Introduction

While secret voting in elections is currently judged essential for protecting voters from intimidation or coercion, secret voting in parliaments is controversial.Footnote ¹ Parliaments in the nineteenth century used secret voting to protect the freedom and autonomy of representatives from interference, or control, by a monarch or royalist government. This was consistent with both the trustee model of representation championed by Edmund Burke and J.S. Mill and the organization of parliaments before the extension of the franchise. Contemporary parliaments have in many cases abolished secret voting, or now only use it for nominations or appointments. This limited use of secret voting is consistent with conceptualizing representation as accountability and organizing modern democratic parliaments around the central role played by political parties. Nothwistanding this, arguments in favor of secret voting survive when the autonomy of deputies is advocated against strict party discipline.

Italy offers an interesting example of the persistence of secret voting from the 1848 constitutional monarchy to the post–World War II republican democracy. The Constitution of the Kingdom of Italy (Statuto Albertino) promulgated by King Charles Albert of Savoy in 1848 gave formal recognition to secret voting, making it compulsory for the final vote on bills.Footnote ² Voting via secret ballot at any stage of the legislative process could be requested by ten deputies or eight senators. This number was increased to twenty in the (lower) Chamber and ten in the Senate in 1868. It was decided that secret voting would prevail over open voting if competing requests were put forth (Rules of Procedure, Article 97). Thereafter, both secret voting for the final vote on bills and the norm of giving precedence to secret voting have been applied in the Italian Chamber,Footnote ³ with the exception of the Fascist period (1922–43).Footnote ⁴

This situation lasted until secret voting for the final vote on bills was abolished in 1988 – 140 years after its introduction into the Italian parliament. The survival of secret voting in the Italian parliament's Rules of Procedure and its retention in specific matters to the present day seems at odds with the principles of popular sovereignty and accountability of representatives that inspired Italy's Republican Constitution of 1948.

Different explanations for this state of affairs may be proposed. On the one hand, the persistence of secret voting could be interpreted as an example of institutional “path dependence.” On the other hand, it may also be argued that secret voting has been instrumental to the (changing) strategic purposes of key political actors, and that secret voting for bills was abolished when it no longer served a useful purpose in the eyes of some party leaders. This conjecture is consistent with an extensive rational choice literature that highlights the effects that institutions have on outcomes, and the “manipulation” of rules in order to secure desired outcomes (Shepsle Reference Shepsle, Rhodes, Binder and Rockman2006). This chapter argues that secret voting was in practice used primarily as a strategic tool by governing parties’ factions and opposition parties to (1) shape the content of legislation and (2) determine the composition of governments.Footnote ⁵

The argument presented in the following pages is structured as follows. In the first section the debate surrounding the use of secret voting in the 1948 constitution-making process is examined. In the second section there is a brief description of the prevailing patterns of legislative-executive relations in the 1948–92 period. This is followed by an account of the main changes in the parliamentary Rules of Procedure from 1948 to 1988. In the fifth section, the strategic use of secret voting in the Italian parliament and its effects on government stability are examined through a number of case studies. In the penultimate section, an account of the parliamentary process that led to the reform of secret ballot in 1988 is presented. I close the chapter with some concluding remarks.

Secret Voting in the Constitution-Making Process

Secret voting in the Italian constitution-making process can be examined from two perspectives. The first one implies looking at the use of secret voting in the workings of the Constituent Assembly, as the Assembly adopted the same Rules of Procedure that had been in force under the constitutional monarchy until 1922. According to these rules, a secret vote could be requested by twenty deputies and the secret vote had precedence over open voting. The second perspective refers to the debate surrounding the decision about whether or not to adopt secret voting as a provision within the future constitution, and the actual decision taken on this matter. Both aspects are relevant for understanding the general principles used to justify secret voting and the strategic considerations by political actors in the constitution-making process (Elster Reference Elster1993).

The Constituent Assembly of the newborn Italian Republic was elected by universal suffrage on June 2, 1946.Footnote ⁶ Its 556 members were elected by proportional representation in 32 multimember districts. A subcommittee of 75 members was nominated in order to draft the Constitution for the Italian Republic. The committee's proposals were then debated and voted through majority rule by the plenary Assembly. Any member of the Assembly could submit amendments. The Assembly held meetings until January 31, 1948, holding 375 public sessions, 170 of which were devoted to constitutional debates.Footnote ⁷

The partisan allocation of seats in the Constituent Assembly shown in Table 5.1 shows that the largest party in the assembly was the Christian Democratic Party (DC), which had 209 seats. The second-largest party was the Communist Party (PCI) with 104 seats. The socialists were divided into two parties with 65 (PSI) and 49 (PSLI) seats each.Footnote ⁸ Other parties were the centrist Liberal (22 seats) and Republican (25 seats) parties. The right-wing ideology was represented by the Fronte Liberale e Democratico dell'Uomo Qualunque (20 seats) and Unione Nazionale (13 seats).

The Constituent Assembly made extensive use of secret voting. Strategic motivations cloaked in arguments of principle are evident from the outset. In the session of April 23, 1947, the first request for a ballot using secret voting was advanced by twenty deputies from liberal, republican, and socialist parties.

The issue at stake was an amendment submitted by a socialist deputy, aiming at abolishing the word “indissolubility” (of marriage) in Article 23 of the future Constitution. It should be noted that deleting the expression “indissolubility of marriage” from the draft constitution was interpreted by many observers as being anti-Catholic and indicating a pro-divorce stance.

The indissolubility of marriage vote followed a debate about procedural decisions, which highlights the tactical use of secret voting. The Christian Democrats were in favor of open voting (and against the substance of the amendment). A Christian Democratic speaker explicitly accused deputies from small parties of requesting a secret vote with the purpose of not revealing their preferences to a predominantly Catholic electorate. The Communists, who were openly in favor of the substance of the amendment, declared that the request for a secret vote had to be accepted because small parties needed institutional protection from being dominated by larger parties. In the final ballot, 194 out of 385 deputies attending the session voted in favor of the amendment and 191 voted against (majority threshold 193). As a result, the word “indissolubility” (of marriage) was removed from the draft constitution.

From then on secret voting was used increasingly in the constitution-making process: forty-three votes on constitutional amendments were taken using a secret ballot and twenty-three through roll-call voting (the first of which was requested on March 22, 1947). A “technical” reason has been suggested for this development (Sparisci Reference Sparisci1987). A secret vote required less time than an open vote because the latter allowed “voting declarations.” Consequently, a request for a secret ballot could be advanced in order to prevent the practice of filibustering through endless deputies’ speeches.

An examination of these sixty-six votes shows that strategic reasons may have played a role, as roll-call voting on constitutional amendments tended to be requested more often by DC deputies, whereas secret voting tended to be requested by opposition parties. This conjecture is reinforced by the fact that the Constituent Assembly attempted to reform the Parliamentary Rules of Procedure, and secret voting in particular. On May 5, 1947, a proposal aiming at increasing the number of deputies required to impose a secret vote from fifteen to thirty was submitted by DC representatives. After a debate during which deputies from minor parties declared their fierce opposition, a secret vote was requested on the decision of whether or not to proceed with reforming the Rules of Procedure. Out of 380 voters, 194 voted in favor of the status quo and 186 voted against. Consequently, the Assembly retained the prevailing rules. Another reform of the Rules of Procedure aiming at giving precedence to roll-call vote over secret ballot in case of concurring requests was proposed by DC deputies on September 26, 1947, but was never debated.

The debate about whether or not secret voting should have become a provision embodied in the future Constitution clearly highlights parties’ positions on the issue. On October 14, 1947, the DC leader Aldo Moro submitted an amendment to article 69 in the draft of the constitution (Article 72 in the definitive version) aiming at removing the norm of secret voting from the constitutional text. In a passionate speech, Moro argued that decisions about regulation of the legislative process should have been left to the ordinary rules of procedure and not embodied into the Constitution. The following day the Assembly approved a new text that removed secret voting from the future Constitution. The Communist Party (PCI) was in favor of retaining the original formulation. To this end, an amendment was submitted by two PCI deputies, but it was rejected through secret voting on October 15, 1947 (135 for, 160 against).

This debate was framed in terms of general principles such as liberty and autonomy of deputies versus their accountability to the electorate. However, it is worth noting that in the transitional Consulta Nazionale installed immediately after the end of World War II (September 25, 1945 to June 2, 1946), the DC and the PCI parties’ positions on the matter were more ambiguous as their future electoral prospects were uncertain. The popular vote for electing the Constituent Assembly made clear that the DC party could now count on a high level of popular support. It was very likely that such support would have been confirmed in the general election scheduled immediately after the approval of the new Constitution. This expectation determined a clear stance in favor of open voting, as this procedure would have facilitated party discipline once in government. On the other hand, the PCI understood that secret voting could become a tool in the hands of the opposition as a means of exploiting divisions in any future governing majority (Curreri Reference Curreri and Labriola1998).

Executive-Legislative Relations in Italy, 1948–1992

Executive-legislative relations are of key importance in understanding the strategic use of secret voting. Consequently, it is important to focus on the main actors in the Italian political system and on how the relationship between ruling coalitions and opposition parties evolved over time.

The first elections in the history of the Italian Republic (April 18, 1948) established the predominance of the DC and shaped in a lasting way the party system. In the Chamber the DC gained 305 seats, mostly at expense of minor centrist parties, while the Popular Front (Communists and Socialists) gained 183 seats. From 1948 to the early 1990s, the configuration of the party system remained relatively stable. Seven parties (PCI, PSI, PSDI, PRI, DC, PLI, and MSI) enjoyed a relatively constant level of electoral support throughout this period. The PCI was the most left-wing party on the left-right policy dimension and the MSI the most right-wing (Laver and Schofield Reference Laver and Schofield1990). Both the PCI and the MSI were confined to permanent opposition given their “anti-system” stance. The DC was the largest party placed in the center of the political spectrum. From the beginning the DC was compelled to form coalition governments.Footnote ⁹ The period from 1948 to the early 1990s may be divided into four distinct phases (Verzichelli and Cotta Reference Verzichelli, Cotta, Müller and Strom2000):

• The centrist coalition phase (1948–60), when the DC formed governments with minor centrist parties.

• The center-left coalition phase (1960–75), when governments included the Socialist Party (PSI).

• The National Solidarity phase (1975–79), when the DC formed minority governments with the abstention or the support of the Communist Party.

• The pentapartito (five parties) coalition phase (1980–92), when the DC governed with PSI, PSDI, PRI, and PLI.

These four phases differ in terms of the prevailing pattern of executive-legislative relationships. Up to the 1980s, the political scenario was dominated by the two largest parties – the DC and the PCI. The PCI was the largest communist party in any Western country. However, it was excluded from government office due to its ties with the Soviet Union. Political elites developed a pattern of cooperation captured in such terms as “consociationalism” (consociativismo) or consensual democracy where ideological polarization, not cultural segmentation, was the main source of division (Lijphart Reference Lijphart1968; Bogaards Reference Bogaards2005).

This cooperation is well documented in the literature that shows that three out of every four laws approved by the Italian parliament passed with the support of opposition parties including the Communist Party (Di Palma Reference Di Palma1977; Cotta Reference Cotta, Cotta and Isernia1996). Such consensus was equally high for laws approved through the so-called decentralized procedure, that is, by parliamentary committees, which was the most common procedure up to 1980s, or through the ordinary procedure, that is, by securing a majority on the floor.Footnote ¹⁰ The literature stresses the micro-sectional character of Italian lawmaking (Zucchini Reference Zucchini1997; Giuliani Reference Giuliani1997).

In the national elections of 1976, the PCI came close to the DC in terms of electoral support, gaining 34.37 percent of the vote while the DC gained 38.71 percent. Other parties entered the parliament for the first time (the Radical Party, an extreme left splinter, and the Greens). The collaboration between DC and PCI – defined as a “historic compromise” by political leaders and intellectuals – culminated in the PCI abstaining in the vote installing a DC minority government in 1976, and the PCI vote of confidence leading to the next DC cabinet formed in 1978. These events occurred when Italian democracy was grappling with the threat posed by political terrorism.Footnote ¹¹

In the 1980s, the pattern of cooperation between the DC and PCI eroded due to the increasing role played by the reformed Socialist Party under the leadership of Bettino Craxi (from 1976 onward). Even though the PSI never became a serious electoral challenger either to the PCI or the DC, its central position on the left-right dimension allowed it to play a pivotal role in the making and breaking of governments. Throughout the 1980s, both the DC and the PCI suffered electoral losses, as in the 1983 elections in which the DC gained 33 percent of the popular vote and the PCI 29 percent of the votes cast. Even though the DC was still the largest party, its bargaining power in government formation was undermined as the DC had to concede the office of prime minister to the Republican Party in 1981 and eventually to the Socialist Party in 1983.

The literature about coalition formation in Italy emphasizes a typical pattern of short-lived governments (the mean cabinet duration was less than a year) coexisting with policy stability, as the coalitions that formed always included the DC as the largest and central party (Giannetti and Sened 2001). The heterogeneous nature of coalitions and especially internal divisions within parties are the key to understanding cabinet instability. The DC party had an institutionalized factional structure, but internal divisions were significant in other parties as well.Footnote ¹² Allocation of executive power among the major factions of each party within coalitions explains the recurrent pattern of government termination, where intraparty factions renegotiated their participation in the following cabinets. This was especially apparent during the five-party (pentapartito) coalition phase. In this period the PCI played a stronger oppositional role as it often succeeded in exploiting the divisions of the governing majority.

To sum up, Italian politics in the 1948–92 period was marked by two different patterns of executive and parliament relationships: a consensual phase up to 1979, and a more adversarial phase that lasted until 1993 when the approval of the electoral reform and the unfolding of corruption scandals led to ending the so-called First Republic (Mershon and Pasquino Reference Mershon and Pasquino1995). As we shall see in the next section, changes in the parliamentary rules of procedure reflect these evolving patterns.

Institutional Background: Parliamentary Procedures 1948–1988

Article 64 of the Italian Constitution of 1948 establishes that “Each Chamber will adopt its own Rules of Procedure through majority voting.” The Chamber of Deputies approved new Rules of Procedure in November 1949 but made only minor changes to the rules that had been in place up to 1922.Footnote ¹³ Article 93 (formerly article 97) reaffirmed (1) the binding use of secret voting for the final vote on bills, (2) the number of twenty deputies for requesting it at any stage of the legislative process, and (3) the precedence of the secret vote over the open roll call procedure.

In the previous section, we described how the legislative-executive relationship evolved from a consensual pattern of decision making that characterized the former phase of the Italian republic especially throughout the 1970s toward a more adversarial pattern that was typical of the 1980s. The Chamber undertook a major reform of its internal rules in 1971. The Rules of Procedure adopted in 1971 formalized the consensual nature of Italian democracy, giving the Parliament a central role vis-à-vis the executive and setting the institutional background for a further evolution of the relations among the largest governing and opposition parties (Labriola Reference Labriola and Violante2001; Lippolis Reference Lippolis and Violante2001).

The main changes may be summarized as follows. Parliamentary parties rather than individual legislators became the main actors in legislative politics. At least twenty deputies were required to form a parliamentary party grouping.Footnote ¹⁴ Parliamentary groups and their chairs (Capigruppo) were endowed with substantial agenda-setting power over the organization of parliamentary activity. What is most important here is that the timing of parliamentary activity had to be established by the Capigruppo using a unanimity rule.Footnote ¹⁵ The unanimity rule implied that there had to be an agreement among all the political groups represented in the parliament for the legislature to work effectively. This is why such a provision has been considered by many commentators as a key symbol of the consociational phase described in the previous section.

Rules constraining the terms of parliamentary debate (articles 39 and 83) and the submission of amendments on the floor were established (article 85), but any leader of a party group could ask for a departure or dispensation from these rules. As a consequence of such wide discretion, only an agreement among the governing and opposition parties could guarantee the enactment of bills. Moreover, the Rules of Procedure adopted in 1971 established that a request for a secret vote on a piece of legislation could be submitted by any chair of a parliamentary party grouping. This rule allowed minor party groups to request secret voting for strategic purposes.

One of the most important provisions of the Rules of Procedure adopted in 1971 was article 116, which regulated the use of the confidence vote. The confidence vote is typically used by governments to stabilize fragile ruling coalitions or speed up the legislative process by attaching the fate of a particular policy to a vote on government survival (Huber Reference Huber1996). According to article 94 of the Constitution, the investiture vote was to be taken using an open vote. However, the Constitution said nothing about the request of a confidence vote by the government. This is because the Rules of Procedure enacted before 1971 did not formally regulate the use of the confidence vote; between 1948 and 1971, the use of this procedure was based on informal rules.Footnote ¹⁶

Article 116 of the 1971 Rules of Procedure established that the confidence vote could not be requested on matters for which the rules prescribed the use of secret voting, such as the final vote on bills. More importantly, article 116 introduced a double vote (open and secret) on bills composed of just one article on which the government had requested a confidence vote. Single-article bills were typically decree laws or legislation that ratified international treatises.Footnote ¹⁷ The strategic importance of this provision can be fully understood by looking at one of the devices that the Italian executive has at its disposal in order to enact its own bills – the decree law procedure.

A decree law becomes operational immediately and remains in effect for sixty days without any parliamentary approval. If after this period the parliament has not “converted” the decree into a regular law, then the status quo prevails. The executive could reissue any number of decree laws that failed to get converted.Footnote ¹⁸ Throughout the 1970s, the use of decree laws increased well beyond the requirements of “extraordinary cases of necessity and urgency” established by the Constitution. Executive decrees became a common device that shaped strategic bargaining between the executive and the legislature (Della Sala and Kreppel Reference Della Sala, Kreppel, Carey and Shugart1998). In short, the main consequence of article 116 was that if the government asked for a confidence vote on a bill composed of a single article, which would have converted the decree law into a law, then immediately following this vote the same bill had to be approved by secret voting (in this order). This provision gave ample room for strategic maneuvering, as will be seen in the next section.

The end of the consensual phase in 1979 paved the way for a substantial revision of the Rules of Procedure during the 1980s. This process of revision occurred in several steps and involved curbing the terms of parliamentary debate and reforming the amendment procedure by allowing the government to group articles and amendments selectively. Among the most important provisions was the abolition of the unanimity rule for setting the agenda of parliamentary activity.

If attention is focused on changes related to the use of secret ballot, an important revision occurred in 1983, whereby the number of deputies required for calling a secret vote was increased to thirty. Moreover, it was established that the request for secret voting could be submitted only by chairs of parliamentary party groupings that, independently or jointly, amounted to thirty members. The use of secret ballot for final vote on bills was finally abolished in 1988. Before analyzing the parliamentary process that led to the reform of secret voting, in the next section we focus on the impact of secret voting on the legislative process and government termination.

Strategic Use of Secret Voting, 1979–1988

In the previous sections it was argued that the main actors in the political game devised parliamentary rules of procedure that were largely instrumental to their interests. Here it is important to stress the point that the secret ballot was seldom used when the relationships among government and opposition parties followed a pattern of cooperation in the legislative process. This is because the threat of requesting a secret vote ran the risk of revealing areas of dissent in the governing majority, and this fact induced government and opposition parties to reach prior agreement on the content of legislation. When the relationships among the governing and opposition parties evolved toward a more adversarial pattern, the use of secret ballot increased dramatically, becoming a constant threat to government survival.

Figure 5.1 shows the sharp increase in the use of secret voting during legislatures VIII (1979–83), IX (1983–87), and X (1987–92). An examination of this data reveals that the use of secret ballot increased after the end of the National Solidarity phase (1971–79) and the beginning of the pentapartito coalition governments (1981–92). To show how secret voting was an important strategic tool that allowed the opposition to exploit the divisions within the governing coalitions, the focus in this section is on government termination.

Figure 5.1. Comparison of two alternative measures of secret voting in a legislature where differences in duration taken into account.

Note: The duration of legislative terms and the number of secret votes have been normalized in order to remove differences that may result from variations in the absolute length of legislative sessions. This normalized estimate attempts to control for variation in absolute values where longer legislative terms would, ceteris paribus, be expected to have more secret votes. An alternative method is to present the average daily number of secret votes taken in a specific legislature term – an estimate that also controls for absolute differences in legislative terms. Source: Author's elaboration from data available at the Italian Chamber of Deputies.

To set the context, it is important to outline the composition and duration of governments immediately prior to the reform of secret voting and its immediate aftermath – that is, between May 1979 and June 1992. Table 5.2 shows the composition and duration of governments during the VIII, IX, and X legislatures. In what follows we focus on four cases of government termination, including (1) the Cossiga government in 1980, (2) the Spadolini government in 1981, (3) the Craxi government in 1986, and (4) the Goria government in 1988.

Cases (1) and (3) followed the same pattern: a decree law on a financial issue was introduced by the government requesting a confidence vote on it; this resulted in two votes in close succession, a confidence vote held by an open ballot and a secret vote on the same decree law; the government won the first (open) vote and then lost the second (secret) vote where between one in ten and one in five governing party MPs defied the party whip. This defeat, despite prior success in a confidence vote, led to the resignation of the prime minister and the immediate collapse of the government. These two episodes are striking because the first and the second votes occurred in sequence. Cases (2) and (4) also stem from being defeated through secret ballot in one or a series of votes, which may or may not have been preceded by a confidence vote.

The Reform of Secret Voting, 1988

The reform of secret voting was the outcome of an explicit attempt by the largest factions within the governing parties to (a) strengthen the role of the government vis-à-vis the parliament, (b) break the consensual pattern of legislative bargaining, and (c) redefine the role of the parliamentary opposition. In their parliamentary speeches asking for a confidence vote, Prime Ministers Spadolini (PRI) in 1982, Craxi (PSI) in 1983, and De Mita (DC) in 1988 explicitly declared their commitment to institutional reforms. It should be noted that these three prime ministers were also leaders of their respective parties.

The procedure for changing parliamentary rules differs from the procedure used for ordinary laws, as reform proposals for the former have to be examined by a special parliamentary committee (Giunta per il Regolamento) composed of representatives of all the major parties. The amendments to original proposals can be grouped by the committee into “super amendments” (principi emendativi) and then submitted to a vote on the floor (requiring simple majority support).Footnote ²⁵ Then the Committee revises the text accordingly and submits it to the floor for a final vote. The final approval requires an absolute majority, or 316 votes.

The parliament started debating the text drafted by the special committee on September 27, 1988. The discussion occupied a total of eleven sessions. The floor addressed three preliminary questions. First, should reform of secret voting be undertaken using open or secret ballots?Footnote ²⁶ Deputies of the MSI, Independent Left, and extreme left parties requested voting by secret ballot. The Chair of the Chamber accepted this request on the basis of the Parliamentary Rules of Procedure, stipulating (as noted earlier) that secret voting had precedence over open ballots for the final vote on bills.

The second key question centered on whether abolishing secret voting was a violation of the Italian Constitution. Three separate motions were tabled addressing this issue. The third question asked: Could proposals for reforming secret voting be postponed to some future date? Six motions were proposed by deputies from the PCI, the Radical Party, and the MSI who wanted to delay discussions aimed at reforming secret voting. In essence, the second and third questions posed by members of the opposition parties were attempts to halt or delay reform of the secret voting procedure.

Plenary votes on the second (constitutional) and third (timing) questions were subject to secret ballots. On the constitutional question, a majority of legislators (258) supported the view that reform of secret voting did not imply any violation of constitutional principles, while 231 argued this reform was in conflict with articles of the Italian Constitution such as Art. 67 (“Each member of the Parliament represents the Nation and carries out its duties without a binding mandate”). On the question of timing, a majority of deputies (267) voted in favor of dealing with reform of secret voting immediately and 247 supported doing this task at some future unspecified date. The narrow divisions on these secret roll-call votes (41 vs. 37 percent and 43 vs. 39 percent, with, respectively, 22 percent and 18 percent abstention rates) demonstrate how controversial and uncertain the process of reform was in late 1988.

Initially, forty amendments were tabled within the special committee charged with making proposals for reforming the secret vote. To reduce the burden of having many plenary votes, the committee aggregated these forty amendments on the basis of their substantive content into a smaller group of six “super” amendments. Then these six “super” amendments were voted on by the floor in eighteen separate votes, as shown in Table 5.4 (parliamentary session of October 7, 1988). The eighteen secret votes dealt with the use of open or secret vote on financial laws, constitutional laws, electoral laws, Parliamentary Rules of Procedure, and parliamentary committees charged with investigating powers. These votes highlight a tight parliamentary battle, in which some secret votes were won by a narrow margin, as shown in Table 5.4.

The evidence presented in this table shows the uncertainty of securing success in reforming the secret voting procedure sought by Italian party leaders. Opposition to the reform came from a number of sources, and different tactics were employed at different stages of the reform process. The second amendment, which proposed introducing an open ballot for financial laws, was passed by a wide margin. In contrast, other amendments were adopted with much narrower majorities: the third amendment allowing open ballot voting within parliamentary committees; the fifth amendment concerning the abolition of the double (final) vote for bills; and the sixth amendment restricting the use of secret ballot. However, the closest vote evident in Table 5.4 was on amendment 1-d, which was defeated by just three votes. Had amendment 1-d been approved, the reform of the secret vote procedure would have been fundamentally weakened, because unanimous agreement among parliamentary party leaders would have been able to secure a secret ballot on any specific bill they desired.

The special committee charged with reforming secret voting submitted its final bill to the floor of the lower chamber of the Italian parliament on October 13, 1988. All of the opposition parties requested that the legislature vote on the secret vote reform bill be undertaken article by article. The opposition parties hoped that this strategic approach would increase the likelihood that the reform of secret voting would fail because there would be more opportunities for defections from the party line. After a heated discussion, the Chair of the Chamber ruled that reform of the secret vote would be undertaken on a take-it-or-leave-it basis, where there would be no scope for legislators to accept or reject specific provisions through a series of ballots.Footnote ²⁷

Eventually the bill reforming secret voting passed by a narrow margin, with 323 supporting reform, 58 opposing, and 222 legislators – mainly from the PCI and Independent Left – abstaining. In this secret vote, the governing majority with 377 seats lost a considerable amount of support (54 votes, or 14 percent), and consequently the reform bill passed with a slender margin of just 7 votes (or 1 percent of the total chamber) above the required majority of 316. The PCI abstained on the final vote for two reasons: the party was officially in favor of introducing open vote for financial laws and maintaining secret vote on all the other laws; the outcome of the vote was highly uncertain, and a defeat could not be excluded notwithstanding PCI abstention.

Following the reform of October 13, 1988, use of secret ballot was, and still is, restricted according to Article 49 of the Parliamentary Rules of Procedure, to the following circumstances:

Under article 49, secret voting can be requested in “votes on amendments to the Rules, on establishing Parliamentary Committees of enquiry, on ordinary laws regarding State constitutional bodies (Parliament, President of the Republic, Government, Constitutional Court) and regional bodies, as well as on electoral laws.” The restrictions concerning the use of secret voting followed Law 362, approved on August 23, 1988, which established a fixed calendar for the submission of the budget bill and its final approval. This law gave the executive agenda-setting power on budgetary issues such as the annual budget bill.

In sum, secret voting could no longer be used in any parliamentary deliberations that had financial implications. More generally, reform of the secret vote marked the end of the “supremacy of parliament” (centralità del Parlamento) vis-à-vis the executive. This in turn contributed to breaking the consensual pattern of executive-legislative relations and paved the way for the more wide-ranging institutional reforms of the early 1990s.

Conclusion

The main purpose of this chapter has been to map out the use of secret voting in the Italian parliament. The secret vote has a long history in Italy; its origin may be traced to the Constitution of the Kingdom of Italy of 1848. In the foregoing pages the focus has been on the use of secret voting during the First Italian Republic. This chapter has shown the importance of legislative voting rules for party discipline, as a comparison among key open and secret votes reveals a dramatic increase of defections from party line when secret voting was used. This strategic use of secret voting was most strongly evident in the termination of Italian governments in instances where success in open confidence votes was nullified by defeats in subsequent secret votes.

Understanding the context in which the secret vote was reformed in Italy is critically important. In the decade prior to reforming secret voting, the use of this legislative rule grew phenomenally – by more than thirteen times between Legislatures VI and VIII. This expansion in the use of secret vote coincided with a period in which executive-legislative relations became increasingly adversarial rather than consensual. With increased competition between the executive and parliament, party leaders promoted the abolition of secret vote because it prevented them from monitoring organized intraparty factions that used secret voting as a tool to undermine incumbent governments.

Factions within the governing coalition parties often voted with the opposition, using the protection offered by the secret ballot, in order to promote their own agenda. The strategic use of secret voting by party factions had two main consequences: it undermined party discipline, and thereby reduced government stability. This, in turn, made policy making more inefficient and attenuated the effectiveness of Italy's postwar system of democratic governance.

The research pursued in this chapter may be further extended by examining three specific questions with important general implications. First, how did the use of different voting rules affect the legislative output produced by the parliament? In the Italian case, unlike most other parliamentary democracies, there are sufficient data to examine this question. For example, if an analysis were undertaken of requests for secret voting for specific amendments to annual budget bills (the final votes were always secret), it should be possible to make some inferences about the capacity of Italian governments vis-á-vis the parliament to affect the content of legislation.

Second, how did secret voting affect the use of confidence votes requested by the government and the use of the decree law procedure? This is an important question because it highlights how changes in the balance of power between executive and parliament are reflected in the strategic use of specific procedural rules. Third, was the power of Italian party factions solely tied to the strategic use of secret voting? There is evidence to suggest that party factions also used strategic abstention on key votes in which failure to reach a quorum would act as an effective block on the legislative process, and/or turned to legislative party switching under the permissive rules applied to the formation of parliamentary parties in Italy when secret vote was no longer available. This implies that the association between intraparty politics and the strategic use of parliamentary rules may be much more pervasive than previously realized.