2.1. Logic of the PLC theory. Contrary to most PBC models (Rogoff, 1990; Rogoff and Sibert, 1988; Shi and Svensson, 2006), which are based on a principal-agent relationship between voters and their representatives, the PLC theory presupposes an interaction between three sets of agents: 1) the government in office, called the “legislator”; 2) interest groups and 3) unorganized voters (Padovano and Petrarca, 2017). This setup is more similar to those of the “common agency model” (Bernheim and Whinston, 1986; Denzau and Munger, 1986; Grossmann and Helpman, 2002) and to the special interest group theory of legislation (McCormick and Tollison, 1981; Becker, 1983). Voters and interest groups represent the demand side of the political market, the legislator the supply side. The logic and the assumptions of the model are straightforward. Voters act individually, have not borne the costs of organizing themselves into an interest group and aim at obtaining general public goods through legislation approved by the legislator. Interest groups, on the other hand, have to bear coordination costs in order to obtain specific private goods for the benefit of their members, again by means of legislative measures approved by the legislator (Lagona and Padovano, 2008; Padovano and Petrarca, 2017). Voters are supposed to have incomplete information about legislation (Nannestad and Paldam, 1994; Lagona and Padovano, 2008), while interest groups, thanks to their organization, have complete information about it (Tollison, 1996; Moe, 1996). Finally, voters are the only ones who can guarantee re-election to the legislator with their votes; interest groups instead can exchange legislation with the legislator in return of information and economic resources (Rowley and Vachris, 1994; Padovano, 1995; Austen-Smith, 1996).
The legislator’s goal is maximizing the probability of being re-elected, thanks to the support he receives from voters and interest groups in return of his legislative production. This production is carried out through two different legislative instruments named, in the jargon of the theory, “laws” and “decrees”. In the logic of the theory, what distinguishes laws from decrees are the information costs they engender. Laws, given their process of approbation, which always involves an explicit vote of the parliament, are supposed to be a highly visible instrument; decrees, instead, are typically an executive act that does not necessarily require the approval of the parliament; hence they are less easily known.5 The model incorporates this stylized fact assuming that decrees are visible only to interest groups (and to legislators who promulgate them), who have complete information; the content of the laws instead is known to all agents of the model, including voters. To the extent that every piece of legislation redistributes property rights (Peltzman 1976; Becker, 1983), the legislator can choose between these two different tools to redistribute specific private goods to interest groups and general public goods to voters. The model shows that the vote maximizing strategy for the legislator is articulated in two stages. In the first part of the legislature, when voters are less active, the legislator redistributes specific private goods in favor of interest groups by means of decrees, to obtain information and economic resources from them; then, in the last months before the elections, the legislator redistributes general public goods to the voters to get their vote; it will do so by passing laws.6 The model therefore predicts two opposite cycles of laws and decrees. Specifically, one should observe a low production of laws in the first years of the legislature, followed by peaks in the last months before the elections, when voters are more alert7. By the same logic, one should witness a larger use of decrees in the first period of the legislature when special interest groups are relatively more active than voters, and a decrease in the production of decrees in the last months of the legislature, when the approaching elections shifts the legislator’s attention from lobbies to voters in order to secure re-election. In other words, the opposite cycles of laws and decrees reflect the different periods of the legislature when either special interest groups or voters are expected to be relatively more active, thus attracting the legislator’s attention.
2.2. Empirical contributions. The existence of a legislative political cycle in legislative production has already been demonstrated both in the legislatures of the Italian Parliament (Lagona and Padovano, 2008; Auteri et al., 2018), as well as in other legislative branches. In particular, evidence of PLC has been found in the European Parliament (Kovats, 2009), in that of the Czech Republic, regardless of the subject matter of the law (Brechler and Gersl, 2011), in the French National Assembly for both legislative and presidential elections (Padovano and Gavoille, 2017) and in that of South Korea, with increasing magnitude as the country became more democratic (Lagona and Padovano, 2021). Finally, Padovano and Sy (2023) exploit a new dataset that includes the legislative activity of 19 countries, mainly from 1975 to the 2010s, to analyze in a comparative setting how institutional features affect the size of the PLC. In line with theory, the magnitude of the cycle is interpreted as a measure of the inefficiency in the agency relationship between voters and elected officials. The panel estimates show that the cycles are larger in countries that adopt a PR electoral systems, in countries characterized by a parliamentary government and where by a higher-than-average degree of government decentralization.
While the proofs on the cycle of laws are now extensive and robust, the studies on the cycle of decrees are few and the evidence more scarce. Official data about the production of decrees are more difficult to obtain than in the case of formal laws; moreover, the institutional characteristics of decrees are more heterogenous than those of laws, which makes the study of decrees more complex. So far, Lagona et al. (2015) is the only research where the “full hypothesis” of an opposite cycle of laws and decrees has ever been tested, receiving positive empirical support. In that paper, however, the time series of different types of decrees are added together into a single variable, to be compared with the dynamics of the other highly aggregated legislative instrument examined, i.e., the formal laws.
More recently, official records on the promulgation of decrees in Italy have become available via two official sources, the Official Collection of the normative acts of the Italian Republic (Raccolta Ufficiale degli Atti Normativi della Repubblica Italiana), and the Official Gazette (Gazzetta Ufficiale della Repubblica Italiana), as well as in the databank Normattiva. These new datasets allow, for the first time, to perform a test of the PLC theory disaggregated by type of decree. In addition to that, they also include all the most important types of decrees featured in the Italian legislative system, classified according to the principles of the Italian legislative system, thus minimizing the discretion of the analyst in the selection and classification of the data.
2.3. Contributions of constitutional law. Italian scholars of constitutional law have usually examined each type of decree separately, analyzing the institutional details that (should) govern their approbation and implementation, according to the Italian Constitution and the other sources of law (Modugno, 2012; Bin and Petruzzella, 2019). While this type-specific approach is certainly useful to gather a proper understanding of the differences between the various types of decrees, several scholars have also pointed out that, over time, many democratic countries witnessed a significant expansion of legislators’ discretionary power to seamlessly adopt different legislative instruments regardless of the subject matter of legislation for reasons of political expediency (Campos, 1989; Della Sala and Kreppiel, 1998; Tiller and Spiller, 1999). This process has increased the substitutability between legislative instruments and blurred the institutional differences that the constitutions and other sources of law had originally established. As a side effect, this greater substitutability seems to have swollen the legislative output directly implemented by the executive branch at the expense of the legislation approved by the parliaments, in Italy and elsewhere (Carey and Shughart, 1998).
To the best of our knowledge (at least in the context of the Italian legislative system) these claims of the constitutional law literature have been substantiated mainly through anectotical evidence, with little resort to statistical inference. Hence the analytical approach provided by the PLC theory seems quite appropriate to the study of the production of decrees, since: a) contarry to the constitutional law literature, it is more general, as it fails to impose any a priori on the legislator’s choice of the legislative instrument, allowing for substitutability between them; b) at the same time, it permits to take into account the specific features of each type of decree; c) it provides testable hypotheses to be submitted to rigorous empirical analysis.