Abstract
This article asks if low political competition is associated with more restricted public procurement processes. Using unique Swedish municipal data from 2009 to 2015, it demonstrates that when one party dominates local politics, noncompetitive outcomes from public procurement processes are more common. What is most striking is that the risk of receiving only one bid, on what is intended to be an open and competitive tender, considerably increases with long-standing one-party rule. The article contributes to a significant body of work on the detrimental effects of low political competition, and the results are particularly interesting from a comparative perspective because Sweden—an old democracy with a meritocratic bureaucracy, characterized by low levels of corruption and clientelism—is a highly unlikely case in which to find such tendencies.
Keywords
Introduction
Abuse of power is an omnipresent risk.
1
Constitutions, laws, and other regulations are written from a perspective that suggests that while the ruling elite must have enough power to do good, they cannot be left unguarded to do bad. It is, for example, a common theme of the Federalist Papers, where James Madison (Hamilton, Madison, & Jay, 1788/1961, p. 260) notes that . . . power to advance the public happiness involves a discretion which may be misapplied and abused. They will see, therefore, that in all cases where power is to be conferred, the point first to be decided is, whether such a power be necessary to the public good; as the next will be, in case of an affirmative decision, to guard as effectually as possible against a perversion of the power to the public detriment.
In democracies, at least, tendencies for such abuse should diminish with increased elite competition (Schumpeter, 1947; for a related argument, including not only democracies, see Acemoglu & Robinson, 2012). As a matter of fact, much of the political economy literature on rent-seeking starts with the assumption that although politicians are assumed to be selfish, political competition is the principal vehicle that creates policies that benefit large parts of the electorate, rather than just the elite and their immediate followers (Besley, Persson, & Sturm, 2010; Gerring & Thacker, 2004; Montinola & Jackman, 2002; Persson & Tabellini, 2003; Rose-Ackerman, 1978).
With this perspective in mind, the public procurement process is an area in which the political system is put to a critical test. Not only does public procurement today involve huge sums of money, amounting to 12% of worldwide GDP (Organisation for Economic Cooperation and Development [OECD], 2017, p. 10), but these considerations are also key when it comes to political elites’ interaction with business elites, wherein the former have the ability to use their position to benefit their constituency, companies to which they have connections, or themselves. As these temptations are ubiquitous, most governance systems have, accordingly, devised legal and bureaucratic checks against them. In Europe, national and European Union (EU) regulation of the common market aims to create a fair and open marketplace for government contracts. EU Public Procurement Directives are devised to level the playing field for all bidders, connected or not, in the interest of the polity as a whole. The civil service and bureaucratic audit systems within countries are supposed to secure compliance with the rules for open and fair access to public resources. Where such systems work, public procurement processes are consequently of much higher quality (Charron, Dahlström, Fazekas, & Lapuente, 2017).
No system is bulletproof, however; in polities with long-standing incumbents, the weight of the bureaucratic balance may erode. With low political competition, for example, loyalty between otherwise neutral and independent civil servants and representatives of the party in power may develop and pervert the system of control. Moreover, politicians can strengthen their position when in power for a longer time, and build political-business networks that might, in themselves, generate pressure for collusion between the two spheres (Campante, Chor, & Do, 2009). In line with recent research in this field (Coviello & Gagliarducci, 2017; Klašnja, 2015), this article argues that low political competition creates “entrenched parties” (Folke, Hirano, & Snyder, 2011, p. 578) able to control public procurements, and with the power to favor certain bidders, regardless of their formal merits.
Therefore, this article asks whether low political competition—in extreme cases, one-party rule—makes it more likely for incumbents to restrict competition in public procurement processes. To this end, it also investigates the weakening of control mechanisms that have been set up to stifle such tendencies.
To answer this question, we turn to Swedish municipalities, arguably, a least likely case of such political manipulation of the rules discussed above. We find that when one party dominates local politics, procurement processes show noncompetitive outcomes more often, while there is a directly opposite effect on turnover. Most striking is that the risk of obtaining only one bid, on what should be an open tender process, increases with political entrenchment. These results also hold up when we use other operationalizations of political and procurement competition, include a broad set of controls, and employ a wide array of estimation techniques. Moreover, further analysis suggests that entrenched parties are able to exert favoristic control over public procurement due to an amicable relationship with opposition parties, more partisan control over local audits, and lower pressure from the media.
The article contributes to the literature in three different ways. First, our case selection is particularly important. While a number of recent studies that are interested in the micro foundations of the links between low political competition and different types of rent-seeking have considerably advanced the field, they have almost exclusively researched cases already known for widespread corruption, such as Italy, Romania, and Brazil (Coviello & Gagliarducci, 2017; Ferraz & Finan, 2008; Klašnja, 2015). Whether the same dynamics are of significant importance in a low-corruption context was far from certain, and finding the same type of empirical patterns in a country such as Sweden is, therefore, very informative when evaluating the universality of the link between low political competition and the risks associated with the abuse of power. Second, our study contributes to the literature on corruption voting. It theoretically describes, and empirically scrutinizes, entrenched parties’ ability to control the salience of misconduct and thereby hamper accountability (Klašnja, Tucker, Deegan-Krause, 2014). Third, it speaks to the New Public Management (NPM) literature (Hood, 1991), as it explains how the success of externally produced goods and services is dependent on political factors, and thereby answers to a call for more studies that put NPM reforms into political context (O’Toole & Meier, 2015).
Entrenched Parties and Public Procurement
We focus on political elites in democratic states and ask about the conditions under which they are able to influence public procurement, whatever their motives might be. 2 The question under direct scrutiny is whether low levels of political competition make it more likely for incumbent rulers to override procedures that are supposed to guarantee open competition in public procurement.
The rationale behind securing public procurement from outside actors, instead of in-house production, concerns a rather straightforward market mechanism. Generally speaking, the expectation is that competition for public contracts pressures prices downward, and quality upward (Brown, Potoski, & Van Slyke, 2006; Christoffersen, Paldam, & Wurtz, 2007; Donahue & Zeckhauser, 2011). This expectation hinges, however, on the idea of open competition. As OECD (2011, p. 147) notes, public procurement is “. . . vulnerable to waste, fraud and corruption due to its complexity, the size of the financial flows it generates and the close interaction between the public and the private sectors.” These hazards are well known by policymakers. Public procurement processes are, therefore, regulated by the EU (OECD, 2011) and in all OECD member states, including Sweden, the specific case under inspection here. 3
Circumventing the law, for example, by way of restricting competition, invites serious risks. First, it breaks with the purpose of opening up to outside providers of goods and services. It removes the vehicle that is supposed to decrease prices and increase quality, namely, competition. Second, deliberate restriction of competition violates not only the spirit but also the letter of the law. Such restrictions cannot be written into tenders, and this, therefore, restricts the transparency that is vital for accountability. Third, single bidding invites an unhealthy relationship between entrenched municipal parties and certain firms, and, therefore, increases the risk of elite collusion.
Based on the robust and growing literature on the detrimental effects of low political competition, we hypothesize that there is a risk that incumbent politicians try to circumvent open competition and thereby put efficiency and quality-enhancing mechanisms out of play. Starting from the idea that elite competition, and especially interparty competition, drives up governance quality in democracies (Schumpeter, 1947), a large comparative literature studying the effects on governance of electoral rules (Persson, Tabellini, & Trebbi, 2003), party systems (Tavits, 2007), and decentralization (Gerring & Thacker, 2004) has developed over recent decades. In very simplified terms, this line of research investigates, often by making broad cross-country comparisons, if there are negative correlations between institutions enhancing political competition and different forms of rent-seeking (Besley et al., 2010; Chang & Golden, 2007; Kunicová & Rose-Ackerman, 2005; Persson & Tabelini, 1999).
The corrective mechanism is thought to stem from political competition, which, in turn, increases accountability, and, thus, ultimately relies on the assumption that voters punish incumbents for the misuse of power. This assumption is, however, only modestly supported by empirical studies from the related field of corruption voting. While it seems reasonably clear that voters, to some extent, cast their votes based on perceptions and experiences of corruption (Xezonakis, Kosmidis, & Dahlberg, 2016), corrupt politicians are surprisingly often reelected (Chang, Golden, & Hill, 2010). Recent papers have suggested that the relatively low electoral cost for corrupt politicians might be caused by the fact that opposition to corruption voting crucially depends on its political salience (Ferraz & Finan, 2008; Klašnja et al., 2014).
For accountability to be efficient, the misconduct must be salient. However, without meaningful political competition, the long tenures that often follow put incumbents in a position where they, to some extent, can control the salience of potential misconduct. We use the terminology from Folke et al. (2011) and refer to such parties as entrenched parties. In their recent paper, Coviello and Gagliarducci (2017) demonstrate that politicians’ length of tenure in office, indeed, affects the outcomes of public procurement processes in Italy. Their findings suggest that longer time in office brings higher risk of corruption, showing that extended mayoral tenures lead to more local winners, more expensive contracts, and lower quality procurement processes in general. We argue that this is most likely an effect of long-tenured incumbents’ ability to disarm internal, as well as external, monitoring functions. If these functions worked as intended, they would, when sounding the alarm, increase the salience of the issue and raise the electoral cost of such manipulations.
Controlling Salience
The key conclusion from existing research is that for accountability to work as intended, issues of power abuse must be salient to voters. Therefore, it is important to understand the formal and informal monitoring mechanisms, monitoring agents’ incentives, and the entrenched parties’ ability to control such monitoring agents. We see five mechanisms that are potentially affected by political entrenchment and one-party dominance.
First, we should consider the possibility that the competing party, or parties, might be less likely to sound the alarm when they are electorally weak. In his seminal book Party Government, E. E. Schattschneider (1942, p. 183) describes why partisanship is not by default an effective control of a powerful local party boss: Professional politicians as a class develop a remarkable solidarity when their privileges are attacked by the public. The bosses of the rival parties in the locality can often lend each other a helping hand. The tendency of the bosses to get together is enormously strengthened in regions where the disparity in the strength of parties is great. If one party is overwhelmingly strong and the other party is correspondingly weak, the temptation of the stronger party to annex the weaker party is very great indeed.
Second, the bureaucracy routinely faces external formal monitoring agents, such as local and national auditors. Using within-country variation in Brazil, Melo Pereira and Figueiredo (2009) show that political competition actually correlates with the effectiveness of audit institutions. Moreover, they demonstrate the importance of the institutional setting—a less volatile party system and more programmatic linkage strategies make the audit institutions more effective, too. In brief, while audits and other similar functions are not expected to exert a strong curb on the abuse of public procurement in general, in highly institutionalized parts of the world, such as Sweden, they may turn out to be effective controls. An entrenched party is, however, in a better position to disarm the auditor, particularly if the auditor is in any way dependent on the incumbent, for example, for their appointment. Entrenched parties are incentivized to use their powers of appointment more aggressively exactly because they are in a situation with low risk of turnover. They do not have to pay the price of other parties doing the same thing following the next turn (Grzymała-Busse, 2007).
Third, media coverage of corruption increases its salience and can, therefore, affect the level of corruption voting (Ferraz & Finan, 2008; Klašnja et al., 2014). This can plausibly be extended to abuse of public procurement for reasons other than corruption. Investigative journalism might very well expose shady procurement deals, and extensive media coverage can ensure that voters will be informed of potential political misbehavior (Svaleryd & Vlachos, 2009). Gordon (2011) documents how vendors in districts crucial for the electoral success of the Republican Party initially won unusually large contracts, and then, after the Washington Post had described this as an effort at manipulation, how this effect disappeared. This check nevertheless hinges on politicians being susceptible to media critique, which is not a given in the context of low political competition (Besley & Prat, 2006). However, with a larger winning margin, or outstanding political craft and experience of power, comes the opportunity, from time to time, to handle critique from outside actors, such as the media, and, thus, accept some electoral cost. Entrenched parties are likely, therefore, to be often in a position where they can ignore the risk of being scrutinized by media.
Fourth, in all contemporary states, the bureaucracy provides an informal and internal check on the incumbent. Such checks introduce a relatively efficient control mechanism on politicians (Dahlström & Lapuente, 2017; Miller & Whitford, 2016). Charron et al. (2017) describe how procurements in Spanish municipalities are sometimes manipulated by politicians to extract rents, and explain how this hinges on politicians’ ability to control bureaucrats using their power over appointments and salaries, for example. Politicians seem not only to appreciate such power, but also to know how to use it. Ting, Snyder, Hirano, and Folke (2012) show that incumbents maintain a desire to keep a patronage bureaucracy, as long as they expect to continue their winning streak, while Folke et al. (2011) demonstrate that political control over the bureaucracy can be transferred into increased votes. Even in civil service systems, such as that in Sweden, long-standing incumbents might put a strain on the neutrality of bureaucrats. For example, bureaucrats normally have the advantage of being inside the system for the long run. However, with low political competition, the longevity advantage attenuates and bureaucrats become more dependent on the politicians of a particular party. It is not unlikely that additional partisan loyalty follows, as entrenched parties are also able to bias applicant selection so that it is mostly those who sympathize with the party in power who get appointed (for a classic, and partly similar, way of reasoning about salary levels in the bureaucracy and corruption, see Becker & Stigler, 1974). Moreover, under normal circumstances, the bureaucrat is the expert, but with long tenures, incumbent politicians and parties are likely to become knowledgeable, too, which makes them less dependent on the bureaucrats. The competence of bureaucrats has another important effect. In a study of bureaucratic turnover in India, Iyer and Mani (2012) describe how highly competent bureaucrats are less susceptible to political pressures, which indicates that with competence comes the ability to resist pressure from entrenched parties. For, as noted by Schattschneider (1942, p. 176), “the boss lives by bad administration.”
Fifth, and finally, entrenchment implies stronger networks with the outside community, including tighter bonds with contractors (Campante et al., 2009; Coviello & Gagliarducci, 2017). Such networks are likely to increase the pressure on politicians to circumvent open competition, as well as their ability to do so. In such cases, well-connected contractors are likely to find themselves in a position to call in favors from time to time. At the same time, politicians who are satisfied with what such contractors have delivered previously are probably tempted to overlook some irregularities to give the contract to someone they know and trust.
Taken together, we expect a negative correlation between limited political competition and noncompetitive outcomes in public procurement. Furthermore, we expect several intermediary mechanisms to be present, namely, a co-opted opposition, a tighter political control over audits, politicians less susceptible to media critique, lower quality bureaucracy, and tighter networks between politics and business, with more local winners of contracts.
Empirical Strategy
In the remainder of the article, we will estimate the relationship between political entrenchment and noncompetitive procurement outcomes in Swedish municipalities. Such an empirical strategy, that is, studying local-level politics within a single polity, is methodologically beneficial for at least two reasons: First, restricting the scope to a single country drastically diminishes the risk of omitted variable bias (Alt & Lassen, 2003); as Sweden is a unitary state of moderate size, this risk is likely to decrease even further. Second, the subnational level of analysis is also an effective antidote to ecological fallacies, derived from what Snyder (2001) refers to as “whole nation bias,” wherein considerable subnational variation is made invisible in the face of national-level averages.
Furthermore, as discussed in the Introduction, Sweden is an almost ideal least likely case for the present purpose, and any affirmative conclusions will, therefore, expand the universe for which the theoretical expectations on political entrenchment and noncompetitive outcomes presented above are applicable. Our underlying expectation is that any positive results indicating problems with institutional quality here are likely to be more limited than in contexts where politicians are generally freer to engage in illicit, clientelistic, or corrupt activities.
The Case of Sweden
Sweden is a medium-sized (10 million inhabitants) European state. Despite its unitary structure, Sweden’s 290 municipalities are unusually autonomous and legally independent entities, in charge of most public services—such as primary and secondary education, as well as child-, social, and elderly care—resulting in the majority of the country’s public servants being employed in the municipal sector (Statistics Sweden, 2014). As with the national level, its proportional electoral system means that local politics tend to involve the representation of a relatively large number of parties, usually the same seven or eight that are represented in the national parliament, along with an increasing, but still limited, presence of local parties. Despite the long-standing dominance of the Social Democratic Party at the national level, local politics have traditionally contained much more ideological variation, with a considerable share of municipalities ruled by center-right or rainbow coalitions (Erlingsson & Wänström, 2015).
As mentioned above, studying political and institutional malpractice and dysfunctionality in a Swedish local context is particularly illuminating, considering the country’s high level of institutional quality. Despite a recent growing trend of political appointees, the nature of Swedish public administration is still decidedly meritocratic, and this remains the case also at the local level (Dahlström, Folke, & Rickne, 2014; Garsten, Rothstein, & Svallfors, 2015). Each municipality has its own audit committee, responsible for overseeing the operational effectiveness of municipal operations (Swedish Association of Local Authorities and Regions [SALAR], 2014). In contrast to the body of public servants, these committees are politically appointed, but the chair is usually a representative of the opposition, and the actual audit reports are almost always written by outside experts. Furthermore, public procurement is regulated through Swedish law, which is largely based on EU Directive for this area. The explicit aim of Swedish regulation is to secure public procurement processes that seek out and take advantage of competition to get best value for money. The fundamental principles for public procurement in Sweden are “the principle of non-discrimination,” “the principle of equal treatment,” “the principle of transparency,” “the principle of proportionality,” and “the principle of mutual recognition” (the Swedish Public Procurement Act, 1:9). The Swedish Competition Authority (SCA) is the designated national agency for monitoring and ensuring that competitive public procurement is achieved according to the principles previously mentioned.
Furthermore, focusing locally is likely the most appropriate level of investigation for our purpose, because pork-barrel politics has been found to have a local flavor in many democracies. In the United States, for example, partisan control over federal expenditures systematically affects which districts receive funds (Kriner & Reeves, 2015). Even in Sweden, there are indications of local spending for partisan purposes (Dahlberg & Johansson, 2002). Other studies have shown that family ties to local politicians in Denmark—another country with high marks for its institutions—increase firm profitability, especially in industries relying on public demand (Amore & Bennedsen, 2013), that children of local politicians in Sweden have higher average earnings, although there might be legitimate reasons for these so-called “dynastic political rents” (Folke, Persson, & Rickne, 2017), and that low political competition is associated with higher legal political rents in local government in Sweden (Svaleryd & Vlachos, 2009; but see Bergh, Erlingsson, Sjölin, & Öhrvall, 2013). Therefore, it is reasonable to expect that biased spending will be traceable primarily at the local level.
Sweden and its municipalities are also an ideal setting in which to study public procurement processes: First, according to the SCA, public works, goods and services worth about 625 billion Swedish kronor (~US$71 billion) were bought by public entities such as municipalities, agencies, and publicly owned companies in 2012. This adds up to nearly a fifth of total Swedish GDP in 2012 (Swedish Competition Authority, 2015, p. 14), and puts Sweden in the upper quartile in comparison with other OECD countries, where the average around that time was about 13% (OECD, 2011, p. 149). Public procurement is, thus, a large and important part of public spending in Sweden, as it is in the rest of the OECD. Furthermore, reflecting their importance as the principal public service providers in the country, 70% of all procurements are made by municipalities and their companies (SCA, 2015, p. 29). In 2016, there was a total of 18,336 open public procurement tenders in Sweden, a large majority of which were announced by municipalities and municipal companies (SCA, 2018, p. 14). There is, thus, a considerable experience of handling public procurement at the local level. Finally, the object of analysis considered herein is not only politically and administratively relevant, but also decidedly local in nature. Even the lowest official EU classification of subnational units (Nomenclature of Territorial Units for Statistics [NUTS] 3-level, corresponding to the county level) contains an average of 14 Swedish municipalities, and although the 13 largest municipalities have a population exceeding 100,000, the median size is a modest 15,235 inhabitants.
Within this institutional framework, an entrenched party has some leeway to influence procurements. As mentioned above, public procurement is regulated by law and monitored locally by auditors and nationally by the SCA. While open competition is clearly stipulated in the law, there are ways to obey the letter but not the spirit of the law. The buyer could, for example, write specifications into solicitations, or invoke extreme urgency exceptions to create noncompetitive outcomes. Municipal politicians cannot, however, circumvent local procurement officers. Generally speaking, local public procurement involves both politicians, on executive boards and/or the council, and bureaucrats. Biased procurement, therefore, has to be accepted by both parties, which under normal circumstances work against deliberate restriction of competition. But with entrenched parties, the preferences of the two groups—as well as the auditors—are more likely to align.
Indeed, there are rather straightforward ways for an entrenched party to influence both auditors and bureaucrats. Parties nominate and elect auditors in the municipal council, wherein an entrenched party will tend to hold a strong position. Moreover, the council is the auditors’ principal, which among other things means that it decides on the auditors’ budget and is the body that receives the auditors’ reports (SALAR, 2014). And while the bureaucracy in Sweden is decidedly meritocratic in principle, the protection for meritocratic recruitment and promotion is not as strong for municipal as for state employees. 4 Furthermore, the highest official in the municipality—the municipal director (Kommundirektören)—is appointed directly by the executive board, led by the mayor and her or his party (the Swedish Municipal Act, 2017:725). In addition, the highest administrative official for each sub-board, for example, on social affairs, is appointed by that board, which again would be dominated by the entrenched party. 5
Still, although it is well known from the literature on politicization that appointees’ loyalties trickle down into the organization (Lewis, 2008), the bias created by a long-standing incumbent party is probably more important than the ability to directly appoint top officials and auditors. After decades with the same entrenched party (in our case, about 40 years), it is not unlikely that those bureaucrats that stay on internalize the incentives of the entrenched party, while those who will not do so, exit.
Swedish Public Procurement Data
In Sweden, the only publicly available source of public procurement data is the EU-wide Tenders Electronic Daily, 6 which only reports large-value contracts regulated by the EU’s Public Procurement Directives. Given that this article’s main interest is municipal public procurement, we, therefore, also collected data on smaller contracts whose value falls below EU reporting thresholds, but which are above national thresholds. 7 As there is no public database of these smaller contracts, we obtained the data directly from a private data provider (Visma Opic), which in effect implements the relevant transparency provisions of the Swedish Public Procurement Act. According to the law, tenders below the EU threshold are either published by Visma Opic directly or another local tendering portal from which Visma Opic collects the information and enters it into a consolidated database. As there is no publication requirement for direct awards below the national threshold, the database only contains such low-value tenders if they were voluntarily published. Due to the fragmented and unregulated public procurement publication process, data formats and contents are very diverse and consolidation into a unique database is problematic; hence, Visma Opic manually collects and enters data where necessary and also searches for missing information where possible. The database covers the key characteristics of the tendering and contract award phases such as date of publication, contract value, name of the winning bidder, name of the buyers, or the product category of the purchase.
In spite of the laborious data collection effort of Visma Opic, data quality is an issue facing our analysis, leading us to use only those variables that are considered to be reliable enough. In total, there are 135,007 unique tenders in the database between 2009 and 2015; roughly 70% belonging to the national regime, and 30% to the EU regime (Tenders Electronic Daily). In addition, to increase the precision in our measures, we restrict the sample in three other important ways. First, we only use contracts awarded by local bodies, that is, municipalities and municipal enterprises, which shrinks the sample to 89,951 unique tenders. Second, we remove noncompetitive markets, defined in line with prior research (Charron et al., 2017). Specifically, we excluded all those markets—defined by a combination of geography (NUTS-1 level) 8 and product group (3-digit CPV) 9 —which have less than five unique bidders winning contracts in the whole of the 2009-2015 period. This equals 521 tenders, or 0.58% of the sample, leading to a tender count of 89,430. Third, we removed those tenders that were cancelled or incomplete (i.e., still pending final contract award decision); this removed 6,979 and 589 tenders, respectively. Combined, these conditions lead to a final sample used in the analysis numbering 81,931 tenders. 10
Risks of Single Bidding in Public Procurement
Our dependent variable is a direct measure of noncompetitive outcomes in public procurement. We operationalize our dependent variable as single bidding, that is, only one bid being submitted in a tendering process in an otherwise competitive market, as this represents the simplest indication of restricted competition. If a municipality, during a given year, has had multiple bidders for all of its tenders, it will receive a score of 0. If all tenders received only one bid, it will receive a score of 100. Hence, the measure we use in the municipality database is the percentage of single-bidder contracts out of all the contracts awarded by a municipality in a given year. We argue that a high percentage of single bidding in a municipality signals repeated occurrence of noncompetitive tendering, which, at the very least, makes deliberate manipulation more likely compared with situations in which there is competitive tendering.
Our interpretation of single bidding as a risky outcome crucially depends on adequately identifying competitive markets. As discussed above, competitive markets are those with at least five unique suppliers. If a market has at least five active suppliers, it is quite likely that two of them will show up as noncorruptly formulated tenders. The fact that markets with less than five suppliers make up only 0.5% of the sample underlines that the overwhelming majority of public procurement in Sweden is done in competitive markets for products that can be supplied by a variety of companies (an alternative competitive market definition, taking 10 suppliers as a cut-off point, removes merely 1.4% of the sample, further strengthening our claim for competitive markets being the predominant market type in Sweden). In addition, as our market definition already incorporates geography (NUTS-1 regions), we expect no bias from less competitive markets in more remote or more sparsely populated regions (recall, our measure does not consider the average number of bidders, only single vs. multiple bidders).
Although single bidding in competitive markets may result from a range of reasons, including incompetence, comparative research has primarily used it as a corruption risk indicator (Charron et al., 2017; Coviello & Gagliarducci, 2017; Klašnja, 2015). While this article is not designed to study the motivation for entrenched parties to restrict competition, it is important for us to show that single bidding might have detrimental effects. For instance, in the Swedish context, we find that non-Swedish firms winning public procurement contracts are close to 10 percentage points more likely to be single bidders if they are registered in a tax haven such as Panama than if they are registered in non-tax haven countries such as Germany (Figure 1). This suggests that illicit proceeds that may have been earned through single bidding contracts are then often channeled through secret jurisdictions, to hide money flows (Shaxson & Christensen, 2013). Furthermore, and similar to findings in other countries (Fazekas & Kocsis, 2017), single bidding is associated with more expensive contracts when compared with initial cost estimates produced by independent experts (there is 12.1% of single bidding in contracts below initial estimates, i.e., a discounted final contract value, as contrasted with 14.3% single bidding in contracts with equal or above initial estimates, i.e., no or negative discount).
Comparing incidences of single bidding (%) among foreign suppliers according to the country of incorporation, Sweden, 2009-2015.
Importantly, reporting quality—due to corruption or other reasons—may bias our measurement of noncompetitive outcomes. Hence, we test whether this is correlated with the dependent variable. Information can be concealed in two ways: First, municipalities may hide contracts altogether by splitting them up into smaller contracts, each falling under the national reporting threshold (e.g., about SEK500,000 [SEK = Swedish Krona] for services); second, they may omit important bits of information from public notices, without which assessing public procurement performance is difficult (e.g., name of the winner and contract value). We measure contract concealment by calculating the proportion of advertised public procurement contract value in our database to total municipal spending on public procurement from local budget statistics, 11 and measure omitted information by counting the number of data points missing from seven mandatory items (buyer address, buyer post code, buyer settlement, contract value, supplier name, number of bids received, and contract award date). Quite reassuringly, at the level of municipalities, neither of these indicators is significantly correlated with the single bidding ratio (linear correlation coefficients are −0.072 and −0.002, respectively).
Independent Variables
Our primary measure of political entrenchment is one-party rule, a dummy variable indicating whether the same party has held the highest political post (in Swedish, kommunstyrelsens ordförande, the chair of the executive board; henceforth, “mayor”) during the entire era of modern Swedish municipalities, which began with a massive wave of mergers in the early 1970s. 12 Although coalition rule is a very common occurrence, and Swedish mayors are indirectly elected by the local assembly, Karlsson and Gilljam (2016, p. 704) note that the mayor is the “undisputed leader of a Swedish municipality,” and local politicians consider this post to hold as much power as the municipal executive board (analogous to its government) at large (Erlingsson & Öhrvall, 2017).
Although most municipalities have experienced at least one turnover in power, by as late as 2015, more than a fifth of Swedish municipalities had not. It should be noted that, due to Sweden’s proportional electoral system, staying in power for an extended period of time requires a large measure of political skill and maneuvering, appeasing both the electorate and other parties in the municipality. For example, if the voters of a ruling party on the right, such as the Conservatives, are not satisfied with the party’s rule or policies, there are three ideologically close alternatives available (the Center Party, Liberals, or Christian Democrats).
Second, we employ an alternative operationalization of political entrenchment through stability, an ordinal scale-variable indicating whether the incumbent party is new for the given term, reelected once, or reelected twice or more. 13 Although one-party rule is likely the best representation of an entrenched political landscape, this additional measure provides a more nuanced and contrasting perspective of the earlier stages of ruling party entrenchment (length of mayoral tenure has been shown to influence procurement performance in Italy; see, for example, Coviello & Gagliarducci, 2017). Even if strongly incentivized to embark on bending the system to its own benefit, new ruling parties are unlikely to achieve this in the short run, as permeation of the political and administrative structure, for example, by strategical staffing, is bound to take time, especially in the Swedish context where meritocratic recruitment largely overshadows any type of spoils system (Dahlström et al., 2014).
Estimation Strategy
To predict single bidding as a function of political entrenchment, we combine cross-sectional and panel regressions, and contract-level matching estimators. First, as the within-municipality variation in one-party rule is too small for meaningful time-series analysis, 14 we start with models exploiting cross-municipal variation, focusing on the one full-term period contained in our sample (2011-2014). 15 To isolate our focal relationship, this approach calls for a fairly comprehensive set of controls. The estimations below will also consider the size of population, 16 and the (logged) land area of each municipality. More populous municipalities are likely to have more competitive markets, and, thus, be prone to receiving a higher number of procurement offers, while working in geographically larger municipalities involves larger transaction costs that may dissuade companies from placing offers. Furthermore, we include median income as wealthier municipalities, it is quite reasonable to assume, will tend to attract more companies. We also include the identity of the ruling party, in part due to the fact that type of political leadership is likely to capture a number of otherwise immeasurable socioeconomic factors; as an example, one will find stark differences between municipalities that have only been ruled by the Social Democrats, which tend to be small industrial towns, and those that have been ruled continuously by the Conservatives, which are generally wealthy metropolitan suburbs. Furthermore, although we have no prior expectations regarding the matter, one cannot exclude the possibility that different parties operate in different ways regarding the political establishment’s views and approaches to public procurement (a notion tested in the Robustness section). Finally, to capture remaining unobserved variation, we include county (N = 21) fixed effects. 17
Second, the fact that stability contains ample within-municipality variation over time 18 allows us to move to a panel format with this variable, using municipal fixed effects (FE) on a municipal-year-level dataset for 2009-2015. Perhaps the main advantage of this model is that it accounts for unobserved time-invariant municipal characteristics, which are likely to plague our goal of approximating causal identification (e.g., municipal size simultaneously determining procurement competition, hence, single bidding; and political competition, hence, one-party rule). As most controls employed in the one-party rule-based models vary little or not at all over time, only median income and party FEs remain in the FEs models, while year FEs are introduced. Because single-bidding ratio is only weakly autocorrelated, we mainly rely on static estimations, but also present dynamic panel estimations, including a lagged dependent variable (LDV). 19 This specification is, however, problematic on two grounds: First, it is well-known that introducing LDVs in FE models gives rise to Nickell bias, resulting in inconsistent estimates, especially in shorter panels such as the present one. Second, the LDV also removes 15% of all observations (i.e., all cases in 2009). As a solution for the former issue (but—notably—not the latter), we employ system- and difference–generalized method of moments (GMM). A further benefit of the GMM framework is that it allows for considering the possibility of endogeneity even further; herein, we treat stability as predetermined.
Third, we also carry out a contract-level propensity score-matching analysis, interpreting one-party rule as the control condition and multiparty rule as the treatment. 20 To fully reflect the degrees of treatment captured by the stability variable (i.e., reelection only once or reelection twice or more), we also include matching with treatment conditions of one, as well as two or more, reelections. The added value of the contract-level matching, on top of the cross-sectional and panel data analyses, is that it takes into account contract-level variance such as contract value distribution, which the organization-level analyses can only reflect imperfectly. In addition, contract-level matching is perfectly suited to deliver tight comparisons of the most similar contracts awarded by treatment and control municipalities, further addressing the potential biases of comparing dissimilar contracts.
In summation, while we lack robust causal identification using random assignment, our diverse set of models aims to address the major sources of observed and unobserved confounders that we can think of, across municipalities, within municipalities, and at contract level. While none of the approaches on their own warrants causal interpretation, taken together, we posit that they suggest that there is a causal link, rather than mere correlation.
Results
To recapitulate, our overarching hypothesis is that politically entrenched municipalities will have less well-functioning public procurement processes, resulting in higher single-bidding ratios, while new ruling parties will be associated with lower single-bidding ratios. Below, we present the results of the main tests of this link, followed by a series of robustness tests, and an investigation into the proposed mechanisms through which any such relationship is likely to flow.
Main Results
First, a simple bivariate look offers initial support to the entrenchment hypothesis. As evident from Figure 2, one-party-rule municipalities are, indeed, associated with a higher propensity for single bidding. Compared with equivalents that have experienced turnover, the single-bidding ratio in such municipalities is 3.3 points higher, an increase of 29%. Contrastingly, municipalities with a new ruling party score 1.9 points (17%) lower in single bidding than municipalities with once-reelected parties, and 2.4 points (22%) lower than incumbents with even longer tenures.
Political entrenchment and single-bidding ratio.
For one-party rule, this relationship is only marginally weakened with the introduction of control variables, which work in the expected direction (sparsely populated, poor, and geographically large municipalities significantly predict higher single bidding). The fully controlled estimation (Table 1, column 7) has one-party rule associated with 3.1-point (p < .01) higher single-bidding ratio. The notion that long-term political entrenchment is related to restricted competition is thereby supported.
Single-Bidding Ratio and One-Party Rule.
Dependent variable: Single-bidding ratio. Data averaged for the 2011-2014 term period. Robust standard errors in parentheses. FEs = fixed effects.
p < .1. **p < .05. ***p < .01.
Next, we shift focus to our secondary operationalization of political entrenchment, stability. The results are in line with the bivariate findings, as well as the results from the analysis on one-party rule above. In brief, we observe a consistent positive association between being reelected and higher single-bidding ratios (Table 2). The introduction of controls does little to shake this observation; coefficient sizes remain highly stable, although the significance levels for both reelection categories (once, and twice or more) oscillate between the 95% and 90% level of significance. The fully controlled static estimation (column 5) reveals that, concordant with expectations, the coefficient for reelected once is smaller than for reelected twice or more, but while the former lands at the stronger side of the 95% threshold of significance (p = .046), the latter ends up on the weaker side (p = .054). When introducing a lagged dependent variable (LDV; column 6), reelected once remains significant (p = .052), while reelected twice or more loses significance (p = .244). Considering that this estimation is associated with Nickell bias, columns 7 and 8 present the GMM results, more appropriate to handle the LDV. Here, the coefficient sizes for both reelected once and reelected twice or more increase compared with all FE estimations, although significance for the latter remains at the 90% level in the system-GMM estimation (p = .062). The insignificant coefficient for the LDV in these specifications does, however, lead us to conclude that the static FE estimator (column 5) is our preferred specification. Despite a modest measure of heterogeneity stemming from the choice of modeling technique, we can detect a consistent and positive association between stability and increased single bidding. While the greater part of this link appears rather immediately with the first reelection, the generally larger coefficient sizes for reelected twice or more, along with the preceding findings regarding one-party rule, hint at a process that compounds over time.
Single-Bidding Ratio and Stability.
Dependent variable: Single-bidding ratio. Robust standard errors in parentheses. Stability reference category: new ruling party. Fixed effects: standard errors, clustered at the municipal level, in parentheses. GMM models use the two-step estimator with Windmeijer correction; stability and LDV treated as predetermined and instrumented GMM-style (lag depth 1-4). Other covariates treated as exogenous and instrumented IV-style. LDV = lagged dependent variable; FEs = fixed effects; Sys-GMM = System Generalized Method of Moments; Diff-GMM = Difference Generalized Method of Moments.
p < .1. **p < .05. ***p < .01.
In summation, while these approaches individually lack the makings of hard causal inference, the combination of the extensive controls in the cross-sectional framework with the FEs and GMM estimation should, at the very least, alleviate the vast majority of potential objections about spuriosity, endogeneity, and omitted variable bias.
Robustness
To further ascertain the robustness of our findings, a number of alterations to the original estimation strategies were devised. In turn, we made adjustments to our modeling specifications, our focal variables, and explored heterogeneities in the results. 21
First, as already mentioned in the discussion on estimation strategy, and to ascertain that our results are not driven by contract-level variance masked by aggregating to the municipal-period level (e.g., sectoral structure of spending), we shifted the unit of observation to the contract level. Such a high degree of granularity allows for controlling for contract characteristics that are likely to influence bidder numbers. In addition, employing propensity score matching also represents an alternative specification to the regression techniques used in the main analysis (see Tables A3-A8; Figures A2-A4 in the online appendix). In addition, we estimated one-party rule with all available data using the between-estimator, as well as pooled ordinary least squares (OLS) with errors clustered at the municipal level (see Table A9 in the online appendix). Both sets of analyses yield results in line with the main results presented above. We also estimate stability in the cross-sectional context (see Table A10 in the online appendix). This specification allows us to leverage the data from an extra term-period back in time, increasing the number of categories to include elected three times or more. Furthermore, shedding the focus on within-municipality allows us to add one-party rule as a fifth category. The results, which are most comparable to the cross-sectional estimations using one-party rule, show that municipalities with one-party rule are most prone to single bidding. Furthermore, new ruling parties are significantly less likely to display single bidding than all categories of reelected incumbents. Interestingly, the coefficient for the reelected three times or more category is relatively small, hinting at heterogeneity within this group that we, unfortunately, cannot explore further using the available data.
Second, we respecified both the main independent and dependent variables. For the former, we substituted our original measures of political competition for one capturing the vote share of the ruling party (Table A11 in the online appendix). While, as we argued above, entrenchment is most likely mainly a result of temporal factors, a stronger mandate from voters is likely to make incumbents more comfortable in influencing bureaucratic affairs. The results are not particularly strong, but they do consistently display a positive relationship between vote shares and single bidding; for our preferred (fully controlled static FE) specification, a 1-percentage point increase in mayoral party vote share is associated with a 0.16 increase in single-bidding ratio (p = .1). For the dependent variable, we relaxed the assumption of linearity by predicting the likelihood of a municipality-year having no single bids at all (Table A12 in the online appendix). Furthermore, we substituted the dependent variable to measure the (discounted) average number of bids, an approach acknowledging the possibility that a municipality consistently receiving only two bidders, but never only one, may plausibly be considered to have lower competition than a municipality receiving a large number of bidders for almost all tenders, but occasionally only receiving a single bid (see Table A13 in the online appendix). 22 These altered estimations garner results in the same direction as the analysis above, bolstering our main finding of a negative relationship between political entrenchment and competition in public procurement, especially in the long term.
Finally, we accounted for the possibility of unobserved heterogeneity in the results. Two factors stand out as particularly relevant for our case: the size and political composition of a given municipality. As Rose-Ackerman (1999, p. 101) notes, collusion is more easily maintained in smaller settings, where there tends to be a more limited number of actors involved in such activity. Furthermore, regardless of whether the goal is corruption or simply pragmatism, smaller municipalities may also be a context more conducive to a higher degree of the informalism inherent in the political-bureaucratic-business nexus, in turn, leading to the political sphere having a comparatively larger sway over bureaucratic matters such as public procurement. Thus, there is reason to believe that simply including population as a control—as we did in the original estimations of one-party rule and the system-GMM estimation with stability—fails to sufficiently account for its influence. Therefore, we reran the stability-estimations with the sample split down the median in terms of population size (15,190.5; see Table A14 in the online appendix). 23 Indeed, the results strongly indicate that political entrenchment is more cogent in smaller municipalities, whose negative coefficient in this setting dramatically increases to nearly twice its original size, while it is rendered null in large municipalities. The same trend is observable for one-party rule; its interaction with the log of population size reveals a positive and strongly significant relationship in small municipalities and an (insignificantly) negative link in large ones. The break-even point appears around the fourth quartile (33,760.5 inhabitants; see Table A15; Figure A5 in the online appendix). Furthermore, the main results could plausibly be conditioned by the identity of the entrenched party—a question of particular relevance in Sweden due to the relative historical dominance of the Social Democratic Party. Interparty differences are, however, not dramatic; both when measured as one-party rule and stability, entrenched parties consistently have higher single-bidding ratios, regardless of which party is in charge (see Table A16 and A17; Figure A6 and A7 in the online appendix).
Mechanisms
Next, we turn to the specific mechanisms accounting for why and how political entrenchment may decrease competition in public procurement. In the theoretical discussion, we identified five potential intermediary mechanisms through which this relationship could operate. First, the opposition in a politically entrenched landscape will tend to be more lenient toward the ruling bloc. Second, political entrenchment may suppress external monitoring functions such as audit committees within a polity. Third, entrenched rulers are similarly less susceptible to media critique. Fourth, entrenched politicians may also silence potential critique internally by ensuring that their own competency is not rivaled by local bureaucrats. Fifth, networks between politicians and local business may have had a longer time to develop when one party has ruled for an extended period of time.
Using cross-sectional averages for 2011-2014, we first analyze the respective associations between one-party rule and indicators capturing the respective suggested mechanisms. We then estimate the association between these mediating variables and single-bidding ratio. First, the oppositional role of the political opposition is captured through a survey item describing the relations between the political majority and opposition (data from a 2012-2013 survey of local politicians, Gilljam & Karlsson, 2013; question framed “The relations between majority and opposition in my municipality are good”). Second, external monitoring is operationalized as whether the chair of the municipal audit committee comes from the ruling majority (data from Statistics Sweden, 2017a, complemented by data collection by the authors). Third, media sensitivity is captured using a survey question for politicians in the ruling majority on the extent to which election promises come to fruition through local media pressure (again using data from Gilljam & Karlsson, 2013). Fourth, for the internal control mechanism, we estimate human capital in the bureaucracy—measured as the share of municipal employees with postsecondary education (data from Kolada, n.d.-a). Finally, we estimate local networks as the share of local winners in municipal public procurement (using our own data, provided by Visma Opic). Each relationship is tested bivariately and with the full set of control variables (i.e., population, area, median income, ruling party ID, and regional FEs). 24
The results, displayed in Figure 3, tell a mixed but predictable story. All of the relationships are in the expected direction, although not all are significant in both steps. Once the battery of controls is considered, majority-opposition relations are significantly better in one-party-rule municipalities, while also predicting (at the 90% level of confidence) higher levels of single bidding (panel A). Second, although one-party rule strongly predicts majority-chaired audit committees, this is, in turn, only insignificantly related to more single bidding (panel B). Media influence (panel C) is significantly lower in one-party rule municipalities and predicts lower levels of single bidding itself. While bureaucratic human capital, as expected, predicts lower single bidding (panel D) and the bivariate association between one-party rule and bureaucratic human capital is strongly negative, this does not hold for the inclusion of controls. Finally, local winner ratio (panel E) is positively yet insignificantly related to one-party rule and single bidding.
Mechanisms.
The fact that all relationships go in the expected direction, but display varying strength and sensitivity to account for structural factors, points to a multicausal story, in which certain plots are more convincing than others. Fierce political opposition, more rare where entrenchment has set in, appears to be a moderate boost to procurement competition. While it seems like the audit function is, indeed, weaker in one-party municipalities, its potency for ascertaining competitive procurement is itself only marginal. Conversely, the media appear to be a more important external check. Similarly, highly skilled bureaucrats appear to be able to use their “alarm” function to a higher degree in turnover municipalities, with better procurement as a result, although contextual factors seem to be playing a large role here. Finally, we find only very weak evidence of local networks disproportionally influencing the procurement process.
Conclusion
We have suggested that the tendency for ruling politicians to manipulate public procurement processes at the expense of the general public is stronger when political competition is low. Employing a unique dataset, including information about local political competition going back decades in time, public procurement contracts between 2009 and 2015, and a large set of other relevant variables in Swedish municipalities, our results demonstrate that when political competition is low—and especially when one party dominates the political landscape for a long time—public procurement processes, indeed, show signs of manipulation, as they are less competitive. Although marginally weaker, we also observe that this trend dissipates when a new ruling party assumes power. These results are robust to a large number of alterations of estimation strategy, and compounded in smaller municipalities, where these problems are already disproportionately severe. Moreover, we propose that when one party dominates the political scene, the control mechanisms within the political system—external as well as internal—will tend to erode, thus, facilitating the influencing of procurement.
Considering its strong history of programmatic parties and low levels of corruption and clientelism, Sweden, in all likelihood, approximates a true least likely case. Thus, our study stands in sharp contrast to recent papers in the same vein, which have tended to focus on young democracies and/or institutionally weak settings (Coviello & Gagliarducci, 2017; Klašnja, 2015).
Furthermore, the results corroborate a fundamental expectation in the political economy literature, showing how low political competition goes together with bad government (Gerring & Thacker, 2004; Montinola & Jackman, 2002; Persson & Tabellini, 2003; Rose-Ackerman, 1978). This is an important contribution in itself; as such an association has previously mostly been studied on the aggregate level, and between countries. Our study also advances knowledge of how entrenched parties can take advantage of the bureaucratic apparatus (Folke et al., 2011), findings that could shed new light on why corrupt politicians are surprisingly often reelected (Chang et al., 2010); if the salience of corruption is critical for corruption voting, as recently suggested (Ferraz & Finan, 2008; Klašnja et al., 2014), then by disarming the mechanisms that could otherwise draw voters’ attention to the issue, the entrenched party holds a considerable advantage.
Finally, our results are also relevant for policymakers and scholars in public administration. This study implies that advocates of marketization in the public sector, and students thereof, should pay close attention not only to the administrative, but also to the party political context in which such NPM reforms are implemented (Hood, 1991; Osborn & Gaebler, 1992; Pollitt & Bouckaerd, 2011). To some extent, it, therefore, answers O’Toole and Meier’s (2015) call for a more general theory of public management that also takes the political context into account. The last decades have seen a dramatic increase in public procurement and other NPM-related reforms (Brown et al., 2006; Hood & Dixon, 2015) and while creating a market for, say, infrastructure or elderly care might hold potential for increased productivity in theory, such reforms might, instead, risk being turned into partisan assets in the hands of local party bosses.
Supplemental Material
bdf_replication – Supplemental material for Political Competition and Public Procurement Outcomes
Supplemental material, bdf_replication for Political Competition and Public Procurement Outcomes by Rasmus Broms, Carl Dahlström and Mihály Fazekas in Comparative Political Studies
Supplemental Material
bdf_replication_dofile – Supplemental material for Political Competition and Public Procurement Outcomes
Supplemental material, bdf_replication_dofile for Political Competition and Public Procurement Outcomes by Rasmus Broms, Carl Dahlström and Mihály Fazekas in Comparative Political Studies
Supplemental Material
Political_competition_Appendix – Supplemental material for Political Competition and Public Procurement Outcomes
Supplemental material, Political_competition_Appendix for Political Competition and Public Procurement Outcomes by Rasmus Broms, Carl Dahlström and Mihály Fazekas in Comparative Political Studies
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research for this article was financially supported by the research project, “Out of Control or Over Controlled? Incentives, Audits and New Public Management,” Riksbankens Jubileumsfond (the Swedish Foundation for Humanities and Social Sciences), Grant No. SGO14-1147:1.
Supplemental Material
Notes
Author Biographies
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
