‘Time to move on’ or ‘taking more time’? How disregarding multiple perspectives on time can increase policy-making conflict

This article argues that when different perspectives on time remain disregarded in a public policy debate, policy-making conflict can increase. We present an in-depth qualitative analysis of media articles from 2005, 2009, and 2014 in the debate surrounding the contested ‘Oosterweel connection,’ a multibillion-euro infrastructure project in Antwerp (Belgium). Although concerns of time management motivated arguments to speed up the policy-process, the insensitivity of policy-makers to multiple perspectives on time increased conflict. Firstly, while administrative actors reasoned mainly from a procedural time perspective and saw time as scarce, citizens reasoned mostly from an impact-based time perspective and saw time as abundant. A binary debate on policy-making tempo (high versus low) ensued. Secondly, political actors often reasoned from political perspectives on time. Their actions, which were intended to appease, did not end the binary debate and sometimes reinforced it. As the debate on the Oosterweel connection persisted, parties increasingly believed that not only were their infrastructure goals incompatible, but so too were their goals for time management. This increased conflict.


Introduction
Time is of the essence in the policy-making process for spatial projects. After all, a widely shared belief is that successful spatial projects are those that finish on time, with project success typically defined as projects 'being delivered on budget, time, and benefits' (Flyvbjerg, 2014b: 11). However, actors in spatial projects do not necessarily think of time in the same way. Whether a project is advancing quickly or slowly depends on whom you ask. The literature on spatial projects and spatial policy-making, however, has not fully recognised the existence and interaction of different perspectives on time. Much of this literature focuses on explaining and countering delay (e.g. Flyvbjerg, 2014a;Hall, 1982;Morris and Hough, 1987). Time is treated here as a measure for project success (being 'on time') rather than as a research subject in its own right. A different strand of literature, on decision-making 'lockin,' does address time as an explicit research subject (Arthur, 1989;Cantarelli et al., 2010). However, in this work time is still treated as an objective point of reference for all actors involved.
We argue that different actors within one spatial project may have different perspectives on time, and that recognising these perspectives helps to explain policy conflict. The idea that different actors can experience time differently has been presented by various sociologists (Adam, 1990(Adam, , 2008Nowotny, 1996), urban theorists (Graham and Healey, 1999: 627;Harvey, 1996: 210-247) and scenario planners (Frame, 2008;Rickards et al., 2014). It has also received increasing attention in the field of policy analysis (Bressers et al., 2013;Eshuis and van Buuren, 2014;Pollitt, 2008: 59-63). The impact of time perspectives on policy-processes, however, still remains understudied (see also : Bressers et al., 2013: 26;Eshuis and van Buuren, 2014: 402).
The theoretical contribution of this article is threefold. Firstly, by studying time not as an objective point of reference but as a social construction, we add a fresh perspective to the literature on spatial projects and spatial policy-making. This article, secondly, answers calls (Eshuis and van Buuren, 2014: 419;Pollitt, 2008: 61-63) from the field of policy-analysis to pay more attention to the way in which time constructions shape policy-processes. More specifically, while the existing literature provides conceptualisations and demonstrates the influence of time perspectives on policy-making, it has not really unravelled mechanisms of how different time perspectives interact with each other beyond stating that this causes 'tensions' (Eshuis and van Buuren 2014: 417). Lastly, and relatedly, this article investigates the role of time perspectives in policy-making conflicts. The literature on (spatial) policy-making acknowledges that those who make plans and set time in policy-processes have a lot of power in setting the conditions for the policy-process. But it does not investigate how the struggle over time management may increase conflict between different parties in a policy-process.
This article focuses on the multiple perspectives on time in one spatial project: the Oosterweel connection in Antwerp (Belgium). The Oosterweel connection is a contested planned highway, currently estimated to cost 3.5 billion euros. We first discuss the main concepts of our study and our methodological approach. Next, we present the most important findings of our mediaanalysis. The following section analyses these findings and presents our main argument: when different time perspectives remain disregarded in a public policy debate over a spatial project, a binary debate on policy-making tempo can ensue, making the act of time management increasingly contested. As a result, the conflict between parties increases.
The management of multiple time perspectives in spatial policy-making Time as success or constraint. The planning and construction of large spatial projects notoriously tends to take more time than initially calculated (Flyvbjerg et al., 2003;Siemiatycki, 2009). Dealing wisely with time is therefore often taken to mean finishing projects on time. The literature on so called 'planning failures' even defines failure in temporal terms, with success typically defined as projects 'being delivered on budget, time, and benefits' (Flyvbjerg, 2014a: 11; for empirical applications, see also : Hall, 1982: 138;Morris and Hough, 1987: 193). This literature provides various explanations for projects not finishing on time. Studies focus, for example, on countering 'optimism bias': the tendency of policy-makers to underestimate time up front (Kahneman and Lovallo, 1993;Kahneman and Tversky, 1979). Such studies analyse why time is estimated poorly and how to improve estimations (e.g. Flyvbjerg, 2008Flyvbjerg, , 2009Kahneman, 2011;Kirkebøen, 2009), but they see time as a measure of success (being 'on time') rather than as the core research subject.
The literature on 'lock-in' in spatial policy-making does take time seriously as a core research subject (Cantarelli et al., 2010). Lock-in literature sees policy-making errors as resulting from actors being captured by decisions from the past, for example, when it comes to choosing one technology over the other (Arthur, 1989). As a result, the decisions that actors make in the present are suboptimal. Lock-in studies are informed by path dependency theory (Hall and Taylor, 1996;Krasner, 1988;Mahoney, 2000;Pierson, 2000) and sunk costs theory (Arkes and Blumer, 1985). Path dependency theory explains historical situations (e.g. failed infrastructure projects) by focusing on the historical path leading up to these particular situations. The term 'sunk costs' refers to investments that have been made and cannot be recovered (and are thus 'sunk'). These should not influence decision-making as they are already lost, but often do (see also: Wilson and Zhang, 1997). Here, time is conceptualised as a uniform constraint on future decisions. Whether the past can be experienced in multiple ways, for example, is not analysed. This paper conceives of time differently. It argues that actors within one policy-making process can have different perspectives on time. If different actors can have different perspectives on time within one process, both these perspectives themselves and their subsequent demands for time management become highly contestable.
Multiple perspectives on time in the policy-process. It is in itself not a novel insight that time can be conceived of differently by different actors. Sociologists such as Adam (1990Adam ( , 2008 and Nowotny (1996), as well as urban theorists (e.g. Graham and Healey, 1999: 627;Harvey, 1996: 210-247) and scenario planners (Frame, 2008;Rickards et al., 2014) have already studied time as a social construction, rather than an objective point of reference. These authors, however, do not investigate the impact of such constructions on concrete policy-making processes. This is also not their goal, as they aim to deconstruct the taken-for-granted notions of space and time in their respective disciplines.
How the constructed nature of time affects policy-making processes has been explored in the field of policy-analysis (Bressers et al., 2013;Eshuis and van Buuren, 2014;Pollitt, 2008: 59-63). However, this field of study is still relatively young. Hence, the existing studies only scratch the surface of how social constructions of time impact policy-processes (Bressers et al., 2013: 26;Eshuis and van Buuren, 2014: 402). Still, evidence gathered thus far demonstrates that differences in how policy actors conceive of time can shape important aspects of policy-making, such as collaboration (Eshuis and van Buuren, 2014) and evaluation (Bressers et al., 2013).
The contribution of Eshuis and van Buuren is especially helpful, as these authors identify four dimension of a perspective on time: time-horizon, time-character, time-budget, and tempo (Eshuis and van Buuren, 2014: 407). The time-horizon explains whether actors have a short-term, medium or long-term vision on a policy issue. This depends on their institutional embeddedness. A civil servant responsible for monitoring the procedural progress of a policy-process obviously has a different horizon from a politician who wants a project built within her term of office. From a time-horizon, then, follows a specific vision on the character of time: the sequentially structured time, ruled by administrative procedures, perceived by a civil servant, may be at odds with the cyclically structured time, ruled by elections, perceived by a politician. Time can also be experienced as unstructured and erratic. Time-horizon and the character of time subsequently shape perspectives on time-budget (is there ample time or little time) and tempo (is a process going fast or slow). A civil servant with a sequential view on time may feel that there should be sufficient time to maintain procedural integrity, while a politician with a cyclical view on time may view the same procedures as eating up time in the face of upcoming elections. Together, these four dimensions make up a time perspective.
Multiple time perspectives and conflicting demands for time management. To this point we have argued that different actors within one spatial project can have different perspectives on time, and that this approach to time remains underdeveloped. We have also conceptualised a perspective on time by distinguishing between time-horizon, time-character, time-budget and tempo. But we have so far left out the topic of who manages time. However, although various stakeholders may have different time perspectives, some actors have more power than others to set the time.
Policy-makers manage time by deciding how to phase the policy-making process and how much time to allow for the various phases. By phasing the process, they automatically set some of the conditions for the policies themselves: policies requiring activities that would delay the process, after all, can be framed as inappropriate. Actors outside of the policy-making process, however, are not powerless when it comes to the setting of time. They, too, can try to influence time management. When different parties in a policy-making process have different demands for time management, this may result in conflict. We define conflict as: 'the process that arises when two or more persons or groups manifest the belief that they have incompatible objectives' (Kriesberg and Dayton, 2017: 2). An illustration is provided by the work of van Eeten (1999: 39-63). He discusses a conflict between protesters and governmental actors over a large dike improvement project in the Netherlands. Protesters asked for more time to study alternatives with a smaller spatial impact. Governmental actors, however, replied that time delays were irresponsible given the risk of flooding. The government's refusal to concede more time, in turn, caused suspicion with protesters and led to increased efforts to block decision-making. A policy stalemate followed that only ended when the area again flooded in the 1990s. In this example, it seems that by prioritising their own time perspective and subsequent demands for time management, policy-makers increased conflict over the dike improvement project instead of ending it.
It is widely acknowledged that policy-makers have significant, but contested, power in reimagining future places when making spatial policies (Healey, 2006;van Dijk, 2011) and that they set the often arbitrary conditions for what counts as rational action in policy-processes (Flyvbjerg, 1998). It is also acknowledged that the setting of time is an attribute of power (Goetz, 2014;Goetz and Meyer-Sahling, 2009). We hope to add to the existing literature by investigating the relationship between different time perspectives and conflict over time management in spatial policy-processes.

Methods
This article studies the perspectives on time in the public debate over the Oosterweel connection in Antwerp. The Oosterweel connection is a planned highway that became the subject of a heated discussion in 2005. Today, the project is still publicly contested and remains in the planning phase. Because the policy conflict has endured for more than a decade, it is a good case for analysing how temporal perspectives interact with each other. The richness of the case thus holds promise for theoretical learning on the interaction of different time perspectives within a policy discourse, as well as for practical learning about the spatial policy-making process (Stake, 1994(Stake, , 2010Timmermans and Tavory, 2012). We studied media articles reporting on the Oosterweel connection in 2005, 2009 and 2014. These years represent the peaks in media attention, offering a rich discourse on the time and timing of the project. The peaks were identified by searching for articles containing the word 'Oosterweel connection' (Dutch: 'Oosterweelverbinding') in the period from 2000 to 2014. We used the online database 'GoPress', which contains all Flemish newspapers. We included the following newspapers: 'De Standaard', 'De Morgen,' 'Gazet van Antwerpen,' 'Het Laatste Nieuws,' 'De Tijd,' 'Het Nieuwsblad,' 'Metro', and 'Het Belang van Limburg.' These represent all Flemish subscription newspapers and one widely read free newspaper (Metro). We chose the Flemish edition for each newspaper. If only regional editions existed, we chose the Antwerp edition, as Antwerp is where the Oosterweel connection is planned for construction. Figure 1 shows the number of articles found for these years.
For each year we first scanned through the titles. Next, we identified several months in each year for an in-depth qualitative analysis. These months were connected to key moments, 1 and also scored highest in terms of media attention. We analysed a total of 739 articles (Table 1).

Figure 1. Articles on the Oosterweel connection.
Coding was performed using the 'Nvivo' software. The first step was coding the peak months for each year inductively. Our goal was to study the way in which time was made sense of in the Oosterweel debate. Coding therefore concentrated on identifying arguments related to time. In terms of how we identified arguments, we focused on two distinct discursive elements. One was the advancement of evidence. For example: actors referring to research to argue for rapid construction of the Oosterweel connection. The other was the advancement of symbolic language. Symbolic language consisted mostly of metaphors, which reveal the way in which one thing is understood in terms of another (Lakoff and Johnson, 1980) and thus attuned us to how the actors imaginatively constructed specific images of Oosterweel to make statements in terms of time. For example: by describing Oosterweel as 'mortgaging' the city of Antwerp. Symbolic language also contained proverbs and the comparison of Oosterweel to other infrastructure projects or to events in the distant past. We attributed a wide range of codes to these discursive elements such as: 'standing still' (for a statement like 'we have lost time, because the planning process has been standing still'); 'going forward' (for a statement like 'the rapid planning procedure is shoving Oosterweel down our throats'); and 'research' (for a statement like 'considering trajectories proven inferior by prior research will only amount to unnecessary delay').
After a first round of coding, we regrouped smaller codes under larger labels. In this second coding step we also recoded pieces of text that, in retrospect, fitted better under a different label. The resulting coding process clearly distinguished between arguments in favour of speed in the planning process and arguments in favour of taking more time to deliberate. Based on this insight, we revisited the various temporal arguments in step 1 and paid specific attention to the way in which actors constructed time perspectives by embedding events in a past, present and future and how this related to time-horizon, time-character, time-budget, and tempo. Our coding process thus followed what is called the abductive approach, with iterations between data and theory (Schwartz-Sea and Yanow, 2012; Timmermans and Tavory, 2012). We found a clear difference between the arguments that presented time as structured through the planning procedure (e.g. delay would be very costly), time as structured through the rhythm of elections (e.g. speed is necessary because of approaching elections) and time as structured according to the impact of infrastructure (e.g. there should be ample time to deliberate on a project having such impact for years to come). Based on this final analysis, we constructed the different temporal perspectives in the debate over Oosterweel. These are presented below, after first discussing the most important temporal arguments in the public debate over Oosterweel.
Temporal arguments in the public debate over Oosterweel In this section, we discuss the most important temporal arguments in the public debate over the Oosterweel connection at three different points in time: 2005, 2009 and 2014. For each year, we distinguish between arguments pleading for keeping up the decision-making tempo on Oosterweel and arguments pleading for decreasing the tempo. We start by briefly discussing the history of the Oosterweel connection before 2005.
The early history of the Oosterweel connection. The region of Antwerp has been dealing with traffic congestion for a long time. It is commonly thought that the make-up of the Antwerp ring road, constructed in the 1960s, is largely responsible for this. The ring road does not form a circle around Antwerp, but something closer to a semi-circle that crosses directly through the city, thus inhibiting through traffic from bypassing Antwerp. At present, 270,000 cars and trucks use the ring road on a typical business day, making it the most crowded highway in Belgium. In 1995, the Flemish road agency 2 declared it a priority to close the circle of the Antwerp ring road (Verelst, 2009). The trajectory for this enclosure was located in the Oosterweel area. Hence the name of this trajectory: the 'Oosterweel connection,' often shortened to 'Oosterweel.' The project consisted of a tunnel under the Scheldt River and an overpass over part of the city. Figure 2 both the Antwerp ring road and the plans for the Oosterweel connection (dotted). In 2005, the scale model for the Oosterweel connection was presented to political actors from the Flemish government and the city of Antwerp. This is when the public debate on Oosterweel begins. In 2005 the Oosterweel project is presented to the wider public. Below we present the temporal arguments expressed in the public debate at key points during this year.
Time to move on. Not long after the presentation of the scale model, a debate arises over the spatial impact of the Oosterweel overpass on the city. Consequently, the city of Antwerp orders a second opinion to investigate whether the overpass could be replaced by a tunnel. The city faces criticism for this, which mainly focuses on timing. Why order a second opinion so late into the process, actors from BAM (the administrative body responsible for project execution) and from the Flemish government wonder, when you previously committed to the same project? Moreover, various political actors from the Flemish government claim that Antwerp's ill-advised timing might jeopardise the larger political deal the Oosterweel connection has been embedded in on the Flemish level. In addition to the Oosterweel connection, that is, plans have been made to execute various bicycle and infrastructure projects in Antwerp. However, after the study concludes that a tunnel would be more expensive than an overpass, the city of Antwerp recommits to Oosterweel and the planning procedure continues.
Nevertheless, the debate on Oosterweel soon resurfaces. At a press conference in September, the action group 'Straten-Generaal' denounces the Oosterweel connection because it would allegedly inhibit future city development in Antwerp. Instead, an alternative trajectory further away from the city is proposed. Administrative and political actors respond that the current plans are the result of many years of discussion and research. Further reflection is thus unwarranted; it is time to move on to the next phase of the planning process.
Taking more time. Rather than arguing for speed, other voices in the public debate argue for taking more time. Various civic actors with expertise on urban planning join the public discussion started by Straten-Generaal. One architect illustrates his reasoning as follows: 'I think that it is more important to take our time today, a few months or even a year, than to decide now on something that will mortgage the city for a hundred years to come' (as cited in Rinckhout, 2005: 8).
Moreover, in addition to the substantive discussion regarding infrastructure, the citizens from Straten-Generaal voice concern over the very insistence on speed by decision-makers. They stress that hardly any Antwerp resident had even heard of the Oosterweel connection before 2005. How could citizens have voiced concerns earlier if they did not even know about the plans, they ask.
2009: Blaming the other for delay.
Four years later, the conflict over Oosterweel remains ongoing. By this time, the action group 'Ademloos' (literally, 'Breathless') has joined the protest. Ademloos resists Oosterweel over health concerns: a busy highway close to residential neighbourhoods would increase the volume of fine particles in the air. In the summer of 2008, faced with growing protests, the government asks the independent 'Arup/ Sum' research consortium to compare the Oosterweel connection to other trajectories. Meanwhile, Ademloos collects enough signatures to hold a referendum on the Oosterweel connection in Antwerp. In 2009, the results of the Arup/Sum study are presented and the municipal referendum is held. We will discuss the temporal arguments from the public discussion in this year.
Cutting the knot. The Arup/Sum study concludes that none of the trajectories is good enough.
They instead propose a new trajectory which strongly resembles the alternative of action groups.
The study is interpreted in various ways. One of these interpretations, voiced by administrative actors from BAM and the responsible minister, is that only the Oosterweel trajectory would guarantee a quick solution to the worsening problem of traffic congestion. This is because choosing an alternative trajectory would require a new Environmental Impact Assessment (EIA), resulting in years of delay. This interpretation is ridiculed by action groups, who ask the question why the study was even ordered if its conclusions would be dismissed.
Despite these different opinions, political actors from the Flemish government feel that they have to make a decision on the issue quickly because the elections for a new Flemish government are approaching. As a compromise, the government decides to further research the alternative Arup/Sum trajectory, but also allows the Oosterweel project to continue in the permitting procedure.
In the lead-up to the Oosterweel referendum in October, actors from the BAM again argue that only Oosterweel could be completed in a timely manner. Moreover, Oosterweel is presented as a necessary project with a long history that should not be stopped by opponents in its final phase. This argument resounds in several newspaper pieces; the conflict over Oosterweel is characterised as one of endless bickering, and it is asked that Antwerp citizens vote for 'moving forward' (Demeester-De meyer, 2009: 29) rather than 'standing still' (Demeester-De meyer, 2009: 29) so that the government can finally 'cut the knot' (Thomaes, 2009: 72) and begin construction.
Pushing through Oosterweel. An alternative reading of the Arup/Sum study not only concludes that the Oosterweel trajectory is inferior, but also complains that it took so long even to find these results. After all, protesters claim, the government resisted appeals for further reflection for several years.
Reminding the public of this history, protesters argue that Oosterweel should certainly not be pushed through just for the sake of moving the procedure forward. When the political compromise is reached of further researching the Arup/Sum trajectory while also letting Oosterweel continue in the permitting procedure, action groups are furious. To simply continue with the Oosterweel project after four years of conflict and an independent study is, in their words, 'simply rude' (as cited in Mels, 2009: 1).
In the campaign for the referendum, protesters argue that cautious rather than fast decisions are needed, as Oosterweel will have effects for decades. They also stress that the government is to be held responsible for the slow decision-making process so far. One reader of a newspaper puts it as follows: 'What have all those people [from BAM] even been doing for the past 11 years with all that money and time, that their only argument remains: if this is voted down, we will lose valuable time' (Leijssen, 2009: 22).
By now, several political actors that had supported the Oosterweel project before, join the action groups in their opposition. Some of them, most notably the mayor of Antwerp, are from political parties that are also members of the Flemish coalition. The changing of positions is met with resentment by other coalition parties, who portray the so called 'U-turn' (e.g. Brinckman, 2009: 7) as a breach of prior commitments for pure electoral gain. The mayor of Antwerp, for example, is accused of changing positions to advance his re-election in 2011. However, those that change position argue that they had been politically pressured during previous terms of office into favouring Oosterweel just for the sake of upholding political commitments.
Eventually, 59.24% of the voting Antwerp citizens reject the Oosterweel connection in the referendum. The referendum, however, is not legally binding, which prompts new debates over what to do next. However, at least for the moment, the planning process for the Oosterweel connection is put on hold.

2014: Learning from history?
In 2014, the Oosterweel debate continues. The original project has now undergone some major revisions, most notably through the decision of 2010 to replace the contested overpass with a tunnel. This restores the political consensus over Oosterweel. It does not, however, appease action groups who stress that they object to the location and not merely the overpass and are frustrated by the forging of the decision in politico-administrative circles rather than in an open dialogue. The revisions do require a new EIA to compare different highway trajectories. It's results are presented in February; in May the elections for a new Flemish government take place.
Breaking ground. Because the EIA states that the Oosterweel project scores highest among all trajectories for mobility, the government concludes that the Oosterweel connection is the preferred trajectory. BAM's chairman asks protesters to stop their resistance, while political parties from the Flemish coalition reconfirm their commitment to Oosterweel in the so-called 'Valentines agreement.' They stress that the decision-making impasse should end now and that ground must be broken as soon as possible.
However, as the Flemish elections draw closer, a new action group called 'Ringland' emerges. Ringland also recognises the impasse, but argues that to get beyond this, no new infrastructure should be built for the moment. Rather, the existing ring road should be tunnelled. In this way, Ringland claims, the existing traffic could be managed better and the space currently used by cars could be reclaimed for citizens (hence the name Ringland). Their plan is praised in various newspaper articles.
Administrative actors and actors from coalition parties alike do not object to tunnelling the ring road as such. However, they refuse to reopen the discussion on Oosterweel. As one politician explains: We have come a long way already in developing the BAM trajectory [another name for the Oosterweel trajectory] and we cannot reset the counter now. It will then take another ten years before anything actually happens. And that would entail economic selfdestruction. (as cited in Moolenaar, 2014: 3) Changing priorities. A different EIA interpretation reads that the track proposed by the action groups performs best on liveability and that in terms of mobility there is almost no difference between this trajectory and Oosterweel. Hence, many citizens critical of Oosterweel are angry when political actors so quickly state that Oosterweel is the better option. This sentiment is illustrated by the following quote from a reader's letter: On Monday the EIA is presented, on Wednesday the newspapers tell us that first changes have been made to the Oosterweel design, on Thursday one can read about additional changes and on Friday the three political parties involved have suddenly decided on everything. In my opinion this is equal to fraud. (Goyvaerts, 2014: 35) Protesters continue to resist Oosterweel. They also get a new ally in the lead-up to the election. One political party that had always backed Oosterweel decides that the plan lacks support and starts opposing it. Meanwhile, the Ringland action group refuses to take a position for or against Oosterweel, but merely stresses the priority of tunnelling. Consequently, Ringland asks that the timing of Oosterweel be adapted. Asked how it would react if political actors choose to ignore their plea, an actor from Ringland warns that this may be a very costly decision in terms of time, as the continued opposition would only lead to more delay in decision-making. At present (April 2017) the planning process for the Oosterweel connection is still ongoing.
How a binary debate on tempo increased policy-making conflict The previous section described the main temporal arguments in the public debate on the Oosterweel connection. We now analyse these arguments in three steps. First, we present three time perspectives that underpin the temporal arguments in the Oosterweel debate. Next, we analyse how these different perspectives interact with each other and, thirdly, how this impacts policy conflict.
Three perspectives on time.
In our theoretical section we explained how to discern perspectives on time by studying timehorizon, time-character, time-budget, and tempo. We can now discern three perspectives on time in the Oosterweel debate, namely: a procedural perspective, an impact-based perspective, and a political perspective.
Firstly, we find a procedural perspective on time, in which new events are interpreted as being part of a planning procedure with phases and milestones. This perspective takes the policymaking procedure as a starting point and the time-horizon, consequently, is short-to mediumterm. The character of time is sequential. Time is structured in a policy-making process with a fixed beginning, a projected ending, and activities that build on each other. The projected end date defines the remaining time-budget. Hence, it is important to keep up the tempo so that time is not lost. In the debate over Oosterweel, this perspective on time stresses rapid completion. The project has been in preparation for many years and has been studied repeatedly. It now needs to move forward in the planning procedure so that it can be completed on time. This perspective is primarily advanced by administrative actors whose task it is to manage the project procedures.
Occasionally, political actors in favour of rapid decision-making and citizens and media growing increasingly tired of the continuing discussion over Oosterweel also adopt this perspective.
Secondly, we find an impact-based perspective on time, in which new events are interpreted as being part of an ongoing debate over Oosterweel's long-term consequences. The time-horizon is long-term and it takes the impact of building new infrastructure as its starting point. It is better to take the time for serious reflection now, it is argued, than to make bad and irreversible decisions that will harm the city for decades to come. The time-character is thus sequential, but the impactbased perspective focuses on the consequences of decisions rather than on the procedure itself. This also means that the time-budget for decision-making is perceived as ample, but as soon as infrastructure has been built there is no going back. When it comes to tempo, cautiousness rather than speed is stressed. This perspective is advanced primarily by citizens and action groups, but is also occasionally adopted by political actors who join the protesters in their resistance.
The third perspective on time is a political one and is advanced almost exclusively by politicians from ruling parties. The time-horizon is short-to medium term and the electoral cycle is the focal point of reasoning. Because of the cyclical nature of elections, the character of time is cyclical. Ruling parties have to uphold the agreements they made in the election aftermath, while elections provide the opportunity for parties to (re)position themselves on policy-issues. The time-budget is defined by the elections. Upcoming elections mean that the time is scarce for ruling parties. Moreover, the campaign offers a brief window to take a political position, which also results in time scarcity. The preferred tempo of policy-making varies, depending on perceived time scarcity. We mostly see demands from ruling parties for a high tempo, either to fulfil electoral promises or to reach decisions before electoral campaigning starts. On rare occasions, such as in the lead-up to the 2009 municipal referendum, ruling parties have also made the argument that the tempo of decision-making should be paused for political reasons. In this case, the referendum seems to have allowed for a political window for repositioning reminiscent of electoral campaigns. Table 2 presents an overview of the various time perspectives in the public debate over Oosterweel, as well as an graphic illustration of these perspectives.

How perspectives on time further diverge in a binary debate
Diverging time perspectives in a binary debate. The previous section explained how different characterisations of time-horizon and time-character engender different time-budgets andpolicymaking tempos. We now argue that in the Oosterweel debate, the persistence of different, unacknowledged time perspectives led to a binary debate over the tempo of decision-making. The binary nature of this debate, in turn, contributed to the different perspectives further diverging.
From the start, a tension between the impact-based perspective pleading for more time and the procedural perspective that resisted this plea because time was seen as too scarce was present.
We can understand this tension when we remember the very different time-horizons of these two perspectives. A procedural time perspective focuses on the short-to medium-term procedure of policy-making, the progression of which is managed by setting deadlines and organising activities that build on each other. An impact-based perspective, on the other hand, focuses on the long-term consequences of the project and the permanence of the built environment. From one perspective, progress is welcomed; from the other, progress may be dangerous if it goes at the expense of caution.
In 2005, pleads by citizens to pause decision-making were disregarded with the argument that enough research had been conducted already. Rather than recognising various perspectives on time, policy-makers argued that the only sensible thing to do was to keep up the pace of decisionmaking. This emphasis on progression, however, stirred the suspicion of those who reasoned from an impact-based perspective on time.
The unwillingness to open up discussion was regarded as proof of a flawed policy-process, more concerned with project completion than with the longterm repercussions of infrastructure. The irreversibility of infrastructure construction, and the negligence policy-makers seemed to display towards this aspect, provided the grounds for protesters to increase their efforts to try and pause decision-making. For example through organising a municipal referendum.
This, in turn, stirred the suspicion of those with a procedural perspective. As the policy procedure got delayed, the onset of delay was seen as proof of obstruction by action groups, rather than as a plea for more discussion. From a procedural perspective on time, after all, delay equalled the loss of valuable time. Oosterweel was presented as a necessary project with a long history that could not be obstructed in its final phase.
As delays increased, procedural thinkers attempted to 'make up' for time 'lost,' which in turn made those with an impact-based perspective even more suspicious. A binary debate on aspired tempo ensued: arguments increasingly focused on either increasing the tempo of decisionmaking, as enough time had allegedly already been lost, or decreasing it, with pleas for speed regarded suspiciously. As a result, these time perspectives that were already in tension with each other to begin with, diverged even more. The dynamic of different time perspectives further diverging in a binary debate on aspired tempo is illustrated in Figure 3. However, there are a few key moments when politicians advanced their own, political, time perspective. Mostly, they stressed the need for project progression, either to make good on political agreements made earlier in their term (in 2005) or because of upcoming elections (in 2009). However, during rare political windows (during elections and during the lead-up to the referendum), ruling politicians have argued for temporarily pausing decision-making. Neither plea, however, has succeeded in putting an end to the binarity in the public over tempo. Moreover, at several points in time, decisions motivated by a political time perspective seems to have strengthened this binarity.
In 2009, for example, political actors decided that because the elections were coming up they could not make any final decisions on the Oosterweel connection. As a result, they allowed both the planning process for Oosterweel to continue while also commissioning additional research into the Arup/Sum trajectory. The Oosterweel project thus continued to move forward in the permitting procedure, strengthening the claim that opting for an alternative trajectory would amount to even more delay. However, the call for additional research also suggested that time would be taken for serious debate, thus increasing the frustration of impact-based reasoners when the debate was not opened up after the elections but instead a compromise on Oosterweel was forged in politico-administrative circles. These actions may have made sense from a political perspective, but ended up strengthening the dichotomy over tempo in the public debate over Oosterweel rather than pacifying it.
How conflict between parties increased. In our theoretical section we defined conflict as the process that arises when two or more persons or groups manifest the belief that they have incompatible objectives. As the debate on the tempo of Oosterweel policy-making became increasingly binary, parties for or against the Oosterweel connection were not merely fighting over different highway trajectories anymore. They also increasingly believed that their objectives for time management were incompatible. Demands for keeping up the decision-making tempo became suspicious in and of themselves, because they were cast as leading to bad and irreversible decisions, while demands for decreasing the tempo became equally suspicious, as they were cast in terms of obstruction. This was evident in 2009, when the delay was used to argue both for (because too much time had already been lost) and against (because effort to push through plans had to be resisted) the need to make decisions quickly. And in 2014, when there was a general acceptance of tunnelling the ring road in the public debate, but the conversation again spiralled into conflict over tempo.
As a result of believing that their objectives for time management were incompatible, the activity of structuring the policy-process became an additional source of conflict. Paradoxically, then, although arguments for a high policy-making tempo were advanced out of concerns for time management, they only added to conflict protraction rather than rapid project completion. In the policy-making process over the Oosterweel connection, the public debate increasingly became splintered over the question of how to order time, in addition to being divided over the question of how to order space.

Conclusion and discussion
Through an in-depth qualitative study of newspaper articles about the contested 'Oosterweel connection' in Antwerp, this article has argued that when different time perspectives remain disregarded in a public policy debate over a spatial project, a binary debate on policy-making tempo can ensue. As a result, the act of time management itself becomes increasingly contested and the conflict between parties wielding different time perspectives grows.
The contribution of this article to the literature on spatial projects and spatial policy-making is threefold. Firstly, it demonstrates that while it is understandable for policy-makers to emphasise procedural progression to prevent delay, insensitivity to multiple time perspectives can actually lead to conflict protraction and hence further delay. Attempts to avoid so-called planning 'failure' (Flyvbjerg 2014a;Hall, 1982;Morris and Hough, 1987), can paradoxically thus contribute to planning failure. Secondly, the analysis demonstrates the largely covert power that policy-makers have when making spatial policies. By structuring the policy-process, they not only set the conditions for which policies make (no) sense in terms of space (Healey, 2006;van Dijk, 2011) but also in terms of time. A time perspective can thus activate a specific rationality (Flyvbjerg, 1998) that policy-makers employ to further their preferred policy at the expense of other options. The latter conclusion, relatedly, may also add to the literature investigating decision-making lock-in. Lock-in, we argue, may result from policy-makers having specific demands for timemanagement that lock-out other options.
The article also contributes to the emerging body of literature in policy-analysis that investigates the role of time constructions in policy-making (Bressers et al., 2013;Eshuis and van Buuren, 2014;Pollitt, 2008). Our analysis teases out two theoretical mechanisms that may be activated when policy-processes harbour multiple perspectives on time. Firstly, it demonstrates that different time perspectives may not only be in tension with each other, but that these perspectives can further polarise as a result of confrontation. In the Oosterweel debate, an increasingly binary debate over policy-making tempo ensued which contributed to further diversion of perspectives. Secondly, the analysis demonstrates that prioritising one perspective on time can make the act of time management increasingly contested and consequently increase the conflict between various parties in a policy-process.
These findings underscore the need for what is also called 'time-sensitivity' (Bressers et al., 2013;Eshuis and van Buuren, 2014) in policy-making. Policy-makers should be sensitive to the time perspective of other parties, as those who fail to do so may ultimately do so at their own peril. In the Oosterweel case, not only did the policy-process become severely delayed, the conflict between parties also became much harder to settle when time management itself became a topic of contestation. Indeed, actors seemed to have lost trust in the very potential of alignment between time perspectives and their demands for time management, and as Eshuis and van Buuren (2014: 419) stress: 'without trust and reciprocity, the process of alignment will be extremely cumbersome.' Operating in time-sensitive ways is no easy task for policy-makers. Firstly, they may not recognise other perspectives on time. Hence, awareness of the importance of time-sensitivity is a first step. Moreover, even if policy-makers recognise other time perspectives, they may find themselves bounded by procedures. There is, for example, an inherent tension between following up on past promises and procedural guidelines and being responsive to what citizens demand 'today.' A second step to improve time-sensitivity may therefore be to redesign some aspects of the spatial policy-making process. Our conclusions could for example question the trend of designing highly integrated, tight planning processes (e.g. Lenferink et al., 2013). When many different aspects of a planning process are temporally linked, the ability of policy-makers to reflect on and respond to changing circumstances may be compromised. As our analysis demonstrates, tight planning with a strong focus on efficiency can prove highly inefficient in the end.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The work was supported by Province of Antwerp and University of Antwerp Research Foundation. Notes 1. The empirical section will further elaborate on these.
2. In Belgium, spatial planning/ public works are governed by the regional governments of Flanders, Wallonia and Brussels. In the case of Oosterweel, the Flemish government has jurisdiction.