The Formal Regulation of Electoral Governance in Mexico
by
Dr. Andreas Schedler
Abstract
Over the past decade, Mexico went through a successful transition from electoral authoritarianism to electoral democracy. Driven by pervasive distrust, political parties subjected the new, democratic electoral authorities to tight legal regulation. As the article argues, however, due to the peculiar nature of electoral governance, the reliance on a 'bureaucratic' mode of control was both more effective and less costly than current 'post-bureaucratic' approaches to public management would lead us to expect.
Over the past 25 years, the 'third wave' of global democratization has washed over 70 countries to the windy shores of electoral democracy. Responding to this unprecedented move towards democratic rule, political science has developed a whole new subfield-- democratization studies. In this new subdiscipline, scholars have explored variegated explanations of either successful, failed, or flawed transitions to democracy. They have studied the influence of a wide range of variables, such as economic well-being, political culture, international pressure, pre-democratic institutions, political leadership, choice and contingency. However, they have failed to recognize one 'crucial variable for explaining the success or the failure of democratic transitions' (Pastor 1999b), namely, electoral governance.
In general, political science has paid very scarce attention to the institutional foundations of democratic elections. In established OECD democracies, scholars could afford to ignore them because 'people take for granted the administrative dimension of elections' (Pastor 1999a: 76). As it seems, in the early transitions from authoritarian rule in Southern Europe and South America something similar happened. Those transitions took place in the context of 'usable' (Linz and Stepan 1996: 10 11) state apparatuses (such as in Spain and Argentina) or else, were able to resuscitate dormant traditions of reliable election management (such as in Uruguay and Chile).
Issues of electoral governance did therefore not acquire much visibility but rather, remained in the background during the transition processes. It was only in later transitions that questions of electoral fraud and electoral integrity moved onto the center-stage of democratization. Above all, in transitions from 'electoral authoritarianism' that strove to overcome an immediate past of manipulative elections, the control of electoral fraud turned into the very axis of the transition process.
Mexico, which experienced one of the most delayed and protracted transitions in Latin America, exemplifies the new salience of electoral governance in a particularly dramatic way. Less than a decade ago, the country's electoral institutions still worked as a machinery of fraud, subservient to the hegemonic Institutional Revolutionary Party (PRI). Vote rigging as a mechanism of last instance to protect the ruling party from losing an election was certainly not the only, but nevertheless a fundamental pillar of the post-revolutionary authoritarian regime.
Between 1990 and 1996, however, the country went through a series of negotiated electoral reforms which profoundly altered the infrastructure of political elections. Ten years ago, it was still the SecretarÌa de GobernaciÛn, the Ministry of the Interior, which was in charge of organizing federal elections. Today this responsibility lies in the hands of an independent, non-partisan administrative body, the Federal Electoral Institute (IFE). As a consequence, in contemporary Mexico, elections are now widely perceived to operate in a fundamentally democratic way as free, competitive, inclusive, clean, and even fair contests.
How did electoral designers accomplish this remarkable transition from electoral fraud to electoral fairness? Which kind of institutional solutions did they put into practice in order to effect such a dramatic change. How did they manage to subject the Mexican Leviathan to democratic rules, despite its historic reputation of subverting or breaking formal rules at its convenience? The present article does not pretend to analyze the complex package of interlocking 'safety measures' reformers designed to keep the new Federal Electoral Institute under control. It limits its attention to one key instrument of restraint: bureaucratic regulation the enactment of a dense network of legal rules and regulations designed to minimize the discretion of election officials.
After a brief introduction into the role of distrust in Mexico's transition process, the article describes three representative dimensions of bureaucratic regulation in Mexico's electoral code: the requirements of record keeping, identity controls, and time schedules. The analysis reveals a pervasive reliance on detailed formal restrictions that runs against the spirit of current 'post-bureaucratic' trends in public administration that put heavy emphasis on deregulation.
Yet, as the article argues, the 'entrepreneurial' option was blocked by the prevailing levels of distrust towards the ruling party. Furthermore, the bureaucratic route was more effective than the 'post-bureaucratic' approach would suggest (since formal constraints were embedded in a comprehensive system of accountability and also, sustained by a new balance of power between government and opposition); and it was less costly than the new public administration school would expect (since the distinctive features of electoral administration mitigate many of the pathologies bureaucratic over-regulation tends to produce in other spheres of public administration).
Democratization and Distrust
The 'hegemonic party system' that reigned in Mexico until the late 1980s represented an unusual kind of authoritarian regime. It was a showcase of 'electoral authoritarianism' that legitimated itself through the regular conduct of non-competitive political elections (since 1916). These elections were not mere one-party rituals. The ruling Institutional Revolutionary Party (PRI) did allow opposition candidates to run. Yet, the only genuine, even if weak and ineffective, opposition party was the conservative National Action Party (PAN), founded in 1939. All others were either short-lived PRI dissident movements or so-called satellite parties, ancillary parties of the PRI who were there to simulate a superficial appearance of inter-party competition.
Opposition parties, however, were not supposed to win elections. Their function was to legitimate not to challenge the majority party. The 'possibility of a rotation in power [was] not envisaged' (Sartori 1976: 230). In order to reproduce its political hegemony at the polls, the PRI could rely on a vast array of undemocratic mechanisms. It could exclude political actors from electoral competition; it could try to co-opt or repress them; it could exploit its monopolistic hold on public resources; it could enforce the subservience of the mass media; but last but not least, it could resort to electoral fraud anytime it seemed convenient.
Controlling the organization of election in a centralized fashion, the PRI was able to alter electoral results either to inflate its victories (cosmetic alchemy) or to 'correct' its defeats (decisive fraud). It is probable, thought, that the main function of vote-rigging was preventive. In its golden times of hegemony, the ruling party had little need to alter electoral outcomes. In many places it did not face any opposition at all, and until 1982, it never received less than three quarters of the vote at presidential contests. Still, the manipulation of electoral results always loomed in the background as a powerful 'mechanism of last resort.' Every actor in the political game knew about it and adjusted his or her expectations accordingly. Power, as we know, is at its strongest when it has to be used less. It is a sure sign of erosion when it has to make good on its threats.
Accordingly, as the PRI's hegemony began to erode, it first reacted by stepping on the brakes of electoral fraud in order to halt the accelerating train of democratization. In the municipal elections of 1983, while the country was hit by its worst economic crisis since the 1930s, the PAN turned to more self-confident and confrontational strategies and actually succeeded to win several major cities in Northern Mexico. In response, president Miguel de la Madrid (1982 88) quickly decided that it was too risky to use the electoral arena as a safety valve for social discontent. In the remaining of his term, he tried to contain the emerging opposition through a mixture of electoral fraud and a reassertion of control over electoral administration.
Privileging electoral control over electoral legitimacy, the PRI was able to hold its exclusive positions of power, excepting some local defeats, at all levels of government. But the perpetuation of power came at a price. The increasing reliance on electoral fraud provoked an escalation of post-electoral conflicts and ended up hollowing out whatever may have remained of the regime's procedural legitimacy. This self-reinforcing cycle of increasing competition, fraud, conflict, and discredit reached its culminating point in the 1988 presidential elections.
In this highly competitive race, the official candidate, Carlos Salinas de Gortari, defeated the conservative candidate Manuel J. Clouthier as well as the PRI breakaway CuauhtÈmoc C·rdenas in what is largely regarded 'the most egregious violation of norms for free and fair elections in modern Mexico' (Eisenstadt 1999: 84). The massive evidence of fraud pushed the country into a the dramatic uncertainties of a 'confusing postelectoral conflict' (Loaeza 1999: 24), laying open an acute crisis of legitimacy.
President Carlos Salinas (1988 94) soon embarked on repairing his tarnished 'legitimacy of origin' by convoking parties to negotiate reforms to the electoral system. These as well as all subsequent reforms faced one central challenge: 'to redress an ancient distrust into electoral procedures' (Woldenberg 1999: 68). The antinomy of distrust and credibility came to represent the guiding distinction of a whole decade of electoral reforms. In the end, after four rounds of intensive inter-party negotiations, these distrust-driven reforms enacted in 1990, 1993, 1994, and 1996 led to a radical transformation of Mexico's electoral institutions.
In order to subject the electoral Leviathan to credible restraint, political parties set up a comprehensive scheme of interlocking institutional constraints. Among other things, they wrote a new electoral law, the Federal Code of Electoral Institutions and Procedures (COFIPE). They took the administration of elections out of the hand of the Ministry of the Interior, and established a permanent, non-partisan, and autonomous election management body, the Federal Electoral Institute (IFE). They designed a specialized judicial tribunal to settle electoral disputes. They formulated a catalogue of electoral crimes backed by harsh penal sanctions, to be enforced by a special persecutor. They revamped the electoral registry and issued new, high-tech and high-security voter ID cards. And they established a 'panoptic' regime of party surveillance, allowing political parties to monitor each step of the electoral process.
It is no exaggeration to conclude that these and other 'confidence building measures' added up to a genuine refoundation of the country's electoral infrastructure. So, distrust bore many children, although the present article will limit its attention to only one of them: the impetus to reduce administrative discretion through tight and detailed bureaucratic regulation. It will document the 'baroque' density of formal rules that characterizes Mexico's electoral law by analyzing three classical areas of bureaucratic constraint: record keeping, identity controls, and time rules.
Record Keeping
The modern bureau, the modern office, is a warehouse of wood (in its processed form as paper). Bureaucratic rule, to paraphrase Lincoln, is the rule of the paper, by the paper, and for the paper. Forms and files are the souls of bureaucratic Leviathan, and the requirement to record and report everything official in writing represents one of its basic principles of action.
Electoral administration in democratic Mexico is a neat case of bureaucratic organization based on record keeping. Here as anywhere else in the world of modern bureaucracy, the written form serves the main purpose of certainty and control. Red tape is, so to speak, the material basis of bureaucratic accountability. Hence, Mexico's electoral reformers drew up a complex network of interlocking pieces of paper documents, applications, forms, files, originals, copies, minutes, certificates, notifications, receipts, and reports in order to ensure that all administrative steps follow uniform, controlled, and, above all, controllable procedures that enable parties to trace eventual irregularities back to their original sources. As a consequence, Mexico's electoral law code turns each federal election into a gigantic paper mill. Four particularly paper-intensive processes may illustrate this point: voter registry, candidate registry, voting, and vote counting.
(1) Voter registry
Since the early 1990s, the construction and regular up-date of the voter registry in Mexico is a thoroughly computerized high-tech enterprise. For all its technological sophistication, however, the registration of voters still gives rise to considerable stocks and flows of paper. For instance, citizens, the main clients of the electoral bureaucracy besides political parties, have to fill out application forms to be included into the electoral list [143.1 and 148.1], submit documents that prove their identity [144.2], receive a receipt of their application [148.3], sign a receipt when being handed over their voter card [144.4], submit documents and application forms to up-date their personal data [146.4], fill in and submit forms to appeal against their eventual exclusion from the electoral list or against the eventual non-receipt of their voter card [151.6], and receive the notification of the corresponding rulings [151.7].
(2) Candidate registry
The formal registration of electoral candidates follows the same principle as the registration of voters: Bureaucracies define the conditions under which external agents may gain recognition as official objects of administration (clients) and trigger internal decision-making processes (administrative acts). More often than not, the bureaucratic conditions for external communication include rigid demands for written documentation and standardized forms. To register candidates for elective office at federal level, political parties have to submit, for each candidate, an application with the candidate's personal data, a declaration of acceptance, a copy of her or his birth certificate, a copy of her or his voter card, and a proof of her or his place of residence [178.1 and 2], a written confirmation that the candidate was selected according to prevailing party rules [178.3], and a certificate that the party has registered the minimum number of candidates it is required to register for plurality deputies [178.4] and plurality senators [178.5].
(3) Voting
While the preparatory steps of Mexico's federal elections such as the accreditation of electoral observers [5.3], the selection of polling station officials [193], the determination of polling station locations [195 6], the nomination of party representatives [202 4], and the delivery of the voting material to polling station officials [207 8] already involve extensive paper work, the requirements for record keeping are particularly tight at voting day. The comprehensive polling station file (expediente de casilla) [234.1] that electoral officials have to compile in the course of the day must contain the voting day file (acta de la jornada electoral) [212.4] composed of a detailed file recording the initiation of the voting process [212.5] and a another one reporting on the closure of the voting process [225]; a file that states the results of vote counting and computing (acta de escrutinio y cÛmputo) [232.1]; and eventual reports on public disorders [220.2] or other incidents [221] as well as eventual notes of protest submitted by party representatives [199.1.f and 234.1.d].
Once these documents are put together, political party representatives (who are free and in some cases even obliged to sign all of them) receive a 'legible copy' of the complete file, against the obligatory receipt [235.1]. The formal requirement of legibility, of course, is symptomatic for the surviving assumption of bad will towards electoral officials who in the past displayed a considerable capacity of subverting formal rules.
(4) Vote counting
If vote counting at district, state, and federal level constitutes the culminating event of the electoral process, it also marks the culmination of electoral paperwork. As before, COFIPE demands files and forms for everything: files to record the reception of voting materials from polling stations [238.6 and 242.2], forms to fill in the results computed at polling stations [243.1.c], forms to state the identity of polling station files and District Council files [247.1.a], files to document procedures and findings in case of altered electoral packages [247.1.d], files to report the aggregate results at district level [247.1.e and g], as well as detailed files to give a summary of vote counting at district level, including its results, eventual incidents, and declarations of validity and eligibility [247.1.i].
Since there are up to three federal elections at a time (for the presidency and the two legislative chambers) which furthermore are conducted on the basis of a mixed electoral system (that combines plurality formula with proportional representation for both the Chamber of Deputies and the Senate), the paperwork associated with vote counting multiplies accordingly. At the end of the vote count, the whole process of large-scale paper production flows into a final, apotheotic act of bureaucratic ingenuity: the distribution of those files to different points in the system of electoral administration and dispute adjudication [see 252 and 253].
Identification
Any administrative apparatus, if it wants to make good on the bureaucratic promise of reliability, has to ensure clarity about and control over its basic units of operation (clients, files, and decisions). It has to define criteria of eligibility that allow to identify the beneficiaries (or victims) of its services. It has to set up systems of storage and communication that prevent files from getting lost, altered, or misdirected. And it has to issue documents that give some recognizable form to its final decisions. Given the country's history of electoral fraud, the organization of elections in Mexico reflects a particularly intense concern with establishing and safeguarding the identity of its main administrative objects. In accordance with the nature of the task at hand putting the democratic principle of 'one person, one vote' into administrative practice the federal electoral law focuses on two fundamental challenges: ensuring the identity of the its core clients (citizens) and guaranteeing the integrity of its core documents (ballots).
The Identity of Citizens
In the golden days of authoritarian PRI hegemony, the voter registry represented a renown major source of vote rigging. In essence, it worked as a mechanism of 'weighting' voting rights: It denied suffrage to those people 'shaved' from the list while multiplying the votes of those with various entries. While some of the inaccuracies may have had its origins in technical problems, opposition parties always suspected, and with good reasons, that the voter list's deficiencies were not randomly distributed but rather reflected a clear political bias over-representing regime loyalists while under-representing presumptive protest voters.
As soon as questions of electoral cleanness (as different from issues of political representation) entered Mexico's electoral reform agenda in the 1980s, it was clear that the country was not to have clean elections without first setting up an accurate voter registry. It was clear as well that this was not to be an easy assignment - in a country where the State does a lousy job in identifying its subjects at birth, counting them in regular national censuses, and registering them for military service, the payment of taxes, or the provision of welfare services; as well as in a country where population mobility (internal and external migration), problems of personal identification (most people do not have a passport), problems of location in space (homonymous and anonymous streets), not to speak of sheer factors of size (numbers and spatial distances), pose formidable objective obstacles to any pretension of subjecting the population to the epistemological requirements of a modern state bureaucracy (see Scott 1998).
This it not the place to recount the enormous efforts in terms of time, money, personnel, expertise, and technology both the IFE and political parties have invested since 1990 in order to set up and run a new fully computerized voter registry as well as to issue (twice) a new high-tech and maximum-security voter card (for an exhaustive review and analysis, see Luj·n 1997). Here, I would just like to list some of the formal rules the electoral code defines in order to prevent that voter lists and voter cards (as well as other related documents) be altered or misused.
The law establishes that voter cards have to fulfill minimum standards of safety (which actual cards exceed by far) in order to render them forgery-proof [164.1 and 2], that non-utilized blanks of voter card will be stored in 'a place that guarantees their integrity' [144.5], that voter cards replaced for whatever reason have to be destroyed [150.2], that invalidated voter cards have to be destroyed [163.4], that District Councils must double-check whether the official voter list corresponds to the list parties received for purposes of control [161.3], that the General Council may carry out further random checks in order to establish the identity of these lists [161.4], that polling station Secretaries mark the voter cards as well as the right-hand thumb (with indelible ink) of voters who have made use of the ballot [218.4], and that an 'institution of renown reputation' certifies the quality of the indelible ink before election day [208.3], while the General Council must carry out a random check of its remainders after election day to make sure that it had not been replaced or altered [208.4].
It may be useful to recall that numerous mechanisms of accountability that give parties as well as citizens a say in the prior stages of document production complement the provisions designed to protect the integrity of the official documents and materials that serve to identify the enfranchised citizenry.
The Integrity of Ballots
The elaborate safety measures the electoral law defines concerning ballots begin with their design. COFIPE determines the minimum information each ballot must contain [205.2 6], prescribes a paper design that allows to maintain full control over the number of ballots in circulation at any stage of the process [205.2.d], and asks the General Council to introduce any further security marks it should deem necessary [205.1]. Once printed, the law further stipulates, ballots cannot be changed anymore [206.1]. About three weeks before election day, IFE has to distribute the ballots to District Councils. Naturally, the electoral Code contains detailed provisions for safeguarding the integrity of ballots throughout their delivery and subsequent storage [207]. 'Authorized personnel' of the Institute must deliver the ballots at some pre-established day, hour, and place [207.2.a]. A detailed file of reception must record the quantity of ballots delivered, the way they are packaged, and the names and positions of the officials witnessing the act [207.2.b].
The District Council must then safeguard the material at some pre-established place, guarantee its integrity through wrappers stamped and signed by all persons present, and again record the whole process in writing [207.2.c]. The next day (at latest) the Council must recount the ballots, stamp them, and group them in packages to be delivered to polling stations the week before election day [207.2.d]. And so forth.
This pre-electoral stream of bureaucratic activities planning, filing, counting, sealing, stamping, signing, storing, and recording aims at retaining control over the number of ballots in circulation as well as at preventing the falsification, manipulation, and destruction of ballots before voting day. During voting day, transparent ballot boxes [209.1] fulfill essentially the same function. In addition, party representatives carry the duty to certify the authenticity of all election day files through their personal signatures [200.2, 214.1, 233.1, 216.3, 220.2, and 225.2]. They may even sign ballots if they wish, both before voting day [207.3] and at the initiation of the voting process [212.3], as well as the whole package of documents and materials polling station officials put together at the end of the voting day [233.1].
Incidentally, this must be considered a sweet irony of bureaucratic control: a bureaucratic apparatus, the archetypal expression of modern impersonal rule, making use of the signature, a typical manifestation of modern individual identity.
Once polling stations close, complex counting procedures strive to ensure that ballots enter the electoral equation the way they should: one ballot, one vote. None lost, none added, none altered. The legal set of rules of vote counting includes rules about the sequencing of vote counts [228 and 229.1]; rules of distinction between valid and nil votes [227.2, 230, and 232.3]; an elaborate division of work and mutual controls between polling station officials [229.1]; the requirement of crossing out counted ballots and putting them into a special envelope [229.1.a], extensive requirements of reporting and record keeping [232.1 and 2]; and the continuing presence of party representatives until the end of the process [237.1].
After this operation, the ballots together with all other files and materials take their way from the polling station to the District Council where they are not just summed up to aggregate results but also double-checked and eventually recounted. In these later phases of vote counting, the electoral law does not lessen at all its firm grip over the ballots and all other electoral documents [see 234, 235, 238, and 242 63]. But at this point, tracing this long chain of 'identity checks' further, would only come at the unacceptable price of boredom.
Time Rules
As JosÈ Woldenberg, the President of the Federal Electoral Institute, rightly states, COFIPE, with all its detailed procedural regulation, represents more than just a legal framework for electoral administration. It rather resembles an exhaustive 'manual of procedures' (1998: 29). One of the most prominent aspects of procedural regulation as contained in the electoral code concerns time rules.
These time rules are of diverse nature. They include rules of duration, rules of sequencing, rules of frequency, as well as rules of initiation and termination. But most of them are rules of timing that define schedules of administrative action, either by delineating periods or points in time when something may or must be done or by fixing deadlines up to which something may or must be done. Let us give just four instructive examples of aspects of the electoral process which are embedded in a dense network of formal temporal constraints: nominal lists, voter cards, candidate registration, and voting day.
(1) To begin with, one of IFE's core activities which is thoroughly constrained by legal time rules is the management of the nominal voter list. Among other things temporal, the law stipulates a deadline for the annual delivery of nominal lists to District Juntas for public exhibition [156.1], a deadline for returning the lists from municipal authorities to District Juntas [157.1], another deadline for returning the lists from District to Local Juntas [157.2], an annual period when parties dispose of the nominal lists [158.1], a deadline for reporting eventual modifications made by the Institute [158.4], a time period during which parties may appeal against these changes [158.5], a day when IFE must deliver magnetic tapes of the nominal list to parties in election years [159.1], a day when it must deliver the hard copies [159.1], a deadline up to which parties may formulate their objections [159.2], a deadline to report about eventual ensuing modifications made by the IFE [159.3], a deadline for parties to appeal against these changes [159.4], a deadline for printing nominal lists with photographs [161.1], and a deadline for delivering them Local Councils [161.1].
(2) With regard to the distribution of voter cards, COFIPE determines deadlines for citizens to pick up their voter cards [144.1 and 163.1], a deadline up to which IFE has to hand over a list of cancelled voter cards to political parties [163.2], a period during which this list must be exhibited in public [163.3], a deadline for soliciting the replacement of lost cards [164.3], the requirement to immediately destroy replaced voter cards [150.2], a deadline for destroying the blanks of unsolicited cards [163.4], and a period during which IFE must store the data of those citizens who, for whatever reason, have their name taken off the voter registry [163.9 and 10].
(3) With respect to the official registration of candidates, the electoral Code defines a period for registering campaign platforms [176.2], periods for registering diverse kinds of candidacies [177.1], a maximum time for IFE to check whether candidates fulfill all legal requisites [179.1], the requirement of immediate notification to parties [179.2], a maximum time for parties to meet missing requisites [179.2] as well as to remedy eventual multiple candidacies [179.3], a maximum time for IFE Councils at all levels to session after these deadlines have passed [179.5], the requirement that Local and District Councils immediately notify the registration of candidates to the General Council and the other way round [179.6 and 179.7], as well as, finally, deadlines for replacing candidates [181.1 and 2].
(4) Another phase in the electoral process which is subject to tight temporal constraints is, of course, voting day. The law establishes the day and hour when polling stations must be installed [212.2], an express prohibition of installing them before [212.6], a contingency schedule for installing polling stations in different ways in case the normal procedure cannot be followed [213.1 and 213.1.f], the requirement of permanent presence of polling station officials during election day [212.7], the provision that general party representatives may remain in polling stations only 'the time necessary' to do their job [219.4], a requirement of immediate notification in case the voting process has to be suspended [216.2], the hour of closure of the voting process plus rules of exception [224], deadlines for delivering electoral results and all documentation to District Councils [238.1], and provisions about their eventual extension [238.2 and 5].
This illustrative list of time rules is, of course, far from exhaustive. The Code sets further temporal constraints for the up-dating of the voter registry [146.1, 147.1, and 151.2 and 3], the distribution of electoral propaganda [190.2], the publication of opinion polls [190.4], the multiple-stage random selection of polling station officials [193.1], the nomination of party representatives at polling stations [198.1 and 201 3], the determination of polling stations' physical location [195.1], the delivery of voting material to polling station officials [208.1], vote counting at district and state levels [246.1 2 and 255.1], and so forth.
Basically, the electoral law attaches a deadline to any right and obligation it assigns. This fulfills two basic functions. On the side of obligations, time rules limit power. Decision-makers do not just have an abstract obligation to do something. They have to fulfill it within determinate time limits, and may be held accountable for it. On the side of rights, time rules legitimate power. If those who hold certain rights, be it political parties or individual citizens, do not make use of them within the time frame set by the law, the responsibility for forfeiting their rights falls back on them themselves. In other words, subjecting administrative agents to deadlines shifts power to their clients, while subjecting claimants to deadlines shifts power to administrative authorities. In both cases, the responsibility for any eventual failure to act in a timely fashion lays with the one who ignores the legally established time limit.
Distrust and Bureaucracy
The dense network of rules Mexico's electoral code spreads out is distinctively and classically bureaucratic in spirit. As such, it runs diametrically against the anti-bureaucratic or post-bureaucratic Zeitgeist of current public administration theory. The country's electoral law embodies everything the new 'managerial' public administration school abhors. Oblivious to the new discourse that hails flexibility, risk-taking, innovation, experimentation, empowerment, personal responsibility, problem solving, employee participation, worker motivation, voluntary compliance, competition, market mechanisms, product orientation, service and customer orientation, and so forth, the electoral code pursues with single-minded determination to establish a paradigmatic Weberian bureaucracy.
Rather than assigning the distinction between politics and administration 'to the dustbin of history' (Altshuler 1992: vii), Mexico's electoral code strives to institutionalize it. Rather than adopting the (so to speak) administratively correct focus on results, it embraces 'the obsolete focus on rules' (Barzelay 1992: 125). Rather than empowering public officials, setting them free 'from the constraints that limit experimentation,' and accepting 'a calculated degree of risk' (Anechiarico and Jacobs 1996: 203), it does everything to minimize their discretion. Rather than fostering an 'entrepreneurial culture' (Thompson and Riccucci 1998: 236) of innovation and creativity, it tries to build a machine-like apparatus that would work with maximum 'precision, constancy, discipline, tightness, and reliability, hence: calculability' (Weber 1972: 128). In sum, rather than 'reinventing government' (Osborne and Gaebler 1992), Mexico's electoral reformers set out to forge an old-fashioned iron 'cage of subservience' (see Weber 1988: 332).
However, one may well ask: How else could it have been? How else, given the starting point a history of fraud and abuse that impregnated electoral governance with deep and pervasive distrust as well as the corresponding task at hand constructing electoral institutions subject to effective as well as credible restraints. After all, the Federal Electoral Institute was never thought to be a simple service organization but an 'agency of restraint' (Collier 1999).
The reinvention movement in public administration rests on one pivotal assumption: We can (and should) trust public officials. Its central recommendation to free state employees from the bureaucratic bondage of rules and regulations makes sense only under the strong premise that public officials will not abuse their enlarged degrees of freedom. The 'post-bureaucratic paradigm' (Barzelay 1992) correctly perceives that the bureaucratic obsession with control, regulation, discipline, surveillance, and punishment flows out of a deeply distrustful attitude toward public officials. Rather than treating employees as 'trusted servants' it views them as 'potential criminals' (Anechiarico and Jacobs 1996: 62).
Trust, however, is a scarce resource that cannot be generated neither by administrative fiat nor by scholarly exhortation. To tell the participants in Mexico's electoral reforms: trust each other! would have hardly paved the way to clean and credible elections. Distrust, once it settles into a self-reinforcing 'equilibrium,' must be regarded a 'hard' social constraint that cannot be simply wished away. It works as a forceful 'social mechanism' (see Hedstrm and Swedberg 1998) that generates a limited set of consequences. Rational choice theory, the anthropology of distrust, has done a good job in specifying the possible answers to distrust. If I distrust you I may either stop dealing with you or else, constrain, monitor, reward or punish you. The bureaucratic elimination of discretion is just one of several plausible strategies to manage distrust. But it is one of the most common ones.
Distrust, thus, breeds bureaucracy. As Thompson and Riccucci state, 'low levels of citizen trust in government encourage administrative approaches at odds with reinvention approaches based on rules, investigations, audits, procedural requirements, and control' (1998: 255). It is accordingly hard to see how Mexico's political parties could have chosen a trust-based strategy of loose 'post-bureaucratic' regulation, rather than their high-security strategy of watertight 'bureaucratic' regulation.
The Embedded Bureaucracy
If the prevailing level of distrust towards the PRI state left electoral reformers with little choice other than locking the new election management body into a bureaucratic cage of detailed rules and regulations, one may still wonder whether those formal constraints were actually effective in preventing fraud (and if so, at what price). As a matter of course, the effectiveness of formal rule changes cannot be taken for granted, and even less so in a context like the Mexican. Formal restrictions are 'frequently blunted, distorted, and subverted so that what remains is form rather than substance' (Anechiarico and Jacobs 1996: 17).
Also, altered formal rules often run against the padded walls of persisting informal rules. Furthermore, in the Latin American context, authors have argued that the region's new democracies represent 'informally institutionalized' polities whose everyday politics suffer from a 'visible and apparently increasing gap' between 'formal rules and actual behavior' (O'Donnell 1996: 35, 40 1). In Mexico, specifically, the pervasive 'rule of lawlessness' has been a recurrent topic in political and academic debate.
Still, as they redesigned the institutional infrastructure of electoral governance, Mexican political parties displayed an enormous faith in legal regulation. Electoral law reform constituted the very axis around which the country's protracted process of political democratization revolved. But formal rule changes did not just provide an illusionary appearance of change. They represented an effective, powerful lever that allowed the democratic opposition to gradually lift the authoritarian regime out of its hinges. Today, most national as well as international observers would agree that these reforms indeed succeeded in bringing the age-old problem of electoral fraud under control; that elections in contemporary Mexico have been effectively transformed from manipulative authoritarian rituals to democratic contests free, fair, competitive, and clean (see, for example, Silva-Herzog 1999 and Loaeza 1999).
How can we explain this successful transition from electoral fraud to electoral integrity? How can we explain the apparent effectiveness of the new set of legal constraints? The most obvious answer is: Formal changes were effective because they did not come alone. As we mentioned above, bureaucratic regulation is embedded in a comprehensive institutional architecture that comprises numerous interlocking mechanisms of restraint. Furthermore, the particularities of electoral administration (as we will analyze them below), above all, its cyclical nature, may help to render traditional instruments of bureaucratic control more effective than in other fields of public administration.
But most importantly, in the course of Mexico's democratic transition, it were not just formal rules that have changed. The underlying realities of power, the prevailing correlation of forces between political parties, have undergone dramatic changes as well. Gone is the extreme asymmetry between state party and opposition that sustained the monopolistic control of electoral governance during decades. Today, opposition parties are not only able to win elections; they are also able to oversee and defend their victories.
In Mexico, electoral fraud and electoral competition have evolved in an inverted-U-shaped curve. First, in the 1980s, as the party system became more competitive, the PRI responded with an intensification of fraud. Later, however, in the 1990s, the increasing competitive strength of the democratic opposition made vote-rigging a practice that was more and more difficult to sustain. Furthermore, it was not just opposition parties that came to act as powerful 'agents of accountability.' Over the past decade, civic associations and mass media, too, have developed impressive abilities to monitor electoral processes.
The Distinctive Bureaucracy
So far we argued that the bureaucratic regulation of electoral governance was quite inevitable given its primary function of providing credible restraint but also quite effective given its embeddedness in a wider network of restraint, enforcement, and monitoring. The question remains, however: How costly has the reliance on a bureaucratic mode of control been? In comparative terms, running elections has been seemingly more expensive in Mexico than in most Western European and Latin American countries (see LÛpez-Pintor 1999: 64 74). In part this reflects the fact that the Mexican electoral budget includes the costs of voter registration, a task which in most other OECD countries falls under the purview of general registration offices within the regular state administration. Also, the growing experience of the Federal Electoral Institute has been leading to considerable efficiency gains over time.
It is unclear, however, how much cheaper the conduct of elections in Mexico would be if the electoral code would leave somewhat larger margins of discretion to electoral officials. But more importantly, it seems that in electoral governance, bureaucratic regulation does not generate the same kind of 'bureaucratic pathologies' as we have learned to know and to fear them in other spheres of public management. Electoral administration is different in at least in four regards which tend to reduce the costs of a bureaucratic mode of control by detailed formal rules.
(1) The periodical task
Complying with the strict and detailed set of rules the Mexican electoral code imposes on election officials may be complicated, burdensome, time-consuming, and expensive. But it's just once every three years. The bulk of the rules do not guide the everyday operation of a bureaucracy producing a permanent flow of homogeneous decisions. The Federal Electoral Institute does accomplish some continuous tasks, such as maintaining the voter registry, auditing political parties, monitoring the mass media, and educating the citizenry. But its main function, the organization of elections, is cyclical. A modern, representative democracy is based on regular elections, on the institutional expression of popular preferences at determinate intervals. It is not what Ernest Renan, in his famous definition, thought a nation to be: a daily plebiscite (Renan 1882: 307).
Thus, much of the bureaucratic framework that regulates the organization of elections is effective only for specific tasks in delimited periods. As a consequence, it will hardly generate the same kind of costs as a permanent straightjacket of formal rules. Rather, it will tend to produce delimited and transitory 'pathologies' only. For example, it is hard to see how the citizen who happens to be chosen to work at a polling station, will develop, in the course of one day, the same blind routines and the 'bureaucratic personalities' that are said to afflict officials in other walks of bureaucratic life.
(2) Citizens as agents
Pursuing the Weberian ideal type of a rule-bound, neutral, and predictable administrative machine, the Federal Electoral Institute has been trying, since its inception (in 1990), to implement 'the quintessential reform in the history of public administration' (Anechiarico and Jacobs 1996: 39): the professionalization of public officials through a merit-based civil service system. Polling stations, however, are not staffed with professional employees. Rather, to staff the over 100.000 polling stations the Institute sets up at each election, it selects ordinary citizens on the basis of a complicated two-stage random procedure. Accordingly, all the 'line agents' who do the face-to-face interaction with voters at voting day are not full-time bureaucrats but ordinary citizens.
The numbers are impressive. At the 1997 mid-term legislative elections, more than 730.000 citizens (Becerra and Salazar 1999: 408) contributed to turn the contest into a 'great democratic party' (Ernesto Zedillo). This looks like an unusual degree of community control, as Vincent Ostrom demanded it more than ten years ago (1989) as a centerpiece of public administration reform.
(3) The procedural product
According to the reinvention school, public administration should abandon its 'obsolete focus on rules' (Barzelay 1992: 125) and concentrate instead on its 'mission.' Products, not procedures have to be the central concerns of the 'post-bureaucratic' public official. As Thompson and Riccucci synthesize, 'employees are to be freed from many of the rules that currently limit their discretion' in order to hold them 'accountable for the results of their activities' (1998: 236).
In the realm of electoral administration, however, the procedures are the product. What are election management bodies supposed to deliver? Very simple: clean procedures. This is what we hold them accountable for: the correct and efficient implementation of a complex body of rules and procedures. They do not have a mission beyond. Election management bodies are not supposed to provide welfare, redemption, or entertainment. They are supposed to provide the procedural foundations of liberal democracy. It would be intelligible to election officials if we told them to focus on their 'mission' instead of worrying about the rules. In the organization of elections, privileging rules over substance is not a symptom of pathological 'goal displacement' (Anechiarico and Jacobs 1996: 179) but the very condition of democratic success. Substituting (post-modern?) substantive legitimation for the modern 'legitimation by procedures' (Luhmann 1983) may be problematic in many spheres of public administration. In the realm of electoral governance, it is simply unthinkable.
(4) The function of rules
'A multitude of rules and regulations appears to be the very essence of a bureaucracy' (Perrow 1986: 20). The corresponding idea, however, that detailed rules and regulations represent the core symptom of the bureaucratic disease, seems to derive from the plethora of existing rules that do not make sense. But rules may be more than meaningless noise. In general, 'rules protect those who are subject to them' (ibid.: 24). In the realm of electoral administration, legal rules translate political rights into political realities. Insofar they are visibly linked to this goal, they are rules that make sense. In Mexico, the function of electoral rules to empower citizens with voting rights as well as their context of pervasive distrust between parties provide tangible justifications for most provisions, even if the same provisions would look abstruse in a different setting characterized by lower stakes and higher trust.
Conclusion
The present article reconstructed the bureaucratic structure of Mexico's election code in three regulatory areas: record keeping requirements, identity controls, and time rules. In our concluding remarks, we would like to highlight the two main general hypotheses we put forward regarding the causes and consequences of bureaucratic regulation in electoral governance. On the one hand, looking at bureaucratic grids of detailed rules as a dependent variable, the article identified distrust as the central 'social mechanism' that makes plausible why political parties in democratizing Mexico devised a bureaucratic set of rules that strives to minimize administrative discretion, rather than opting for a less restrictive 'managerial' design of electoral administration. 'Post-bureaucratic' approaches that privilege responsibility over control presuppose a decision to trust public officials. Strategic actors who are stuck into low-trust equilibria would be foolish to take such a risky decision. They will always prefer security over trust.
On the other hand, looking at bureaucratic regulation as an independent variable, the article concluded that Mexico's electoral code contradicts the widespread assumption that bureaucratic rules and procedures are 'not a very effective mode of control' (Przeworski 1999: 27). Moreover, as one starts to think about the distinctive nature of electoral governance, one is pressed to conclude that in this realm, a high density of formal constraints is less likely to generate the same kind of 'bureaucratic pathologies' as in other spheres of public administration.
In sum, our explorative expedition into the under-researched terrain of electoral governance took us some steps towards the goal of 'specifying the conditions under which one management style might succeed or fail' (Thompson and Riccucci 1998: 254). Distrust breeds bureaucracy, we stated, and in new democracies, electoral governance is likely to be infected by the aggressive virus of distrust. Yet, at the same time, the field is also likely to be more resistant to the corrosive disease of bureaucratic pathologies.
Mexico City, FLACSO, October 1999
Notes
I am indebted to the Austrian Academy of Sciences for supporting work on this paper through the Austrian Program for Advanced Research and Technology (APART). Also, I would like to thank Carlos Santiso and Richard Snyder for valuable comments on an earlier version of this article.
References
Altshuler, Alan A. (1992), 'Foreword,' Breaking through Bureaucracy: A New Vision for Managing in Government, Michael Barzelay, with Babak J. Armajani (Berkeley: University of California Press), pp. vii xii.
Anechiarico, Frank and James B. Jacobs (1996), The Pursuit of Absolute Integrity: How Corruption Control Makes Government Ineffective (Chicago and London: University of Chicago Press).
Barzelay, Michael, with Babak J. Armajani (1992), Breaking through Bureaucracy: A New Vision for Managing in Government (Berkeley: University of California Press).
Becerra, Ricardo and Pedro Salazar (1999), 'La elecciÛn federal de 1997: una descripciÛn general,' 1997: Elecciones y transiciÛn a la democracia en MÈxico, ed. Luis Salazar (Mexico City: Cal y arena), pp. 43 75.
Collier, Paul (1999), 'Learning from Failure: The International Financial Institutions as Agencies of Restraint in Africa,' The Self-Restraining State: Power and Accountability in New Democracies, eds. Andreas Schedler,
Larry Diamond, and Marc F. Plattner (Boulder and London: Lynne Rienner Publishers, 1999), pp. 313 30.
Diamond, Larry (1999), Developing Democracy: Toward Consolidation (Baltimore and London: Johns Hopkins University Press).
Eisenstadt, Todd A. (1999), 'Off the Streets and into the Courtrooms: Resolving Postelectoral Conflicts in Mexico,' The Self-Restraining State: Power and Accountability in New Democracies, eds. Andreas Schedler, Larry Diamond, and Marc F. Plattner (Boulder and London: Lynne Rienner Publishers, 1999), pp. 83 103.
Foucault, Michel (1979), Discipline and Punish: The Birth of the Prison (New York: Vintage).
Greif, Avner, Paul Milgrom, and Barry R. Weingast (1995), 'Coordination, Commitment, and Enforcement: The Case of the Merchant Guild,' Explaining Social Institutions, eds. Jack Knight and Itai Sened (Ann Arbor: University of Michigan Press), pp. 27 56.
Hedstrm, Peter and Richard Swedberg (1998), Social Mechanisms: An Analytical Approach to Social Theory (Cambridge, UK: Cambridge University Press).
Instituto Federal Electoral (1996), CÛdigo Federal de Instituciones y Procedimientos Electorales y otros ordenamientos electorales (MÈxico, DF: IFE).
Kiewiet, D. Roderick and Mathew D. McCubbins (1991), The Logic of Delegation: Congressional Parties and the Appropriations Process (Chicago: University of Chicago Press).
Loaeza, Soledad (1999), Incertidumbre y riesgo en transiciones prolongadas: La experiencia mexicana y el Partido AcciÛn Nacional, paper prepared for presentation at the conference 'Pathways to Democracy: Specifying Protracted Transitions', Mexico City, El Colegio de MÈxico, 19 21 May.
LÛpez-Pintor, Rafael (1999), Electoral Management Bodies as Institutions of Governance (New York: UNDP), unpublished manuscript.
Luhmann, Niklas (1983), Legitimation durch Verfahren (Frankfurt/Main: Suhrkamp).
Luj·n Ponce, NoemÌ (1997), TecnologÌa y procesos electorales en MÈxico (1988 1994): Entre la democracia y el control polÌtico (Mexico City: FLACSO), PhD Dissertation.
Miller, Gary J. (1992), Managerial Dilemmas: The Political Economy of Hierarchy (Cambridge: Cambridge University Press).
Morales M., Rodrigo (1995), 'La sociedad civil y las elecciones en MÈxico: Observadores e interventores,' La voz de los votos: Un an·lisis crÌtico de las elecciones de 1994, Germ·n PÈrez
Fern·ndez del Castillo, Arturo Alvarado M., and Arturo S·nchez GutiÈrrez (Mexico City: Miguel Angel Porra and FLACSO), pp. 87 98.
Nohlen, Dieter, Sonia Picardo, and Daniel Zovatto (eds.) (1998), Tratado de derecho electoral comparado de AmÈrica Latina (Mexico City: Fondo de Cultura EconÛmica).
North, Douglass C. and Barry R. Weingast (1996), 'Constitutions and commitment: the evolution of institutions governing public choice in seventeenth century England,' Empirical Studies in Institutional Change, eds. Lee J. Alston, Thr·inn Eggertsson, and Douglass C. North (Cambridge, UK: Cambridge University Press), pp. 134 65.
O'Donnell, Guillermo (1996), 'Illusions about Consolidation' Journal of Democracy 7/2 (April): 34 51.
Osborne, David and Ted Gaebler (1992), Reinventing Government: How the Entrepreneurial Spirit Is Transforming the Public Sector from Schoolhouse to Statehouse, City Hall to the Pentagon (Reading: Addison-Wesley).
Ostrom, Vincent (1989), The Intellectual Crisis in American Public Administration (Tuscaloosa: University of Alabama Press).
Pastor, Robert (1999a), 'A Brief History of Electoral Commissions,' The Self-Restraining State: Power and Accountability in New Democracies, eds. Andreas Schedler, Larry Diamond, and
Marc F. Plattner (Boulder and London: Lynne Rienner Publishers).
Pastor, Robert A. (1999b), 'The Role of Electoral Administration in Democratic Transitions: Implications for Policy and Research,' Democratization 6/4 (Winter): forthcoming.
Perrow, Charles (1986), Complex Organizations: A Critical Essay (New York: McGraw-Hill), 3rd edition.
Przeworski, Adam (1999), 'On the Design of the State: A Principal-Agent Perspective,' Reforming the State: Managerial Public Administration in Latin America, ed. Luiz Carlos Bresser Pereira (Boulder and London: Lynne Rienner Publishers), pp. 15 39.
Renan, Ernest (1882), 'Qu'est-ce qu' une nation?' Discours et confÈrences (Paris: Calmann-LÈvy).
Rose-Ackerman, Susan (1999), Corruption and Government: Causes, Consequences, and Reform (Cambridge, UK: Cambridge University Press).
Sartori, Giovanni (1976), Parties and party systems: A framework for analysis (Cambridge: Cambridge University Press).
Schedler, Andreas (1997), A Brief History of Electoral Reforms in Mexico (Vienna: Institute for Advanced Studies), unpublished manuscript.
. (1999a), 'Restraining the State: Conflicts and Agents of Accountability,' The Self-Restraining State: Power and Accountability in New Democracies, eds. Andreas Schedler, Larry Diamond, and Marc F. Plattner (Boulder and London: Lynne Rienner Publishers, 1999), pp. 333 350.
. (1999b), 'Las comisiones y la pir·mide: La recentralizaciÛn conflictiva del poder en el IFE,' PolÌtica y Gobierno 6/1 (1999): 187 222.
. (1999c), Democracy by Delegation: The Path-Dependent Logic of Electoral Reform in Mexico, paper prepared for presentation at the 95th Annual Meeting of the American Political Science Association (APSA), Atlanta, 2 5 September 1999.
. (1999d), 'Civil Society and Political Elections: A Culture of Distrust?' The Annals of the American Academy of Political and Social Science 565 (September 1999): 126 41.
Schedler, Andreas and Javier Santiso (1998), 'Democracy and Time: An Invitation,' International Political Science Review 19/1 (January 1998): 5-18.
Scott, James C. (1998), Seeing like a State: How Certain Schemes to Improve the Human Condition Have Failed (New Haven and London: Yale University Press).
Silva-Herzog M·rquez, Jess (1999), El antiguo rÈgimen y la transiciÛn a la democracia (Mexico City: Planeta y JoaquÛn Mortiz).
Thompson, Frank J. and Norma M. Riccucci (1998), 'Reinventing Government,' Annual Review of Political Science, Volume 1, ed. Nelson W. Polsby (Palo Alto: Annual Reviews), pp. 231 57.
Weber, Max (1972), Wirtschaft und Gesellschaft: Grundriþ der verstehenden Soziologie, ed. Johannes Winckelmann (T¸bingen: Mohr), 5th edition.
. (1988), 'Beamtenherrschaft und politisches F¸hrertum,' Gesammelte Aufstze zur Wissenschaftslehre, ed. Johannes Winckelmann (T¸bingen: Mohr), pp. 320 50, 7th edition.
Woldenberg, JosÈ, (1998), 'Votar en el extranjero: Hechos y preguntas antes un tema difÌcil,' EtcÈtera 288 (6 August 1998): 26 30.
. (1999), 'La transiciÛn a la democracia,' Nexos 261 (September): 65 70.
