Crafting or mass-producing decisions: Technology as professional or managerial imperative in public policy implementation

Abstract

Policy implementation is characterized by professional public service workers who make decisions about clients using knowledge and skill-sets acquired through years of training and experience. Their unique position separates them from other workers, provides them with autonomy, and enables them to challenge managerial directives. Information and communications technology is used to tame this power. Whereas public service workers have been criticized for having too much influence, technology may shift decision-making from a professional craft to technology-driven mass-production. This article studies how technology impacts policy implementation in seven sub-stages resulting in alternating professional and managerial imperatives in all sub-stages except for discretionary practices. Whereas managers, public service workers, and clients can appreciate that professional norms are strengthened, and managerial goals achieved, there is a growing concern about the role of technology and its influence on public service workers responsible for policy implementation. The article ends with suggestions for future research.

Keywords

Technology e-government policy implementation decision-making public innovation

1. Introduction

Public service workers responsible for policy implementation struggle with several challenges. They experience policy uncertainties, increased workloads, and budget cuts. Yet, they are expected to make decisions of high professional standards implying fair and equal treatment of clients. These expectations often reflect a rationalistic view of policy implementation assuming that public service workers have the time and resources they need to gather all relevant information about each case and compare this information to policy maker intentions and similar cases (Hogwood & Gunn, 1984). In reality, they have limited time to gather information (sometimes they make decisions on-the-spot), the required combination of resources may be unavailable (Hogwood & Gunn, 1984), and pieces of the puzzle are often lacking (Lipsky, 2010). Thus, they use the power and freedom that come with their professional status to adapt their work situation and even challenge managerial directives (Noordegraaf, 2007). Inspired by private sector management principles and reflecting a hierarchical view of policy implementation (i.e., policies are expected to be implemented according to policy maker intentions), information and communications technology has been introduced to address these challenges. However, whereas policy implementation has been criticized for a too large public service worker footprint, technology-driven decisions may disclose too much algorithmic imprint. Thus, excessive goal management can lead to more focus on managerial goals (how) on behalf of professional norms (why). In the industry, craftsmanship has long ago given way to the mass-production of goods. Customers are generally satisfied with this development; they still get quality products and to reasonable prices. Products are not tailormade but meet most of the needs customers have. Even though tailormade products would be preferable, cost and efficiency gains by far surpass what is lost through mass-production. Likewise, standardized and mass-produced decisions in public service provision are highly efficient and cost-saving. Instead of crafting decisions, using time and resources to consider each case, they are more quickly decided upon based on fixed criteria. Individual adjustment and care might be lost for other benefits.

Technology has been associated with performance goals leading to the monitoring formal aspects of policy implementation such as the number of active cases, how fast cases are completed, the number of complaints, and control issues (Buffat, 2015). Other studies have paid attention to how technology can be utilized to save costs by automating or transforming work processes (Bovens & Zouridis, 2002). Whereas technology has been associated with reduced professional autonomy and acquiescence to managerial goals, it also has the potential to enhance professional norms according to the conduct, aims, or qualities that characterize a profession (Freidson, 2001). Technology can contribute to strengthening the quality of public services by preventing bureaucratic and personal biases (e.g., Wenger & Wilkins, 2009), manipulation of information streams (Snellen, 2002), and corruption (e.g., Smith, 2011). It can further provide a more solid foundation for decision-making (De Witte et al., 2016; Larsson & Jacobsson, 2013), and replace repetitive tasks with tasks requiring analytical skills (Cordella & Tempini, 2015).

Technology in policy implementation has been criticized. Discussing discretion, the freedom public service workers have to make decisions within policy boundaries, Lipsky (2010, p. 164) claimed that “the nature of service provision calls for human judgment that cannot be programmed and for which machines cannot substitute”. He argued that discretion is a necessary tool that brings ‘policy as performed’ closer to ‘policy as written’. Peckover et al. (2008) claimed that clients have a need for social interaction that computers cannot satisfy and De Witte et al. (2016) found that social workers got more concerned with processing and monitoring information than paying attention to client needs. Some studies even suggest that technology, contrary to its intentions, can result in hiding considerations made in discretionary practices instead of opening these practices up more (Wihlborg et al., 2016). Thus, technology tends to reveal formal aspects of discretionary practices instead of informal aspects (Jorna & Wagenaar, 2007). The potential severity of decisions adds to the decision complexity since some of the decisions public service workers make can have a serious impact on clients’ lives. Yet other studies suggest that a managerial-oriented use of technology can lead to deprofessionalization of public service workers and subsequent demotivation since they become more occupied with operating computers than enacting public service face to face with clients. Whereas Smith (2011) found that using new technologies could tip the power balance in favor of government institutions, other scholars have suggested that technology use may shift power in favor of system designers (e.g., Bovens & Zouridis, 2002). In some cases, unqualified people were enabled for public service work through technology use leading to tensions within the organization (e.g., Pithouse et al., 2011).

This imprint, however, is not to be taken for granted. First, technology can be purposefully co-designed in cooperation with its professional users thus taking into consideration the needs of users. Furthermore, whereas technology is designed with a specific purpose in mind, its use may be appropriated by its users and institutionalized over time (Orlikowski, 1992). This appropriation can be moderated by organizational characteristics of the workplace such as hierarchical structure, managerial ideology, and the expertise and power possessed by the public service workers (Orlikowski, 1992): “The United States, it appears, is finding that the mass production ethos is very hard to escape. Japanese workers and managers have a much greater understanding of what modern technology can do and rely far more on the judgement of people on the shop floor” (Powell, 1987, p. 196). Thus, the impact technology has in policy implementation comes down to, not only the technology itself, but how technological artifacts are designed for a policy environment and used therein.

This article investigates how the use of technology influences decision-making in policy implementation. The purpose of this study is to empirically show and critically discuss how technology use can change decision-making in sub-stages of policy implementation through the achievement of multi-complex goals. To do so, we have interviewed judges and engaged in participant observations in a Norwegian district court and interviewed case workers in a Norwegian Tax Administration office to support our claims. These informants are what Bovens and Zouridis (2002) call screen-level bureaucrats. Screen-level bureaucrats represent a large group of public service workers whose work has been considerably influenced by digital tools. Traditionally, they have interacted closely with clients and exercised considerable control of their daily work routines. Both case organizations have implemented electronic case management systems (ECMS). These systems represent technology common in many public agencies and their use is well established since their implementation 10–15 years ago. The court and the tax administration office differ in terms of societal mission, type of service provision, and task characteristics. Judges are independent law experts that handle a wide variety of cases such as murder, drunken driving, and child custody. Tax administration caseworkers report to higher authorities and specialize in tax matters. Considering their different societal roles and tasks, we believe these organizations serve as interesting cases for the study of technological impact in policy implementation.

2. Technology in policy implementation: Decision crafters vs. decision factory workers

Many studies of public policies have been conducted by analyzing the policy process in stages (and sub-stages) from the recognition of societal problems to the evaluation of policies (e.g., Dror, 1989; Hogwood & Gunn, 1984). These models often look at the phases as a set of successive steps in a particular chronological and hierarchical order1 (Hill & Hupe, 2014). Policy implementation is “the stage in the policy process concerned with turning policy intentions into action” (John, 1998, p. 204). This is the step where clients are expected to experience the practical outcomes of political priorities. However, as Anderson (1975, p. 79) points out, outcomes of public policies may be substantially modified during the implementation stage: “policy is made as it is being administered and administered as it is being made”. A stage model of the policy implementation phase is useful since it provides scholars with an analytical tool where the tasks conducted by public service workers in different sub-stages can be studied together with related and differing technological influences. Figure 1 presents sub-stages of policy implementation with their associated tasks and actions.

Figure 1.

The policy implementation process (based on Hill & Hupe, 2014; Lipsky, 2010).

Technology has been utilized to support decision-making in several of these steps. Scholars studying technology in policy implementation have viewed public service workers as ‘decision crafters’ or ‘decision factory workers’. In the former camp, scholars see technology as a tool to assist public service workers in the decision-making process. The burden of routine tasks is offloaded and tasks that require professional expertise gain increased attention (Cordella & Tempini, 2015). In this view, the professional norms that public service workers adhere to are not impaired. Instead, technology is a resource that can support professional practices. The latter camp of scholars looks at how technology makes public service workers into ‘decision factory workers’ (Bovens & Zouridis, 2002). Instead of crafting decisions, they are ultimately reduced to mere system facilitators. They work in public agencies that appear as factories “assembling” decisions on a semi or fully automated assembly line. In this view, managerial goals are emphasized on behalf of professional norms.

2.1 Decision crafters: Technology as professional imperative

Crafters are people who use personal skill sets acquired through training and experience. They are proud of their work appreciating both the process of crafting and the output. Likewise, decision crafters are public service workers motivated by helping clients (Tummers & Rocco, 2015) taking pride in their work. To them, technology is an ‘action resource’ (Hupe & Buffat, 2014; Orlikowski, 1992) that promotes professional norms such as fair and unbiased decision-making (e.g., Wenger & Wilkins, 2009), increased information quality (De Witte et al., 2016; Larsson & Jacobsson, 2013), and reduced corruption (Reddick et al., 2011; Smith, 2011). Professionalism is a set of beliefs that consider a certain type of work to be so specialized that it cannot be conducted by people without the required training and expertise (Freidson, 2001). The specialization implies that professionals have a specific set of qualifications, unique to their profession, that enables them to make decisions about clients. Since work is specialized, it cannot be standardized or rationalized (Freidson, 2001).

E-government has been a success in terms of providing clients with platforms where they can apply for services easily and inform public service agencies about their rights (Scott & Golden, 2009). These opportunities are of particular importance for clients not aware of or unable to apply for certain public services due physical or mental disorders (Busch, 2018). E-government services are also used by third-parties through inquiries such as worries about work environments or public agencies alerting about unusual financial activities (Henriksen, 2018). Technology has lowered the bar required to ask for help or notify about worrying behavior, thus, shaping the cognitions and actions of its users (Orlikowski, 1992). Public service workers are enabled by technologies since they fulfill policy maker intentions by helping more clients.

Another aspect is information quality (De Witte et al., 2016; Larsson & Jacobsson, 2013). Public service workers can receive more correct information from clients and public agencies. Correct data input is important for decision-making (Henriksen, 2018). Forms can provide built-in guidance, forced use patterns, and control mechanisms to ensure correct data input. Searchable databases provide public service workers with relevant information previously difficult to obtain (e.g., De Witte et al., 2016; Keymolen & Broeders, 2011). Information is vital for public service provision since it constitutes the factual evidence upon which public service workers make their judgments. Without certain information, they are forced to assess a case according to their discretion creating more room for mistakes. Whereas technology can provide public service workers with more relevant information than before, they still have the autonomy to select and assess information (De Witte et al., 2016; Larsson & Jacobsson, 2013). In addition, technology can provide them with more critical information about their work (Kalu, 2001).

Several factors may explain incorrect or incomplete data such as public agencies storing and handling data from clients multiple times (Busch, 2018), forms difficult to understand (Scholta, 2017), fragmented phrases in different form boxes instead of narratives (Høybye-Mortensen, 2013), recollection of information in stressful life situations (Dawes & Helbig, 2015), and clients deliberately providing wrong information. Jorna and Wagenaar (2007) pointed out how data from systems can be incorrect requiring discretion. Thus, whereas technology is intended to support professional norms, they can also be impeded.

Technology has often been discussed related to the exercise of discretion in decision-making (Buffat, 2015; Busch & Henriksen, 2018). Discretionary power has been disputed since it can lead to incorrect, or even biased or corrupt decisions. Studies have shown that technology can reduce corruption and biases (e.g., Reddick et al., 2011; Smith, 2011). Wenger and Wilkins (2009) found that more women received social benefits when they used technological intermediaries rather than contacting public service workers directly. Snellen (2002) argued that the most important influence technology has on public service workers is the lack of ability to manipulate information streams. When cases are created either by clients or other third-parties, this camp of scholars look at how technology forces information into stringent forms that cannot be changed or manipulated in any way. Furthermore, technology is found to increase decision quality (e.g., De Witte et al., 2016; Larsson & Jacobsson, 2013), and promote more fair and uniform decision-making (e.g., De Witte et al., 2016; Pithouse et al., 2011).

Finally, technology can be used to control decision-making. Errors occur in public service provision and peer reviews can reduce the potential for wrong decision outcomes and increase client trust (e.g., Houston, 2015). The extent to which decisions made by public service workers are controlled varies. In some public agencies, every decision is controlled. Random controls of decisions by managers and peers are more common. Decisions made by independent public service workers such as judges cannot be controlled since the constitution prevents such actions.

Keymolen and Broeders (2011) discussed how the work of professionals and agencies within child care became more transparent due to features inherent in the technologies. This visibility could have both positive and negative effects; positive as it could lead to more reporting since none of the professionals want to be the one who missed the risk, and negative as professionals could resist this visibility for confidentiality reasons.

2.2 Decision factory workers: Technology as managerial imperative

Factory workers are people who, to a substantial extent, operate and supervise machines and technologies to accomplish certain predefined tasks. They use their skills to ensure that these machines function as smoothly and efficiently as possible. Decision factory workers resemble these workers. They operate computers and systems that streamline decision-making. In this perspective, characteristics of decision crafters typifying professional public service workers are influenced by various technologies, and thus, they perceive technology as an ‘action prescription’ (Hupe & Buffat, 2014; Orlikowski, 1992). Technology is utilized to promote managerial goals such as increased efficiency (Houston, 2015; Wihlborg et al., 2016), reduced costs (Pithouse et al., 2011; Reddick, 2005), and standardized routines to avoid policy discrepancies (Busch, 2018; Lipsky, 2010). Managerial concerns are characterized by accountability for performance goals such as efficiency and cost gains (Vigoda-Gadot & Meiri, 2008).

Based on the culture in an organization, technology can be used to standardize decision alternatives and automate decisions (Orlikowski, 1992). Based on the collected information, public service workers need to identify decision alternatives within policy boundaries. Technology for decision support has been extensively discussed in the literature, most prominently in the last century. Sheridan (1992) provided a 10-point scale describing the extent to which technology can support decision-making. At its lowest level, the human is autonomous, and the computer offers no assistance. At its highest level, the computer decides on actions without human interference.

Most often, routinized tasks have been automated in the public sector. For example, when students apply for student grants, the application process can be fully automated from the client’s announced need for a grant to a decision is ready. Standardization forces cases to be categorized into broader categories. The ability to adapt decisions to individual situations is reduced or even removed. From a top-down perspective, the advantage is that decision outcomes are more likely to be equal for similar situations and that public service workers have less opportunities to tweak decisions according to their own desires (Buffat, 2015). Wenger and Wilkins (2009) found that women were discriminated by rogue agents within employment services. Automating the process, more women received benefits while having no effect on men. Smith (2011) studied how increased automation was favored since the influence from personal factors such as mood and recent life events could be removed from the decision-making.

Another aspect of increased automation is automated controls. Technology has been used to support work processes devised by process law. To ensure that clients get fair and just treatments, technologies force certain use patterns and require specific tasks to be conducted from which the public service worker cannot deviate (Orlikowski, 1992). Scholars have recognized that technology is more appropriate for controlling formal aspects of decision-making rather than informal aspects (Buffat, 2015). Whereas technology can measure the number of completed cases and completion time, it is difficult to control underlying actions. For example, Jorna and Wagenaar (2007) showed that whereas technology provided increased managerial control over formal aspects such as the quantity of handled applications and the number of inconsistencies, it failed to capture informal dimensions of decision-making such as the interpretation of information and the content of applied standards. Wihlborg et al. (2016) found that such controls could lead to obscure discretionary practices instead of open practices.

Because of decision factoring, public service workers experience less ownership to decisions (“these are not my decisions”) and become distant from the outcomes of public services. The personal footprint is considerably reduced or even removed. Since they are heavily constrained in their actions, they experience deprofessionalization (Wihlborg et al., 2016) and demotivation (Tummers & Rocco, 2015). Instead of helping clients, they end up helping computers instead.

Technology can influence accountability by reducing judgment costs. Judgment costs are those efforts and pains that public service workers can expect if they make decisions that are unpopular or disfavor clients. For example, teachers make considerations about judgment costs when grading exams. Perhaps scores close to a better grade can be reconsidered to avoid complaints? Since public service workers can hide behind computers, clients face challenges with holding someone accountable for decisions. Not only can public service workers hide behind computerized decisions, Dennis (2006, p. 574) found that decision factories empowered public service workers since clients seem to question authority and decisions less than before: “I mean ‘the computer says what the computer says’ is the way we can present it and they seem less willing to protest against the outcomes”.

3. Research context and methods

The empirical foundation for this study is data collected in a Norwegian district court and a Norwegian tax administration office. The study was approved by the Norwegian Centre for Research Data. In the court, we interviewed judges and engaged in participant observations, and in the tax administration office, we interviewed caseworkers. The district court is responsible for thousands of cases annually regarding matters as diverse as child custody, theft, and corruption. Judges are independent and highly trusted, and their societal role makes the court interesting to study. The tax administration office handles cases on various tax matters for clients and subordinated to other public agencies. The office is divided into teams each taking care of specialized work tasks. The decisions that caseworkers make can be changed by the office manager as well as their peers during controls. The informants used different technological tools (Lovisa, SL, Law Data, and Court Data) to support decision-making as presented in Table 1.

Table 1
Decision-making tools

	Lovisa	SL	Law data and court data
Technology	Adaptive case management system.	Case management system.	Databases.
Technology aim(s)	Support procedural rules, handle workflow, and prevent unnecessary delays and errors.	Provide computerized controls, comprehensive information collection, and inter-agency collaboration.	Provide access to online legal resources such as legislation, decisions, and academic literature.
Policy area	Used in district courts and appeal courts.	Used in tax administration offices.	Used in a wide variety of public agencies conducting legal assessments.
Policy aim(s)	Ensure independent and fair assessments of clients’ cases.	Ensure safe and simple procedures for tax reporting and assessment.	Ensure high-quality foundations for decision-making.
Meta-level questions to answer	How should clients be judged based upon identified sequences of events, law, and previous practices?	How should tax matters of clients be assessed based on financial arrangements?	What does the law and academics say about legal matters, and how have other clients in similar situations been assessed?

Seven judges and nine caseworkers were interviewed face-to-face. Informants were selected based on theoretical sampling: they are screen-level bureaucrats who experience a major influence of technology on their decision-making practices (Bovens & Zouridis, 2002). This sample allowed us to study how commonly used technologies are utilized in various sub-stages of policy implementation. Informants were managers as well as employees in permanent and qualifying positions. Work experience varied from one year to several decades which made us able to grasp a variety of experiences. Informants were introduced to the study either by the scholar, or by their manager who requested their participation. Interviews were semi-structured and formulated with open-ended questions. The interviews were conducted during a period of nine months and varied in length from 40 to 100 minutes averaging 50 minutes. They were audio-recorded and transcribed verbatim using qualitative analysis software. Informants were given the opportunity to correct any errors in the transcribed text. The interview questions concerned topics such as how work tasks were structured, how information systems supported their work tasks though the various steps of policy implementation, and managerial control.

One scholar engaged in participant observations as a lay judge in four one-day trials during a period of two years observing how the trials were conducted in general, how information about the cases was collected, the use of the ECMS, and how a verdict was decided. The observed trials dealt with matters such as violence, misconduct, and drunk driving. After each trial, field notes capturing the essence of events and communication were written down. Key events observed included pre-trial meetings, trials, meetings during the trials, and post-trial meetings discussing final verdicts.

The analysis was guided by the views of public service workers as decision crafters or decision factory workers. Of special interest were manifestations of technological influence in the different sub-stages of policy implementation (see Fig. 1). In the first step of the analysis, the data were coded using codes close to the data. The detailed codes were assessed to find out whether they could be understood as manifestations of decision crafting or decision factory work. In the second step, we looked for relationships between the detailed codes which could consolidate them into higher-order categories. As the last step, the data were reassessed to ensure that no novel concepts emerged from the data, and that the analysis failed to reveal any new relationships between codes. The extracts presented in this article have been chosen to portray how public service workers use technology for various tasks in the different steps of policy implementation. Representations are translated into idiomatic English.

4. Technology in policy implementation: An empirical analysis

The court and the tax administration office investigated in this study have undergone a wide range of changes as results of e-government efforts spanning several years. In the court, these efforts are intended to ensure legal processes according to law requirements, efficient communication between actors in the court, digitization of paper documents, and archival of court testimonies. The tax administration is considered the “digitalization flagship” in Norway since they have automated tax reporting, implemented sophisticated business intelligence systems analyzing clients based on societal trends and other parameters, and shifted their focus from manual routinized tasks towards more analytical work. As a result of these efforts, the agency has achieved substantial increases in tax incomes while reducing operational costs.

Technology has significantly impacted the preparatory stages of policy implementation. Digital tools have been utilized in both case organizations to raise decision needs. In court, administrative staff can cooperate with police authorities and exchange information. When an indictment against a client is raised by the police, they can use their own systems and submit this information directly to the court instead of sending it through traditional mail services. These practices improve the quality and efficiency of their service since the required information is saved in the ECMS directly without being re-entered as previous practices required. The tax administration office has achieved similar gains. Clients can use online forms to raise awareness of issues that need assessment, for example, to inform about the use of a vehicle owned by its employer. Since such use is acceptable to a certain degree without tax deduction, this use needs to be clarified. Compared to older practices, the client does not need to contact a tax administration office to raise this issue. On other occasions, their own business intelligence systems may alert them about suspicious behavior requiring their attention. Information can also come from national or international police.

Furthermore, public service workers can collect more relevant information about a case using less time than before. By providing more relevant information, the quality of decisions increases since the collected information provides a better foundation for decision-making. The chief judge commented on this development:

Matheo: You get more information in each case. Through technology, we now have access to more legal sources than we had before when we had to go and look in heavy books. We even lacked access to some of the legal sources that we now have access to. So, technology influences us by providing a better basis for making decisions.

These gains are considerable. The work that judges do is a good example of how technology can improve decisions. In the court, printed books are almost never used to collect information anymore. Compared to books, databases store new verdicts and updated policies as soon as they are available. And thus, judges are provided with tools that help them consider the latest developments within a certain legal area. Public service workers in the tax administration have achieved similar gains. However, their practices are somewhat less prone for rapid changes. Thus, using a guide book may still be satisfying. A caseworker in the tax administration described how technological skills influenced how technology is used:

Lars: Much is based on self-learning, and when you grow old like me, things are not as simple as before. [ $\ldots$ ] Especially the SL system. It is extraordinary complicated. [ $\ldots$ ] But once you have learned what you need, it is okay.

Since technology to a significant extent is used to provide information from clients and practice, the result is that public service workers trust the information on the computer screen and do not look any further for other information. Moreover, when more information is retrieved, the need for discretion can be reduced. A caseworker in the tax administration office explained this phenomenon:

Trygve: We exercise discretion to find the most correct facts. But clearly, the more information we have, the better we manage to make a decision in terms of correctness. The less information we can collect, the more challenging it will be to get the decision right. So, facts limit the room for exercising discretion.

Another aspect of the information that technology can provide is directed at management, which is informed about the number of active cases, time spent on different types of cases. By using this information, resources may be reallocated:

John: [The system] provides information on case management status in relation to resources. Eh. more judges. More case managers.

Even though more information reduces the need for discretion, technology does not seem to influence their discretionary practices considerably. The influence is indirect and habitual, rather than direct and normative. The subtle influence is illustrated by a judge who acknowledged habitual effects of technology using templates for routine tasks:

Linda: The use of templates may reduce discretion [ $\ldots$ ] We base our decisions on the information in the template without exercising too much discretion [ $\ldots$ ] And that is a risk that we must be aware of.

However, one of the other judges described how she chose to write the verdict in full dismissing the template because she felt it was necessary in a specific case:

Hermine: It was a specific decision where I removed the template text and wrote it in full. I thought it was necessary. And then I got a call from one of the lawyers afterwards who thought it was very good that I had written more than just [ $\ldots$ ] because they observe that the same text is repeated in every case [ $\ldots$ ] So I realized that the dismissal of the template was noticed.

More information seems to be more influential for their discretionary practices. This influence is not considered negative, on the contrary, it is intentional. After all, discretion is constrained by policies, previous practices, and relevant information for the case. The less uncertainty, the less public service workers need to address this uncertainty. It is difficult for technology to influence discretion in traditional public service provision. The complexity of cases, the need clients have to interact with public service workers, and the potential severity of cases are factors that can explain why discretion is considered necessary. While the tax administration has automated tax reporting in general, caseworkers are more reluctant to automation in cases where clients have unusual circumstances:

Lars: You have to communicate with taxpayers, right? And assess the taxpayer’s written answers. It is not easy to automate the information that taxpayers provide. You have to [ $\ldots$ ] assess it against what is stated in the tax report and what the law says. Yes, that is how you do it.

With standardization comes increased categorization and less granulation. Therefore, more clients will be treated equally regardless of individual circumstances. This is challenging since not only must various factors be assessed, but also their individual weight. For these reasons, discretion is considered an instrument that can reduce unreasonable outcomes. One judge working specifically with matters of child care elaborated on this matter:

Sophie: But in cases of child care, there are many variables [ $\ldots$ ] eh [ $\ldots$ ] to be taken into consideration and they affect the extent to which you can detail a rule. Especially since different factors will have different weight.

Another judge elaborates on the case complexity that public service workers are faced with:

Matheo: Life comes in so many facets [ $\ldots$ ]. If you can exercise discretion, then a rule may be adapted, and the result will be correct.

He argues that rules can and should be adapted since it is impossible for legislators to make policies that can fit every situation. Thus, the use of technology is useful only as long as it can support necessary adjustments of policies to real-life situations.

Views on the use of technology to evaluate decisions vary between the two cases. The main reason for this is that judges are independent through the constitution. Because of that, no one can instruct a judge to sentence a certain verdict. Instead, the evaluation of their decisions can only be made through appeals to higher courts. One of the judges reflected on the difference between a previous position as legal advisor at the county governor and her current position as a judge:

Hermine: It is a paradox. I [ $\ldots$ ] eh [ $\ldots$ ] at the county governor every little decision was sent to the manager for approval. Here is nothing of it.

In the tax administration office, technology was widely used by managers and peers to control decisions. Their digital tools allow them to control parts of or whole cases. These practices served several purposes and had numerous benefits. Previously, tax assessments were made in municipal offices of which some were very small. Therefore, public service workers often knew their clients personally or at least through acquaintances. These relations caused problems and raised discussions about unfair treatment of clients. Now, technology allows them to control cases of clients who are located in other parts of the country. Moreover, the ECMS also feature controls where one public service worker assesses only one minor detail and another public service worker in a different part of the country assesses another minor detail. A caseworker in the tax administration reflected on how technology had changed their practices compared to several years ago:

Sara: It has become completely different. From paper reports assessing each record [ $\ldots$ ] until now where we do not look at anything more. The way tax reports are delivered today [ $\ldots$ ] it is only if a report is selected for control that you look at a record.

The public service workers in the tax administration conduct their controls without interfering with others’ work. Issues of corruption and biases are almost eliminated as a result. In addition, the tax administration has made use of sophisticated business intelligence systems that can capture criminal trends internationally and nationally in addition to detect suspicious behavior of clients. For example, if money laundering through restaurants are more common, more controls will be conducted of this type of companies. These practices were favored by most tax administration employees including the manager:

Trygve: Clearly, now we have much more structure [ $\ldots$ ] on this. So, we spend our time more efficiently. And much more uniform. [ $\ldots$ ] I cannot see anything but win-win in all areas.

Public service workers in both case organizations described the implementation of their decisions to be mostly independent of technology. They use technology to inform clients or third-party actors about decisions, but the actual implementation is not influenced. The chief judge explained how verdicts used to announce decisions are written up:

Matheo: The judge has a sentencing hearing, and there are some routines when the judge writes the verdict afterwards. The caseworker has pasted the indictment into the document, [ $\ldots$ ] and just written a little. But everything that concerns the court’s remarks and the case are written by the judge. So there aren’t any prescriptive guidance or anything.

Another aspect is accountability since technology has the potential to remove responsibility from the public service worker. For both judges and caseworkers, technology had no noteworthy influence on the opportunity to hold them accountable for their decisions. One of the judges discussed accountability in terms of judgment costs. In another part of Norway, a chief judge experienced a serious threat against his family. Luckily, no one was injured, but the outcome could have been worse:

Emil: The most serious example is in a case [ $\ldots$ ] where a chief judge got his house set on fire. [ $\ldots$ ] The judge had judged a childcare dispute. The daughter of the judge and a friend were sleeping in the basement and rescued by neighbors. So, a guy is charged for attempted murder of a judge and his family. And this is something that we have not seen in Norway before.

Whereas the incident is no argument for removing responsibility from a judge, the result can be reduced accountability if discretionary practices are influenced by technology.

5. Discussion and conclusion

The literature concludes that technology has been successful in terms of promoting both professional norms and managerial goals (Busch & Henriksen, 2018), in particular by taking over routinized work and leaving tasks that require analytical skills to humans (Cordella & Tempini, 2015). Furthermore, it points at the significance of studying the tasks at hand and the context various technologies are used within to understand their impact on the policy implementation process (Busch & Henriksen, 2018). In this study, we contribute with a better understanding of how technology impacts decision-making through seven sub-stages of policy implementation. We focus on whether technology serves professional aims or managerial goals to suggest if technology is shifting the professional imperative traditionally associated with public service provision (Lipsky, 2010).

5.1 Technological impact in policy implementation sub-stages

The use of performance goals was more widespread in the tax administration office than in the court. Performance goals involved monitoring the number of completed cases, completion time, and costs. Professional norms focused on how the quality of decisions can be improved by emphasizing the unique characteristics in each case and assessing them in comparison with similar cases. For screen-level bureaucrats, we found that the impact of technology on professional norms and managerial goals varies in each sub-stage of the policy implementation process. Whereas professional norms dominate in some of the sub-stages, managerial goals dominate in other sub-stages. In particular, the exercise of discretion seems to be difficult to influence by technology. Table 2 summarizes technological impact in each policy implementation sub-stage based on our empirical analysis and extant literature.

Table 2
Technological impact in policy implementation sub-stages

Policy implementation sub-stage	Problem addressed	Role of technology	Technological impact on professional norms	Technological impact on managerial goals
1. decision need occurs	Accessibility Ignorance	Online forms API Enlightenment	Positive impacts – accessibility – fairness Negative impacts – client interaction	Positive impacts – efficiency Negative impacts – increased costs
2. collect information	Inaccessible data Incorrect data	Data collection Inter-agency data sharing	Positive impacts – data availability – information quality Negative impacts – information errors	Positive impacts – efficiency – cost reductions Negative impacts – worker empowerment
3. identify decision alternatives	Identify decision boundaries	Data sharing	Positive impacts – data availability – information quality Negative impacts – habitual behavior	Positive impacts – efficiency – cost reductions
4. weigh evidence	Discretion	Adherence to policy maker intentions	Positive impacts – anti-corruption – prevention of biases and errors Negative impacts – worker demotivation	Positive impacts – efficiency – cost reductions – client focus – role change
5. choose decision alternative	Discretion	Adherence to policy maker intentions	Positive impacts – anti-corruption – prevention of biases and errors Negative impacts – worker demotivation	Positive impacts – efficiency – cost reductions – client focus – role change
6. evaluate	Incorrect decisions	Internal control	Positive impacts – decision quality – impartiality Negative impacts – worker demotivation	Positive impacts – efficiency – cost reductions – client focus – role change
7. implement action	Informing clients Inter-agency communication	Communication Inter-agency data sharing	Positive impacts – service quality Negative impacts – client interaction	Positive impacts – efficiency – cost reductions Negative impacts – accountability

In the first and second sub-stages, technology is utilized to a significant extent. Clients and third-parties can use websites and online forms to raise needs for decisive actions. Clients can apply for services such as tax reductions, positions in kindergartens for their children, and paternal leave. Third-parties can report concerns such as suspicious financial activities and childcare issues. Furthermore, public agencies can communicate with each other to inform about client decisions concerning other agencies. In field work, police officers may use technology to identify drivers overspeeding and doctors can be notified about low values on certain medical parameters requiring action. However, teachers are not notified about pupils with special needs nor do technology play any role when social workers discover elderly people living alone. From our analysis, we can see that managerial goals such as efficiency and client satisfaction have gained momentum in the first two sub-stages compared to previous practices. Technology use in these sub-stages is considerably discussed in the public management and e-government literature reflecting managerial goals.

Technology is, to a substantial extent, used to identify decision alternatives. Legal processes focus on delimiting decision alternatives based on current law and previous decisions. Digital tools assist public service workers by identifying the exact rule to apply in certain situations and present how the rule has been applied in other situations. Since they seem to trust the information the computer provides and do not look further for information, the technological influence becomes substantial. Our analysis suggests that technology reflects managerial goals to a large extent by turning the attention of public service workers towards efficiency gains, and away from their own considerations about appropriate legal sources.

Discretionary practices seem to be less influenced by technology for traditional public services such as judging and teaching. For other types of public services, in particular mass-transactional tasks (Bovens & Zouridis, 2002), technology has to a large extent reduced or eliminated discretionary practices. This development is illustrated through the shift from manual treatment of tax reports in the tax administration office to automated tax reporting. Other services have also been automated such as applying for and deciding on student grant loans (Wihlborg et al., 2016). These services are characterized by the need for numerical data and schematic rule sets. Data is often made available for government agencies either through inter-agency cooperation or mandated reporting from financial institutions and employers. Even though there are developments towards reducing and removing discretion, the influence on screen-level bureaucrats seem to be slightly increasing (Busch & Henriksen, 2018). Closer assessments about tax matters and infringements must be made by humans. Thus, discretionary practices seem to reflect a professional imperative. However, technological developments such as the use of algorithms for decision-support is already prevalent in US courts (Israni, 2017).

In the sixth sub-stage, technology can be used to evaluate decisions made by peers. This internal control is highly dependent on the status of the public service worker. In our study, decisions made by judges can only be changed through appeals to higher courts. In court, no peer can instruct a judge nor evaluate and change the decision he or she has made. In the tax administration office, peer review is common practice. Technology assists them in controlling decisions. Another benefit coming from the use of digital controls is combating impartiality and corruption. Public service workers can control decisions by colleagues about clients located in other parts of the country. This practice is so established that public service workers do not react negatively to it. In this sub-stage, technology reflects both a professional and a managerial imperative. Whereas technology constrains public service workers through inherent use patterns and limited options to influence others’ decisions, it still leaves the professional assessment of cases to public service workers.

In the final sub-stage of implementing actions according to decisions made, technology is most commonly supporting communicative tasks. It is first and foremost used to inform clients about decisions. Moreover, technology is used to inform other agencies about decisions since they may influence client rights and obligations. Other tasks within this sub-stage is not significantly influenced by technology. These tasks involve practical arrangements on public services. Technology is mostly used to achieve efficiency gains and its use does not impact professional norms negatively. Practical arrangements concerning how the decision will influence the clients still lay upon public service workers.

5.2 Technology as professional or managerial imperative?

The analysis presented in this study concerns the impact technology has on professional norms and managerial goals. Whereas technology can impact both professional norms and managerial goals as our findings hint at, we are particularly interested in revealing whether technology shifts public service provision from a traditional professional imperative to a managerial imperative. Acknowledging the limitations of the cases we have analyzed, we argue that they represent public service workers characterized by heavy workload, lack of resources, close interaction with clients, and the exercise of a substantial amount of discretionary power. The similarities characterizing much of public service work suggest that our findings are transferable to other types of public service workers given that they reside in a country like Norway with somewhat homogeneous demographics, and where services are organized uniformly, policies are applied consistently, and decision-making is supported by similar technologies.

Drawing on extant literature and our analysis, our findings suggest that technology has a significant influence in all sub-stages of policy implementation except for discretionary practices. Whereas certain sub-stages of policy implementation have been largely influenced by managerial goals, technology has supported professional norms in other sub-stages. Most often, both professional norms and managerial goals are influenced. The result is an alternation between a professional imperative, a combination of professional and managerial imperatives, and a managerial imperative. Figure 2 demonstrates this alternation in policy implementation sub-stages.

Figure 2.

Alternation between professional and managerial imperatives.

The alternation has several potential implications for present and future public service provision. First, in terms of how public service workers and clients experience decision quality, there is no inevitable conflict between professional norms and managerial goals. Both managers, public service workers, and clients can appreciate that professional norms are strengthened, and managerial goals are achieved. In fact, public service workers have a clear preference for helping clients (Tummers & Rocco, 2015) and whenever technology can help clients, they are satisfied regardless of professional norms and managerial goals. An important aspect of how clients experience policy implementation is related to fair decision-making. Our empirical analysis suggests that clients appreciate objectives that are associated with both professionalism and managerialism; experiencing respectful and caring encounters with public service workers as well as timeliness influenced their overall view of the policy implementation process.

Second, there is a significant potential for conflict caused by the dual nature of technology promoting both professional norms and managerial goals (Orlikowski, 1992). The modest influence on discretion can be explained by the lack of willingness public service workers have to reduce their discretionary power which they consider a vital prerequisite for exercising professional judgment (Busch & Henriksen, 2018). Discretion is considered vital because of the complexity of policy implementation where they need to deal with information uncertainty, unclear and vague legislation, and decisions with potentially severe consequences. Public service workers are concerned that technology can shift their attention towards managerial goals incommensurable with professional norms, and that their professional identity might be in jeopardy. This is important in policy implementation in general, but especially for judges in court. They are independent through the constitution and given the privilege of making assessments about cases based merely on law without political pressures. If technology creates a managerial imperative, it may have severe consequences. Judges need time to assess cases carefully. Moreover, if decision-making in court is automated, it would not make much sense because of the nature of the work in the courts; if a decision can be automated, there is no independent assessment of an impartial third-party.

Whereas a technological influence on discretionary practices is debated, its influence on other sub-stages of policy implementation seems to be less controversial. Here, technology can assist such as making communications with clients easier, taking care of routine tasks, and providing more accurate information to support sound decisions. The potential implications that can be derived from this technology use are that professional public service workers can shift their attention towards tasks that require their professional expertise and pay less attention to routine tasks. Moreover, such a shift would coincide with their motivations to work in public service provision perhaps leading to increased work motivation.

Third, system designers can now make choices previously made by public service workers (Bovens & Zouridis, 2002). They decide on considerations to be made and convert vague legal terms into algorithms and decision trees that can be decisive for policy implementation outcomes (Henriksen, 2018). This development raises one fundamental problem: system designers are not professionals. They have not spent years in training, neither have they first-handed experienced the many different fates of clients. By letting algorithms support decision-making, a clear shift towards a managerial imperative would be the result. However, more recently, public service design has received much attention in research focusing on taking considerations of both clients and public service workers into account aiming at co-designing and co-creating technologies. Doing so will enable professionals to specify necessary features of the technology to be implemented in public service work.

Finally, we can ask if technology is too easily available for certain tasks. If technology can create habitual practices, do this phenomenon mean that public service workers presume that there are no extraordinary circumstances in their particular case warranting their attention? Perhaps technology becomes such a helpful tool that what was meant as a support tool is institutionalized into a de facto standard instead?

5.3 Future research agenda

Our research context is limited to two types of public service provision in Norway. Thus, findings may reflect a specific culture. Future research should carry out cross-sectional studies spanning several countries, cultures, public services, and technologies.

This study has investigated digitized public services in two street-level bureaucracies that use similar types of technologies. E-government is not a unified term but rather an overall term for technologies ‘with different technical functions and capabilities, and as a consequence, different possibilities for influencing processes and structures in the public sector’ (Lips & Schuppan, 2009, p. 742). Further research can benefit from studying the impact of these varying types of technologies including technological innovations. For example, artificial intelligence, bots, self-service automats, databases, and websites (Agarwal, 2018). Handheld devices such as tablets and smartphones should be of interest. For example, whereas it is common for several groups of public service workers to make decisions “on-the-spot”, extant knowledge is scarce about how technology can influence this type of decision-making. Such developments are suggested and tried out in areas such as sports but to a less extent in the public sector. For example, in the 2018 football world cup, a video assistant referee (VAR) was introduced where a team of referees could suggest another decision alternative for the main referee on the pitch. In policy implementation, groups of public service workers such as police officers and social workers often make decisions out on the streets, and they can do so without technical support. If a police officer stops a citizen for public misbehavior, he or she must decide there and then whether to issue a fine or not. Often operating alone, future research endeavors should investigate how technology can assist public service workers on the move such as through the use of body-worn cameras by police officers in the United States and Canada (Bromberg et al., 2018).

Even more, the impact technologies have on professional norms and managerial goals seems to be influenced by public service workers’ technological skills and training. Public service workers that make better use of technological features will also more likely be influenced by the habitual effect of technology. It could be that public service workers with higher computer self-efficacy have more knowledge about what technology can offer and not offer, and therefore are more likely to favor it, especially in sub-stages of policy implementation where their discretionary power is not influenced. Thus, future research should investigate how technological skills and training influence the impact of technology in policy implementation.

Finally, whereas this study took a public service worker perspective, another promising area for further research is to investigate how clients react to crafted vs. mass-produced decisions. The premise for lot of research within this area is that decision crafting is necessary and inevitable for ensuring just and fair outcomes. If clients are satisfied with mass-produced decisions, perhaps this research has focused too much on the public service worker perspective?

Footnotes

Stage models of the policy process have been criticized for over-simplifying a more complex and messy reality (e.g., Sabatier, 2007). Whereas these claims are not without merit, stage models still serve as useful analytical frameworks to organize studies of public policies.

Conflict of interest

The author declares no potential conflicts of interest with respect to the research, authorship, and publication of this article.

References

Agarwal

P. K

. (2018). Public Administration Challenges in the World of AI and Bots. Public Administration Review, 78(6), 917-921.

Anderson

J. E

. (1975). Public Policy-Making. New York, NY: Praeger.

Bovens

., & Zouridis

. (2002). From Street-Level to System-Level Bureaucracies: How Information and Communication Technology Is Transforming Administrative Discretion and Constitutional Control. Public Administration Review, 62(2), 174-184.

Bromberg,

D. E

. Charbonneau

., & Smith

. (2018). Body-Worn Cameras and Policing: A List Experiment of Citizen Overt and True Support. Public Administration Review, 78(6), 883-891.

Buffat

. (2015). Street-Level Bureaucracy and E-Government. Public Management Review, 17(1), 149-161.

Busch

P. A

. (2018). Conceptualizing Digital Discretion Acceptance in Public Service Provision: A Policy Maker Perspective. In Proceedings of the 22nd Pacific Asia Conference on Information Systems (PACIS). Yokohama.

Busch

P. A

., & Henriksen

H. Z

. (2018). Digital Discretion: A Systematic Literature Review of ICT and Street-Level Discretion. Information Polity, 23(1), 3-28.

Cordella

., & Tempini

. (2015). E-Government and Organizational Change: Reappraising the Role of ICT and Bureaucracy in Public Service Delivery. Government Information Quarterly, 32(3), 279-286.

Dawes,

S. S

., & Helbig,

. (2015). The Value and Limits of Government Information Resources for Policy Informatics. In Johnston

E. W.

(Ed.), Governance in the Information Era: Theory and Practice of Policy Informatics. London, UK: Routledge, pp. 25-34.

10.

De Witte,

. Declercq

., & Hermans

. (2016). Street-Level Strategies of Child Welfare Social Workers in Flanders: The Use of Electronic Client Records in Practice. British Journal of Social Work, 46(5), 1249-1265.

11.

Dennis

M. R

. (2006). Proletarian or Promethean? Impacts of Automation and Program Integration on Social Service Workers and Their Clients. Journal of Contemporary Ethnography, 35(5), 552-582.

12.

Dror

. (1989). Public Policy Making Reexamined (2nd ed.). New Brunswick, NJ: Transaction Publishers.

13.

Freidson

. (2001). Professionalism, the Third Logic: On the Practice of Knowledge. Chicago, IL: University of Chicago Press.

14.

Henriksen,

H. Z

. (2018). One Step Forward and Two Steps Back: E-Government Policies in Practice. In Gil-Garcia

Pardo

, & Luna-Reyes

(Eds.), Policy Analytics, Modelling, and Informatics (Vol. 24, pp. 79-97). Cham: Springer.

15.

Hill

., & Hupe,

. (2014). Implementing Public Policy: An Introduction to the Study of Operational Governance (3rd ed.). London: SAGE.

16.

Hogwood,

B. W

., & Gunn,

L. A

. (1984). Policy Analysis for the Real World. Oxford: Oxford University Press.

17.

Houston,

. (2015). Reducing Child Protection Error in Social Work: Towards a Holistic-Rational Perspective. Journal of Social Work Practice, 29(4), 379-393.

18.

Hupe

., & Buffat

. (2014). A Public Service Gap: Capturing Contexts in a Comparative Approach of Street-Level Bureaucracy. Public Management Review, 16(4), 548-569.

19.

Høybye-Mortensen

. (2013). Decision-Making Tools and Their Influence on Caseworkers’ Room for Discretion. The British Journal of Social Work, 45(2), 600-615.

20.

Israni,

E. T

. (2017 Oct. 26th). When an Algorithm Helps Send You to Prison. New York Times. Retrieved from https://www.nytimes.com/2017/10/26/opinion/algorithm-compas-sentencing-bias.html.

21.

John

. (1998). Analyzing Public Policy. London: Pinter.

22.

Jorna

., & Wagenaar

. (2007). The ‘Iron Cage’ Strengthened? Discretion and Digital Discipline. Public Administration, 85(1), 189-214.

23.

Kalu

K. N

. (2001). Leadership and Discretionary Decision-Making in a Technocratic Administration: Confronting a New Praxis. Administrative Theory & Praxis, 23(3), 311-336.

24.

Keymolen

., & Broeders

. (2011). Innocence Lost: Care and Control in Dutch Digital Youth Care. British Journal of Social Work, 43(1), 41-63.

25.

Larsson

., & Jacobsson,

. (2013). Discretion in the “Backyard of Law”: Case Handling of Debt Relief in Sweden. Professions and Professionalism, 3(1), 1-17.

26.

Lips,

A. M. B

., & Schuppan

. (2009). Editorial: Transforming E-Government Knowledge Through Public Management Research. Public Management Review, 11(6), 739-749.

27.

Lipsky

. (2010). Street Level Bureaucracy: Dilemmas of the Individual in Public Service (2nd rev. ed.). New York, NY: Russell Sage Foundation.

28.

Noordegraaf

. (2007). From “Pure” to “Hybrid” Professionalism: Present-Day Professionalism in Ambiguous Public Domains. Administration & Society, 39(6), 761-785.

29.

Orlikowski

W. J

. (1992). The Duality of Technology: Rethinking the Concept of Technology in Organizations. Organization Science, 3(3), 398-427.

30.

Peckover

. White

, & Hall,

C. J

. (2008). Making and Managing Electronic Children: E-Assessment in Child Welfare. Information, Communication & Society, 11(3), 375-394.

31.

Pithouse,

Broadhurst,

Hall

C. J.

Peckover

Wastell

, & White

. (2011). Trust, Risk and the (Mis)Management of Contingency and Discretion through New Information Technologies in Children’s Services. Journal of Social Work, 12(2), 158-178.

32.

Powell

W. W

. (1987). Review Essay: Explaining Technological Change. American Journal of Sociology, 93(1), 185-197.

33.

Reddick

. (2005). Citizen Interaction with E-Government: From the Streets to Servers? Government Information Quarterly, 22(1), 38-57.

34.

Reddick

Abdelsalam,

H. M.

, & Elkadi,

. (2011). The Influence of E-Government on Administrative Discretion: The Case of Local Governments in Egypt. Public Administration & Development, 31(5), 390-407.

35.

Sabatier

P. A

. (2007). Theories of the Policy Process (2nd ed.). Boulder, CO: Westview Press.

36.

Scholta,

. (2017). The Same but Still Different: Forms in E-Government. In Proceedings of the 50th Hawaii International Conference on System Sciences (HICSS) (pp. 2559-2568). Big Island, HI.

37.

Scott,

., & Golden,

. (2009). Understanding Net Benefits: A Citizen-Based Perspective on E-Government Success. In Proceedings of the 30th International Conference on Information Systems (ICIS). Phoenix, AZ.

38.

Sheridan

T. B

. (1992. Telerobotics, Automation, and Human Supervisory Control. Cambridge, MA: MIT Press.

39.

Smith

M. L

. (2011). Limitations to Building Institutional Trustworthiness through E-Government: A Comparative Study of Two E-Services in Chile. Journal of Information Technology, 26(1), 78-93.

40.

Snellen

. (2002). Electronic Governance: Implications for Citizens, Politicians and Public Servants. International Review of Administrative Sciences, 68(2), 183-198.

41.

Tummers

., & Rocco

. (2015). Serving Clients When the Server Crashes: How Frontline Workers Cope with E-Government Challenges. Public Administration Review, 75(6), 817-827.

42.

Vigoda-Gadot

., & Meiri

. (2008). New Public Management Values and Person-Organization Fit: A Socio-Psychological Approach and Empirical Examination Among Public Sector Personnel. Public Administration, 86(1), 111-131.

43.

Wenger

J. B

., & Wilkins

V. M

. (2009). At the Discretion of Rogue Agents: How Automation Improves Women’s Outcomes in Unemployment Insurance. Journal of Public Administration Research and Theory, 19(2), 313-333.

44.

Wihlborg

Larsson

., & Hedström

. (2016). “The Computer Says No!” – a Case Study on Automated Decision-Making in Public Authorities. In Proceedings of the 49th Hawaii International Conference on System Sciences (HICSS) (pp. 2903-2912). Koloa, HI.

Crafting or mass-producing decisions: Technology as professional or managerial imperative in public policy implementation

Abstract

Keywords

1. Introduction

2. Technology in policy implementation: Decision crafters vs. decision factory workers

2.2 Decision factory workers: Technology as managerial imperative

3. Research context and methods

Table 1 Decision-making tools

5. Discussion and conclusion

5.1 Technological impact in policy implementation sub-stages

Table 2 Technological impact in policy implementation sub-stages

Footnotes

Conflict of interest

References

Table 1
Decision-making tools

Table 2
Technological impact in policy implementation sub-stages