security. Such conditions most likely do not apply to the average voter across our time period across all policy domains. Moreover, uncertainty about the importance of policies can be the result of voters not having enough information about the relationship between their well-being and the policies offered by the government. As we discuss below, our framework produces estimates of price signal uncertainty to address the extent to which the voters form clear priorities given the information they do acquire. It may arise because parties, the media and other elites are not sufficiently clear about the importance of policy problems, but those elites cannot discover the precise probability model for determining public priorities.
34 Eisinger, Robert, The Evolution of Presidential Polling (Cambridge: Cambridge University Press, 2003)Google Scholar
37 Rose, Richard and Davies, Phillip, Inheritance in Public Policy (New Haven, Conn.: Yale University Press, 1994)Google Scholar
38 Taken together with the claim that commitments narrow the scope for the creation of new assets, our argument has empirical validity. Between 1971 and 2000, the policy asset classes we employ in the empirical work that follows capture between 82 and 97 per cent of the quasi-sentences in the Speech from the Throne, with a very high average of 89 per cent.
44 McLean, Iain, What's Wrong with the British Constitution (Oxford: Oxford University Press, 2010)Google Scholar
Anthony King, ed., Why is Britain Becoming Harder to Govern? (London: BBC, 1976)Google Scholar
Whitehead, Phillip, The Writing on the Wall (London: Michael Joseph/Channel 4, 1985)Google Scholar
49 Bacon, Robert and Eltis, Walter, Britain's Economic Problem: Too Few Producers (London: Macmillan, 1975)Google Scholar
50 Hennessy, Peter, The Prime Minister: The Office and Its Holders since 1945 (Basingstoke, Hants.: Macmillan, 2001)Google Scholar
51 Information on the UK Policy Agendas Project can be found at http://www.policyagendas.org.uk. The project coded the Speech from the Throne according to 19 major and 225 minor topic codes, which are common to the Policy Agendas Project and apply to policies comparatively. See Baumgartner and Jones, Agendas and Instability in American Politics.
52 We thank Ian Budge for advice on interpreting and bridging the categories and Judith Bara for supplying the data.
Gerard Breeman, David Lowery, Caelesta Poppelaars, Sandra Resodihardjo, Arco Timmermans and Jouke de Vries, ‘Political Attention in a Coalition System: Analysing Queen's Speeches in the Netherlands 1945–2007’, Acta Politica, 44 (2009), 1–27CrossRefGoogle Scholar
John, Peter and Jennings, Will, ‘Punctuations and Turning Points in British Politics: The Policy Agenda of the Queen's speeches’, British Journal of Political Science, 40 (2010), 561–586CrossRefGoogle Scholar
54 Jones and Baumgartner, The Politics of Attention, p. 240Google Scholar
Soroka and Wlezien, Degrees of Democracy, pp. 65–66Google Scholar
56 The emphasis on foreign policy in the speech is high, but this reflects the traditions of the speech in drawing attention to such matters by the head of state.
59 Wlezien, Christopher, ‘On the Salience of Political Issues: The Problem With “Most Important Problem” ’, Electoral Studies, 24 (2005)Google Scholar
60 King, Anthony and Wybrow, Robert, eds, British Political Opinion 1937–2000: The Gallup Polls (London: Politico's, 2001)Google Scholar
61 Our measure of inflation is the monthly relative price index from the Global Financial Data electronic database. The measure is drawn from a hypothetical bundle of items reflecting typical household expenditure. See, for example, E. H. Phelps Brown and Sheila V. Hopkins, ‘Seven Centuries of the Price of Consumables, Compared with Builders’ Wage-Rates’, Economica, new series, 23 (1952), 296–314.
62 Unemployment is measured by the number of claimants in the United Kingdom as a percentage of the sum of employees, self-employed individuals, members of the armed forces, and government training programme enrolees and the claimant total. This measure is available from the Global Financial Data Series.
63 The daily FT 30 index was obtained from Global Financial Data and is based on an equally weighted sum of values for 30 constituent stocks chosen by editors of the Financial Times. Monthly averages for the index are employed in our analysis.
66 Kolenikov, Stas and Angeles, Gustavo, ‘The Use of Discrete Data in Principal Component Analysis: Theory, Simulations, and Applications to Socioeconomic Indices’ (Working Paper, No. WP-04-85, Carolina Population Center, University of North Carolina, 2004)Google Scholar
67 We performed an orthogonal varimax rotation of the factor loadings and noted correlations between the indicators and loadings. The macro-economic factor was highly positively correlated with inflation (0.9), oil prices (0.3), the FT 30 (0.9). The foreign policy factor was positively correlated with interstate crises (0.7) and the cost of living factor positively correlated with inflation (0.4) and oil prices (0.7). The factor we call government personnel displayed a positive correlation with ministerial resignations (0.3). ‘Ministerial resignations’ is the measured indicator with the highest degree of uniqueness, also correlating positively with the macro-economic factor (0.3). Unemployment also had high uniqueness in correlating negatively with the foreign policy factor (−0.6). Our labels were based on our own best interpretation of these correlations.
68 The mixed factor analysis is appropriate because of the varying distributions of the observed indicators. Inflation, unemployment, and FT30 measures are continuous whilst interstate crises and ministerial resignations are categorical. We use a polychoric routine in combination with the principal factors routine to form a correlation matrix (factormat) for the STATA software package to do the estimation (see Stas Kolenikov and Gustavo Angeles, ‘The Use of Discrete Data in Principal Component Analysis: Theory, Simulations, and Applications to Socioeconomic Indices’, Working Paper, No. WP-04-85, Carolina Population Center, University of North Carolina, 2004). Four positive eigenvalues were recovered. Because Equation 1 defines asset values conditional on the factors, but does not define them substantively, our approach is to use all factors that positively explain variance in the factor indicators (e.g., interstate crises). An eigenvalue describes the variance explained by the factor across all indicators. As a result, we include four factors in our analysis, which taken together fully explain the shared variance in the indicators. A likelihood-ratio test suggests that the four-factor model fits the data better than a model with no factors, χ2 (15 d.f.) = 1349.18.
70 Efron, Bradley and Tibshirani, Robert, An Introduction to the Bootstrap (New York: Chapman & Hall/CRC, 1994)Google Scholar
71 King, Gary, Tomz, Michael and Wittenberg, Jason, ‘Making the Most of Statistical Analyses: Improving Interpretation and Presentation’, American Journal of Political Science, 44 (2000), 347–361CrossRefGoogle Scholar
72 A longer and broader application of our framework to British politics will appear in our monograph, Public Policy Investment: A New Theory of British Statecraft.
75 Speech to Conservative Party Conference, 10 October 1980, see http://www.margaretthatcher.org/document/104431, accessed 4 April 2011.
76 Kavanagh, Dennis, ‘The Heath Government’, in Peter Hennessy and Anthony Seldon, eds, Ruling Performance (Oxford: Basil Blackwell, 1987)Google Scholar
Campbell, John, Edward Heath: A Biography (London: Jonathan Cape, 1993)Google Scholar
Hennessy, Peter, The Prime Minister: The Office and Its Holders since 1945 (Basingstoke: Macmillan, 2001)Google Scholar
79 Two sample t-tests for mean differences support the claim. For risk, the manifesto mean of 9.953 is significantly less than the rebalanced mean of 11.161 (|t| = 2.902, p = 0.003). For return, the manifesto mean of 9.532 is less than the rebalanced mean of 10.162 (|t| = 1.290, p = 0.101), though at a lower level of statistical significance.
80 Standard errors on the portfolio return show it is not possible to rule out the hypothesis that the return is zero in most years. This means that policy capital is largely unchanged, a very good result for a government as it maintains capital after winning an election.
82 A paired-samples t-test of the difference between other premiers and Heath provides evidence for this claim overall (t = −5.306, p = 0.000) and versus Thatcher specifically (t = −3.437, p = 0.002).
83 A paired sample t-test that Thatcher faced smaller factor loadings than a pool of other premiers in regard to social policy supports the claim for the foreign policy (|t| = 1.725, p = 0.048) and government personnel (|t| = 1.557, p = 0.065) factors.
84 A paired sample t-test that Blair faced smaller factor loadings than other premiers in regard to social policy shows this true for foreign policy (|t| = 1.517, p = 0.070) and the cost of living (|t| = 1.463, p = 0.077) factors.
85 A paired sample t-test that Heath faced smaller factor loadings than a pool of other premiers in regard to foreign and defence supports the claim for the macro-economic (|t| = 1.632, p = 0.057), foreign policy (|t| = 2.625, p = 0.007), and cost of living (|t| = 2.879, p = 0.004) factors.
86 A paired sample t-test that Thatcher faced smaller factor loadings than a pool of other premiers in regard to the macro-economic policy asset shows this for the foreign policy (|t| = 2.235, p = 0.017) and cost of living (|t| = 2.086, p = 0.023) factors.
87 A paired sample t-test that Heath faced larger factor loadings than a pool of other premiers in regard to environment and natural resources upholds the claim for the foreign policy (|t| = 2.093, p = 0.023) and cost of living (|t| = 3.137, p = 0.002) factors.
88 A paired sample t-test that Heath faced smaller factor loadings than a pool of other premiers in regard to infrastructure and investment supports the claim for the foreign policy (|t| = 1.496, p = 0.073), cost of living (|t| = 1.461, p = 0.073), and government personnel (|t| = 1.886, p = 0.035) factors.
89 A paired sample t-test that the Wilson and Callaghan governments pooled together faced smaller factor loadings than a pool of other premiers in regard to law, order and civil rights supports the claim for the foreign policy (|t| = 1.789, p = 0.042), cost of living (|t| = 2.308, p = 0.014), and government personnel (|t| = 1.657, p = 0.054) factors.
90 Manin, Bernard, Przeworski, Adam and Stokes, Susan, ‘Introduction’, in Adam Przeworski, Bernard Manin and Susan Stokes, eds, Democracy, Accountability, and Representation (New York: Cambridge University Press, 1999)Google Scholar
93 House of Commons Library, Election Statistics: UK 1918–2007, Research Paper 08/12, pp. 38–45. See http://www.parliament.uk/documents/commons/lib/research/…/rp08-012.pdf, Accessed 5 April 2011.
97 Durbin, James and Watson, Geoffrey S., ‘Testing for Serial Correlation in Least Squares Regression, I’, Biometrika, 37 (1950), 409–428Google Scholar
101 By asset, speech from the throne investment weights are distributed as follows: social policy (mean = 9.07, SD = 4.79), foreign and defence (mean = 43.55, SD = 9.57), macro-economic policy (mean = 18.62, SD = 5.25), environment and natural resources (mean = 10.00, SD = 3.74), and law, order and civil rights (mean = 7.58, SD = 3.36).
102 By asset, returns are distributed as follows: social policy (mean = −11.22, SD = 84.82), foreign and defence (mean = −9.18, SD = 29.63), macro-economic policy (mean = 29.46, SD = 344.90), environment and natural resources (mean = −1.45, SD = 11.89), and law, order and civil rights (mean = −72.54, SD = 282.07).
103 By asset, price signal uncertainty is distributed as follows: social policy (mean = 186.40, SD = 375.22), foreign and defence (mean = 51.55, SD = 52.39), macro-economic policy (mean = 595.99, SD = 641.42), Environment and Natural Resources (mean = 30.34, SD = 40.35), and law, order and civil rights (mean = 361.70, SD = 441.69).
104 By asset, factor instability is distributed as follows: social policy (mean = 3.25, SD = 1.89), foreign and defence (mean = 3.27, SD = 1.96), macro-economic policy (mean = 3.74, SD = 1.75), environment and natural resources (mean = 1.16, SD = 2.25), and law, order and civil rights (mean = 4.31, SD = 2.09).
105 By asset, manifesto investment weights are distributed as follows: social policy (mean = 15.40, SD = 3.69), foreign and defence (mean = 11.41, SD = 3.12), macro-economic policy (mean = 29.12, SD = 7.69), environment and natural resources (mean = 8.57, SD = 3.57), and law, order and civil rights (mean = 17.47, SD = 6.51).