Indicators and Policy Formulation
Many of the expected functions of indicators fall outside of the scope of policy formulation as defined in Chapter 1: in particular composite and headline indicators are designed to influence phases preceding policy formulation, notably agenda-setting and problem identification. In policy formulation, indicators can be crucial in characterizing the current situation; already the choice of the indicators for describing the current policy situation both reflects and shapes perceptions of which elements in decision making situations are deemed important. By virtue of their perceived rigour and accuracy, indicators can assist problem conceptualization - for instance by informing the development of formal or informal models (Seaford 2013), and construction of scenarios. More generally, by quantifying and simplifying, indicators render problems more manageable. Indicators also help to shape perceptions of which policies are deemed viable and relevant, that is, the identification of policy options. For instance, the choice of the parameters to constitute an air pollution index, or the choices of the appropriate biodiversity or climate change indicators shape the range and viability of alternative policy responses. Indicators are frequently used in assessing and comparing potential policy options. Finally, while indicators are not designed to recommend and/or propose a specific policy design, in reality they are frequently used to justify a given (often a pre-existing) policy design.
Instrumental, Conceptual and Political Functions of Indicators
The types of intended use of indicators often tend to focus on what the knowledge-utilization literature terms instrumental use (for example, Weiss 1999), entailing in our case the use of indicators as direct input to specific decisions, in line with the linear rational-positivist model of policymaking, typically involving 'single-loop' learning concerning the consequences of specific actions or policy options (Argyris and Schon 1978). Expectations concerning performance management indicators typically fall within this category. Yet many of the intended functions mentioned above can be considered as conceptual, as indicators are expected to constitute a part of a broad information base for decisions, shape conceptual frameworks and mental models of actors, and ultimately generate 'enlightenment' (Weiss 1999). Hence, indicators should foster especially the more complex types of social learning in the spirit of Habermasian 'communicative rationality'. Finally, some of the expected functions are political, especially when indicators are expected to influence agenda-setting and problem definition, highlight neglected issues, or (de)stabilize and (de)legitimize prevailing frameworks of thought.
Often the political use of indicators is overlooked and portrayed in a negative light, as misuse, abuse, attempts to conceal, cheat, delay and manipulate (for example, Hezri 2006). Alternatively, the absence of proof that indicators have influenced policy is often taken as a proof of failure. 'Indicator advocates' hence often regret the fact that indicators are either ignored or used selectively, or 'misused' for strategic (in other words, illegitimate) purposes, without due regard for their inherent technical limitations. Numerous measures are then often suggested to minimize 'misuse', such as informing and educating the users or the introduction of 'statistical health warnings' (for example, Jackson 2011, p. 20; Pinter et al. 2005, p. 7; Hood 2007, p. 100; Grupp and Schubert 2010, p. 76). But the 'political' use of indicators can involve more than outright legitimization of decisions already made and 'symbolic' use. It can also entail necessary efforts to strengthen the legitimacy of democratic decision making, or advocacy for socially progressive objectives (Parris and Kates 2003). Regarding the former, Stirling (2006) qualifies this potentially constructive type of legitimization as 'weak' justification, targeted at legitimizing processes and/or institutions, whereas 'strong' justification would focus on justifying substantive policy outcomes (in other words, ex post legitimization of decisions that have been made on other grounds than those explicitly mentioned). Political use and functions of indicators can involve double- or triple-loop learning, including 'political learning', which concerns the 'political feasibility of a given idea or prospects for advancing a given problem through manoeuvring within and manipulation of policy processes' (Hezri 2006, p. 101).
As Table 4.1 (Behn 2003) suggests, the roles of indicators extend well beyond the direct policy formulation tasks, covering in particular crucial feedback functions of evaluation, control, learning, motivation and even various 'symbolic' functions. Such functions are, however, de facto rather than intended consequences of indicators. The following sections will review the empirical experience concerning the actually observed and potential unintended consequences of indicators.