The question of why compliance norms exist is complemented by the puzzle of why compliance behavior persists. While most scholars agree that states are capable of gener- ating compliance norms through the application of sanctions, a longstanding controversy concerns what happens once the coercive apparatus is removed or relaxed. According to the theory of willed or voluntary compliance that is dominant in behavioral political sci- ence, the effect of a coercive state is to habituate individuals into rational calculation of the costs and benefits of compliance, eroding the “good,” voluntary and non-rational bases of cooperation, such as shared identity or communal ties, which are seen as more durable in the long run. Once these incentives no longer exist, individuals rationally, and rapidly, revert to non-compliance. Putnam (1993) famously made such an argument with respect to the Norman states of southern Italy, arguing that the hierarchical authority relations of the feudal state undermined local reciprocal ties, and while this created a strong and powerful regime for some time, it ultimately resulted in a rapid decline into criminality and amoral
familism following Italian unification (Banfield 1958). Rose (1994), Rose et al. (1997) and Howard (2003) make a similar argument with respect to the communist regimes of Central and Eastern Europe, in that the pervasive monitoring of the authoritarian state led individuals to see this sanction as the only basis for compliance, resulting in widespread non-compliance once the regime had collapsed. Boix and Posner (1998) and Huysseune (2003) similarly argue that the state is at best a potential destroyer of social capital and trust, while Ullman-Margalit (2004: 75), for example, argues that states do not generate social trust, but simply render such trust unecessary, by socializing individuals into the expectation of sanctions against untrustworthy behaviour. This constitutes what Levi and colleagues (1998) term “cooperation without trust,” which, under the aegis of a coercive state, we may also simply call compliance. According to such a view, the public order of the despotic state is merely a chimera, an illusory order, that undermines the long-term basis for voluntary adhesion to the public good.
I believe that this view is greatly mistaken, and that compliance, once generated, tends to persist. While it may be true that authoritarian states undermine or supplant horizontal and civic ties (cf. Bernhard and Karakoc¸ 2007), the hypothesis that strong and centralized states undermine their own long-term ability to govern stands quite poorly as a general- ized hypothesis. The historical evidence provides many more examples of the persistence of compliance, than of high compliance societies which collapsed into state failure. As argued by Ertman (1997) the distinction between bureaucratic and non-bureaucratic state structures may be more important than the degree of political openness, as bureaucratic states, even when authoritarian, build infrastructural capacity. Therefore absolutist regimes in Europe often gave rise to bureaucratic states, as in the cases of Denmark or Sweden. Historical studies by Jensen (2014) and Bagenholm (2014) show that absolutist rulers in Denmark and Sweden were able to deliver a large reduction in corruption, that were durable following transition to a less authoritarian regime. Fukuyama (2014) has made a similar argument regarding the sequencing of state formation and democratization more generally: in countries such as Greece, which democratized prior to state centralization, clientistic patterns of rule persisted into the modern era, whereas in cases where absolutism first es- tablished a central state, such as Denmark, democratic institutions emerged without clien- telism or patronage. A similar argument was made by Huntington (1968), who reprised the
argument of Elton (1953, 1960) regarding the legacy of Tudor absolutism for norm com- pliance in England, and added its transferral to the founding settlers of the United States. Outside of Europe, Gerner (2012) shows that non-European countries with a longer history of state formation were significantly more likely to become stable autocracies, implying again a persistence rather than a reversal of norms of public compliance with authority.
This can be shown considering both longitudinal and cross-sectional data. Gurr (1981) for example shows that a large decline in homicide occurred under English state formation during the late medieval era, while Eisner (2004) replicates these results for a large sample of countries in Europe. Finally, cross-country evidence also supports such a hypothesis: there is a striking correlation between measures of historical state formation, such as the State Antiquity Index developed by Putterman (2014), and measures of crime such as cross- country homicide rates. Further to this, authors have challenged the explanations of Putnam (1993) or Rose (1994) regarding the breakdown in legal, fiscal, and bureaucratic compli- ance in southern Italy and eastern Europe. The few examples favored by Tocquevillian social capital theorists often have alternative explanations for subsequent state weakness. Ziblatt (2006) shows that regions of southern Italy had weak state capacity, as measured by numbers of bureaucrats, soldiers, and taxation, even in the period before Italian unification, with the implication that the low compliance of the Italian south may be a persistent legacy of state weakness24. Putnam’s narrative of regional variation in Italy stands the facts on
their head: in the period preceding Italian unification, the northern polities of Piedmont and Tuscany were the best examples of centralizing, bureaucratic, and absolutist regimes, while the southern Kingdom of Naples was a decentralized, feudal regime, whose very ab- sence of infrastructural power allowed the development of organizations such as the mafia.
The Formation of Compliance Norms
The persistence of compliance norms in spite of changes of institutions and incentives requires additional mechanisms to be added to the simple rational choice framework out- lined in the previous section. Thinking about compliance in terms of monitoring and en-
24It might be added that, while Putnam (1993) had cited Machiavelli in support of northern Italian republi-
can government, Machiavelli’s own contrast was not with Naples, but rather with France, which no-one today would cite as an example of a weak state.
forcement fails to account for why such behaviors would persist over time, or explain how patterns of compliance can vary within a single formal institutional environment. There- fore, to understand variation in compliance, it is necessary to explain the formation of com- pliance “cultures.” In the first instance, exposure to the state’s regulatory control does not simply generate a stream of successive compliance decisions, but also leads to the forma- tion of compliance “norms”: default modes of action, or behavioural heuristics, in response to an assessment (conscious or otherwise) of the relative costs and benefits under some re- curring situation. If a person arrives in a new city in the United States, for example, they may initially avoid jaywalking because they do not know whether such laws are enforced, and continue to do so for a long while (leading to a large cumulative time loss) because updating their prior assumption is too potentially costly at any given moment, relative to the cost of waiting a few extra seconds for a cross signal. Such a norm may originate as an assessment of relative costs and benefits, but persist long after the balance of incen- tives has altered, as long as information remains imperfect and there remain search costs to updating this knowledge25. Behavioral economics has also revealed another feature of human psychology which makes such norms stubborn, which is that of time inconsistent preferences - in particular, weighting present returns over future returns, or preferring an immediate, small loss or gain, for example, over a larger upfront cost for a large net future gain (Loewenstein et al. 2003). Because updating priors involves an immediate cost in the present while payoffs to behavioral change lie cumulatively over an indefinite future, the sub-optimal equilibrium for many individuals is simply to continue complying with the law26. Regarding compliance with public authority in general, assessments are made
throughout adolescence and adult life whether to comply, based on limited or imperfect in- formation regarding enforcement. As long as the demands of public authority remain low,
25The classic demonstration in psychology is the phenomenon of ‘learned helplessness’, whereby animal
subjects ‘give up’ on engaging in actions that would avert a negative stimuli, after a series of trials in which efforts to do so are frustrated (Seligman 1967, 1975). In this case a norm develops (not to make further efforts) and is sustained based on the experience of a few failed attempts, when continued trial and assessment would have allowed for a better assessment of the situation.
26The irrationality of human time preference can be used to explain the persistence of ‘bad habits’ that
harm individuals in the long run, yet, as in this instance, also explain the persistence of ‘good habits’ that benefit society. Most individuals are ‘addicted’ to the law - continually opting for the smaller, short-term pain of state compliance, rather than incur the larger upfront cost of discovering individual freedom of action.
and information regarding methods and strategies of enforcement remain uncertain, norms of compliance are likely to persist.
However, norms alone are unable to explain persistent cultures of compliance because as soon as a state begins to make increased demands, for example by introducing compul- sory service (e.g. corv´ee labor or conscription), or individuals become more familiar with the enforcement strategies of public authorities, a threshold may be crossed at which ratio- nal choices regarding the decision to comply become more frequent, and instinctive norms of compliance begin to erode. Additional theories are needed to explain why individuals continue to comply.
One such theory is that once compliance has been generated, it constitutes an equilib- rium, because the ability of the state to enforce compliance depends upon collective action dynamics among the population as a whole. The expected cost of tax evasion is quite low when every other citizen is avoiding to pay their taxes - the state, evidently, cannot pros- ecute everyone at once - but the expected cost becomes rather high when one is the first person to break rank. Similarly, a single individual throwing objects at the police is at high risk of arrest: an entire crowd is not. In this way, there are collective action dynamics to public compliance, as the social trust paradigm has proposed, but regarding enforcement rather than cooperation. If a citizen believes that other citizens are not going to pay a new tax, they are indeed less likely to do so themselves. But this is not because they fear being a ‘dope’, as the social trust argument asserts (by the logic of rational choice, they will always be a ‘dope’ if they had the option of being a free rider, regardless of whether others comply or not). Rather, it is because once others defect from payment, they are also more likely to get away with it, whereas were others to keep paying, their chances of discovery, shame, and sanction are significantly higher.
Second, and by extension, this equilibrium dynamic is extended further by the “multi- level” nature of the compliance game (Putnam 1988). An individual’s incentive to comply with the law depends on the investigative and enforcement powers of the state, but the state is engaged in a compliance dilemma with its enforcers: in their capacity as its agents, they may also choose not to comply, and in this way alter the incentives for those who are under their jurisdiction. The decision of the police to enforce depends both on the sanctions placed upon them by their authorities, and the level of sanction or resistance from society.
If social actors are resistant, this increases the incentive of the police to forebear rather than enforce. This is another of the attributes of the non-compliance equilibrium: the degree of non-compliance determines not only the likelihood of individual sanction, but also the absolute level of enforcement. Intermediate authorities are likely to give up on attempting to enforce laws that are pervasively violated, and realise that social resistance may be incurred from attempts to shift to a compliance equilibrium. But once social actors have been pacified, the confidence of authorities to enforce laws will increase.
Third, there are individuals in every population prepared to engage in what Herrmann et al. (2011) term “pro-social punishment” - i.e., the willingness to undertake a short- term personal cost in order to sanction those who refuse to comply with the rules of the political community - and such individuals are likely to be more widespread under com- pliance equilibrium. The increased prevalence of pro-social monitoring and enforcement in a functioning state changes the incentive for individual compliance. In a compliance equilibrium, there are likely to be more individuals who benefit from the state’s provision of public goods, and whose moral and material investment in the status quo will lead them to act as pro-social enforcers. This may be as simple as engaging in minor, low-cost actions such as reporting to the authorities a neighbour who is known to avoid paying her munici- pal tax or ‘shaming’ associates into obeying local regulations, or as extensive as engaging in direct retaliatory action, that the existence of third-party enforcement by the state typi- cally vitiates the need for such behaviour. In any circumstance, the presence of pro-social enforcement greatly increases the expected cost of non-compliance, and the provision of public goods and services by the state increases the numbers of those prepared to act as so- cial enforcers. As Herrmann et al. (2011) report in their cross-national data on compliance games - there is a high contribution to public welfare and pro-social punishment in coun- tries such as Switzerland and Australia, and low contributions in anti-social punishment is more common, such as Saudi Arabia, Oman and Greece.
Finally, it would not be possible to account for the exceptional persistence of compli- ance norms without acknowledging the role of belief-systems, as well as the role of states in sustaining and propagating such beliefs, as long emphasized by the Marxist literature on ideology (e.g. Therborn 1979, Althusser 1975). It is easy to comprehend that normative belief-systems, or beliefs regarding ethical or political values, affect willingness to com-
ply in a very direct way, and collective memories, identities, loyalties, and ethical views concerning compliance are fostered by political rulers and their institutions, and these are the arguments made by “voluntaristic” theories of compliance. However positive belief- systems, or beliefs about the world, are also a mechanism for affecting compliance, by altering individual assessments of the likelihood of detection (p(d)) and the severity of the sanction (S), and have been less an object of study for political science. The paradigmatic example is that provided by religious belief systems, which achieve compliance with their rules entirely through belief in the monitoring and eventual sanction of a deity – without apparent need for actual divine intervention – and for this reason many early states have solved the problem of order on a theological basis. Having officers of state motivated by belief in otherworldly rewards alleviates the sovereign, at least partially, from having to provide such rewards himself, or impose penalties through constant surveillance and dis- cipline of his auxillaries27. Yet even where state power is not supported by a complex
metaphysical system, ideology has an important role in maintaining compliance. This can be evidenced from surveys which show individuals to systematically overestimate their likelihood of being caught when engaged in a range of non-legal activities, as well as the typical sanction they would receive if caught, suggesting that belief in state power, rather than the reality of such power, remains a basis for adherence to the law (Scholz and Pinney 1995, Alm et al. 1992).
Another example of how belief-systems raise compliance is the example given by ter- ror under totalitarian regimes, which does so not through enforcement alone, but also by changing the perception of the degree of enforcement, for example through show-trials or
27Regarding the effect of religious ideology on compliance, Shariff and Rhemtulla (2012) find that, while
there is no correlation between religiousity and crime rate, as measured by the level of homicide, there is a strong association between lawfulness and belief in divine punishment (belief in “hell”). The authors con- clude that belief in sanctioning explains compliant behaviors, an intuition borne out in an earlier experimental in which Shariff and Norenzayan (2011) found that students who believed in a more ‘forgiving’ God were more likely to cheat on a test. Finally Barro and McCleary (2004) find that while there is no relationship between religiosity and economic growth, nonetheless cultures that believe in hell have significantly higher growth rates, which they attribute to the effect of norm compliance on social cohesion and the rule of law. With regard to early political order, early states have often united political and theological power, either be- cause in Samuel Huntington’s (1996) phrase “Caesar is God” (as in the case of ancient Egypt or China) or because “God is Caesar” (as in the early Islamic caliphates). Belief in divine sanction is a classic means of solving the Platonic dilemma of who ‘guards the guardians’.
propaganda, which create an impression of the state’s enhanced reach. Thus even if the surveillance capacities of the state are not objectively greater, it is enough for individuals to believe that public authorities have the capacity to detect or expose their wrongdoing, in order for individuals to rationally refrain from such acts. In a fascinating discussion of compliance in Soviet Russia, for example, Herreros (2004, 2006) argues that the effort of the Soviet authorities to engage in “show trials” was done with a conscious goal – to shift expectations among the rest of the population, as the “plots” of which people were accused and confessed, though fabricated, were perceived as real conspiracies that were detected by the secret police. Thus an impression was generated of the NKVD’s omniscience, such that any plot against the regime would invite comparable detection and punishment. Through social history, we know that such regimes contained substantially greater norm defiance than either external or internal actors understood at the time (Peukert 1988, Shulman 2012). But if the total control of the Soviet state was largely fictitious, still it became a little more true by virtue of people’s belief in it, and this is especially so given that a majority of the public, behaving according to the precautionary principle, was inclined to avoid any ac- tion that might lead to trouble. Regimes of terror are very effective in raising compliance, and reliance on perceived, rather than actual, capacity only becomes problematic once this belief evaporates28.
Because of the ability of states to alter not only individual incentives, but also norms and beliefs, sociologists such as Norbert Elias (1982) or Michel Foucault (1976) have long studied the ways in which the state, through disciplinary institutions such as schools, pris- ons, and the bureaucracy, is capable of instilling in individuals a set of behavioral heuristics that guide individuals into compliant behaviors, often against an objective calculation of ra- tional self-interest. As mechanisms of socialization become more extensive and elaborate, the corpus of rules expands; in this way Foucault (1961, 1963, 1975) argued that the rise of the early modern state was accompanied by profound discursive shifts in medicine, demog-