Reconceptualizing Administrative Burden Around Onerous Experiences

ABSTRACT

Despite ongoing discussions on the need to improve the conceptualization and measurement of administrative burden, several conceptual problems remain. This study offers the first systematic analysis and evaluation of this increasingly central public management concept. Using an ontological-semantic approach, I show that the current conceptualization fails to fully and directly account for individuals’ onerous experiences. I address five interrelated issues, including the overlap of cost categories and the conflation of state actions with onerous experiences. While psychological costs should be retained, I argue for abandoning the other cost categories. Building on previous reconceptualization efforts, I propose a new framework focused on time, money, effort, and psychological costs. Additionally, I explore the structure of the concept and propose specific indicators for each dimension. I then discuss the independence of these dimensions, their capacity to reflect the distributive nature of burdens, and avenues for empirical validation.

INTRODUCTION

In the broadest sense, administrative burden refers to the costs individuals experience when interacting with the state. Public administration scholars have shown a growing interest in this phenomenon (Daigneault, Defacqz, and Dupuy 2024; Halling and Bækgaard 2024) and this literature examines various themes, including the political or administrative origins of administrative burdens (Herd and Moynihan 2018; Peeters 2020); accessing social benefits (e.g. Daigneault and Macé 2020; Heinrich et al. 2022; Negoita, Levin, and Paprocki 2022); exercising democratic rights (Herd and Moynihan 2018); psychological well-being (Baekgaard et al. 2021); the unequal impact of administrative burdens (Bell, Christensen, and Hansen 2023; Tiggelar and George 2023); citizens’ reaction to and negotiation of administrative burdens (Bell, Christensen, and Hansen 2022; Döring and Madsen 2022; Nielsen and Nielsen 2024); and policy feedback (Dupuy and Defacqz 2022; Halling et al. 2022).

With this growing interest, there is also a strand of the literature focused on improving the conceptualization and measurement of administrative burden (Bækgaard and Tankink 2021; Campbell, Pandey, and Arnesen 2023; Madsen, Mikkelsen, and Moynihan 2022; Nisar and Masood 2023). Various scholars have examined the shared and distinctive attributes of administrative burden in relation to neighboring concepts such as red tape and sludge (Campbell, Pandey, and Arnesen 2023; Madsen, Mikkelsen, and Moynihan 2022). Others have contributed to our understanding of the cost categories that currently constitute administrative burden (Baekgaard and Tankink 2021; Nisar and Masood 2023). There have also been calls for and efforts toward developing better indicators of administrative burden (Bækgaard and Tankink 2021; Jilke et al. 2024). However, problems persist, in particular around the conceptual specification of the cost categories. Currently, the cost categories of administrative burden are inadequate; they overlap and do not directly refer to individuals’ onerous experiences, instead focusing on state actions. These issues undermine the analytical traction of the administrative burden framework and hinder the development of cumulative knowledge.

In this article, I address and elaborate on these issues. I first review how scholars have defined administrative burden and examine the conceptual problems identified. I then introduce the ontological-semantic framework guiding my analysis and evaluation of administrative burden. Building on these two sections, I discuss five interrelated problems with the current conceptualization of administrative burden and its measurement, advocating for the abandonment of all cost categories except psychological costs. Next, drawing on Nisar and Masood (2023), I propose an alternative categorization based on time, money, effort, and psychological costs. I discuss concept structure, potential indicators, and the operationalization of citizen-state interactions. In the discussion, I address three key issues: the independence of the alternative cost categories, their capacity to capture the unequal distribution of administrative burden, and avenues for empirical validation.

THE ORIGINAL ADMINISTRATIVE BURDEN CONCEPT

In their foundational article, Burden et al. (2012, 741) defined administrative burden as “an individual’s experience of policy implementation as onerous.” This original definition was later enriched by an influential classification that unpacked the concept into three types of costs (Herd and Moynihan 2018; Herd et al. 2013; Moynihan, Herd, and Harvey 2015). Learning costs involve acquiring information about public programs, including understanding their nature, benefits, eligibility criteria, and the process for receiving them. While the nature of these costs was implicit in earlier definitions, Herd and Moynihan (2018, Table 1.1, 23) clarified that they involve time and effort. Compliance costs encompass meeting administrative rules, including form completion and providing documentation for eligibility. These may involve both direct (e.g. fees) and indirect (e.g. travel costs) financial burdens (Herd and Moynihan 2018, Table 1.1, 23). Psychological costs include stigma, stress, power loss, and frustration (Herd and Moynihan 2018, 23–9).

There have been two subsequent additions to these cost categories. Redemption costs, defined as a subset of learning costs, refer to the burden of using or attempting to use public benefits (Barnes 2020; Barnes and Riel 2022). For instance, in the US Special Supplemental Nutrition Assistance Program for Women, Infants, and Children, eligible individuals must navigate which products are covered under program rules and locate participating grocery stores. However, using vouchers can be challenging due to inconsistencies in store adherence to program rules and potential mislabeling of eligible items. Correction costs¹, on the other hand, refer to the burdens citizens face when attempting to correct administrative errors (Widlak and Peeters 2020). This process poses significant challenges for citizens in terms of learning, compliance, and psychological costs (Holler et al. 2024; Holler and Tarshish 2022; Widlak and Peeters 2020). For instance, Widlak and Peeters (2020) described a case where a woman’s car was stolen in the Netherlands. Over a 16-year period, she encountered high learning costs in identifying the initial police error’s source and understanding its impacts on other administrative databases and services. There were also substantial compliance costs, such as settling debts through expensive legal procedures, and psychological costs, including stress and fears about losing parental authority.

Herd and Moynihan (2018, 3–12) argued that administrative burdens are pervasive, distributive (i.e. hurting some individuals more than others), constructed (i.e. resulting from political and administrative decisions), and consequential, impacting the take-up of social benefits, immigration, and political participation. The distributive nature of administrative burden requires two clarifications. Means-tested programs are, by design, more burdensome than universal programs. Moreover, disadvantaged individuals have fewer (e.g. financial, cognitive, and social) resources that help to mitigate burdens. As Herd and Moynihan (2018, 8) explained, “…not only are policies targeted at the poor more burdensome, but the poor are also more likely to experience government as routinely burdensome.” The administrative burden framework captures variations in the costs experienced by individuals (Herd and Moynihan 2018, 22), highlighting its relevance to a growing number of scholars.

The original definition of administrative burden is experiential—it directly focuses on the costs individual experiences rather than state actions (Bækgaard, Tankink 2021; Herd and Moynihan 2018, 22; Madsen, Mikkelsen, and Moynihan 2022). Experiences are conscious events that individuals go through and to which they ascribe meaning; they depend on “how individuals interpret the world” (Holler and Tarshish 2022, 2). There is another, though less influential, conception of administrative burden: the impositions definition (Madsen, Mikkelsen, and Moynihan 2022). This definition is not directly experiential as it focuses on “the interactions with government that impose (or lessen) burdens on individuals and organizations” (Heinrich 2018, 216).

This definitional divide creates conceptual confusion that may hinder the production of cumulative knowledge. Indeed, different scholars claim to be studying administrative burdens while examining entirely different phenomena (Baekgaard and Tankink 2021). Individual’s onerous experiences differ from state actions, defined as the “rules, regulations, and other actions made by the state” (Baekgaard and Tankink 2021, 17). Despite many scholars claiming an experiential perspective, their indicators solely measure state actions. When these scholars study the relationship between state actions and citizen outcomes (e.g. accessing social benefits), they treat individuals as a “black box” (Baekgaard and Tankink 2021; Halling and Baekgaard 2024). This approach obscures the multifaceted and subjective nature of individual experiences, overlooking significant variations in how state actions affect different individuals. For instance, individuals with cognitive limitations or low administrative literacy face higher costs than the average citizen (Döring and Madsen 2022; Negoita et al. 2024), and are more likely to engage with public services while being less equipped to manage those burdens (Christensen et al. 2020; Herd et al. 2023).

Alongside these variations in individual experiences, Baekgaard and Tankink (2021) also highlighted potential issues with the current cost categories. The recent addition of redemption and correction costs may suggest that the current classification is incomplete and in need of further development. Furthermore, the distinctions between the current costs are not always clear-cut, leading to potential overlaps and ambiguity. Echoing Moynihan, Herd, and Harvey (2015), Baekgaard and Tankink (2021) noted that learning, compliance, and psychological costs are difficult to separate empirically. Similarly, Nisar and Masood (2023) argued that the cost categories are not mutually exclusive, making them difficult to disentangle in practice. Nisar and Masood (2023) further noted that the actual categories do not allow individuals to pursue trade-offs between different costs. For example, individuals cannot exchange learning costs for psychological costs. Therefore, Nisar and Masood (2023) proposed replacing learning and compliance costs with financial and time costs, which should also facilitate trade-offs and comparisons between policies and programs.

These studies contributed to our understanding of administrative burden by highlighting continuing problems with the specification of the concept. These problems underscore the need for a systematic analysis and reformation of the concept. Indeed, the original framework loses some of its utility if we cannot independently measure each cost. Moreover, focusing on state actions rather than individual experiences hinders our ability to study variations in perceptual experiences. This issue is particularly critical for understanding how individual characteristics such as poor health or low human capital impact administrative burden (Bækgaard and Tankink 2021; Bell et al. 2023; Christensen et al. 2020; Halling and Bækgaard 2024). For instance, Bell et al. (2023) found that individuals suffering from physical pain, anxiety, depression, and attention disorders experience higher costs than others, and were more likely to see their access to social benefits curtailed. This finding underscores how “administrative arrangements that are nominally equal in their design and implementation can result in unequal outcomes because of health differences” (Bell et al. 2023, 385). Similarly, Döring’s (2021) hypothesis that higher levels of administrative literacy (i.e. the independent variable) decrease learning costs (i.e. the dependent variable) illustrates these two critical problems. First, if the cost categories overlap, our indicators for learning costs (the dependent variable) may inadvertently include aspects of compliance and psychological costs. This contamination could distort our understanding of the true impact of administrative literacy on learning costs. However, if we measure learning costs by focusing on state actions (i.e. rules and requirements) rather than onerous experiences, we might fail to capture variations in the dependent variable within specific contexts. This limitation would prevent us from effectively testing whether individuals with differing levels of administrative literacy genuinely experience varying levels of administrative burden. Hence, there is a pressing need to develop clearer cost categories and indicators that directly and comprehensively reflect individuals’ experiences. Enhanced conceptualization and valid indicators can also promote scholarly communication and much-needed comparative research on administrative burdens, thereby ensuring the cumulative advancement of scholarly knowledge (Halling and Bækgaard 2024; Nisar and Masood 2023).

THEORETICAL FRAMEWORK

The conceptualization approach in this article emphasizes making conscious, explicit, and informed choices about concepts (Sartori 1970; Saylor 2013; Sartori 2009 [1984]). It espouses an “ontological-semantic” perspective (Goertz 2020), which prioritizes eliciting a concept’s meaning over its measurement. While measurement remains crucial in this approach, valid inferences rely on robust and well-defined concepts. Concepts, as described by Sartori (1970), serve as data containers that enable scholars to delineate what is included and what is excluded. This approach stands in contrast to the latent-variable model, where conceptualization and measurement are not distinct. The latent-variable approach is premised on an unobserved variable causing the indicators:

The ontological-semantic approach does not have causal relationships within the basic framework. […] In the basic framework, the focus is first and above all on the definitions and then secondarily on the empirical indicators; for latent variables, the focus is on “measurement models.” (Goertz 2020, 5).

From an ontological-semantic perspective, social science concepts are understood as multilevel and multidimensional (Goertz 2020). At the basic level, concepts are employed in descriptive and causal propositions, such as the hypothesis that right-wing policymakers exhibit greater tolerance for administrative burdens in social welfare programs compared to left-wing policymakers (Bækgaard et al. 2020). The secondary level encompasses the defining attributes or constitutive dimensions of the concept (Goertz 2020; see also Saylor 2013). Many social science concepts are inherently multidimensional, like Perry’s (1996) concept of public service motivation, which includes commitment to the public interest, compassion, self-sacrifice, and attraction to public policymaking. Finally, at the indicator-data level, concepts are operationalized through empirical referents corresponding to each constitutive dimension.

Concept (re-)construction involves three fundamental operations (Bélanger and Daigneault 2021; Goertz 2020; Saylor 2013). The first is to elicit the meaning of multidimensional concepts by selecting their fundamental attributes from the literature. These fundamental attributes must accurately describe real-world phenomena and be collectively exhaustive, covering concepts’ entire theoretical expanse (Goertz’s [2020] completeness rule). While they do not need to be mutually exclusive (in contrast to typologies in the classical approach, e.g. Halas 2021, Sartori 1970), they should be conceptually distinct (i.e. their semantic content should not overlap). Nevertheless, fundamental attributes can be empirically related. The quality-of-life concept proposed by the Organisation for Economic Co-operation and Development (OECD) is a case in point (Goertz 2020, 33), encompassing eight fundamental attributes, including health status and education. Whereas education strongly influences health status (e.g. Albert and Davia 2011), these concepts remain distinct. Moreover, the core dimensions of a concept should be selected for their causal powers (Goertz 2020), indicating their ability to explain the relationships between the concept and other concepts (e.g. psychological costs and non-take-up of social programs).

The second operation involves identifying indicators for each core dimension. Core dimensions refer to the fundamental nature of a concept—what it really is—while indicators capture its observable manifestations. Decisions about the choice and number of indicators are generally less theory-laden and more focused on methodological rigour and feasibility, although they must align with the concept’s ontology (Saylor 2013). Thus, indicators should directly measure the “essence” or “true nature” of a dimension. For instance, physical health should be measured using direct indicators like blood pressure or resting heart rate, rather than indirect ones that reflect its determinants (e.g. smoking, dietary habits) or consequences (e.g. absenteeism due to illness).

The third operation involves determining concept structure—that is, how dimensions and indicators are logically connected. Conceptual maps are essential to identify assumptions and problems (Goertz 2020, 38). Though rarely discussed in public administration (but see Carboni et al. 2019), concept structure and aggregation decisions significantly impact measurement. They determine how dimensions and indicators are associated to calculate the overall score or level of basic concepts, such as administrative burden. There is an extensive array of options available for concept structure, but according to Goertz (2020), there are primarily three main choices. The “weakest-link” structure relies on necessary and jointly sufficient conditions (logical “AND”), where the overall score is determined by the minimum dimension score. Conversely, the “best-shot” structure, employing logical “OR,” assumes perfect substitutability between dimensions, with the overall score determined by the maximum. The third option calculates the overall score by averaging scores across all dimensions. Thus, in this third option, the overall score of the concept is derived from the average scores of all dimensions.

ASSESSING THE ORIGINAL CONCEPT

I apply the above framework to the original administrative burden. I first map the original concept. I analyze its basic, secondary, and indicator-data levels following Goertz’s (2020) conventions for conceptual analysis (see figure 1). The original concept comprises three constitutive dimensions—learning, compliance, and psychological costs—consistent with the literature (e.g. Christensen et al. 2020; Madsen, Mikkelsen, and Moynihan 2022; Moynihan, Herd, and Harvey 2015). Figure 1 leaves the number and content of indicators for each dimension undefined to accommodate the diverse measurement strategies found in the literature. Following Goertz (2020), question marks (“?”) indicate unresolved decisions regarding concept structure. This underscores the ongoing discussion on how dimensions combine to constitute the basic level of administrative burden, as well as the choice of logical operators. Implicitly, scholars who measure each dimension independently and aggregate them typically employ methods like averaging or factor analysis (e.g. Bell et al. 2023; Fox, Feng, and Reynolds 2023; Johnson and Kroll 2021). It is essential to differentiate concept structure from the empirical relationships between dimensions, such as causality between different types of costs.

Evaluating the Original Concept and its Measurement

I contend that administrative burden, as currently conceptualized and measured, inadequately recognizes and captures individuals’ onerous experience for five reasons: (1) learning and compliance costs conflate state actions with onerous experience; (2) the “proxy approach” leads to inconsistencies between concept and measurement; (3) psychological costs overlap with other cost categories; (4) redemption and correction costs are also problematic; and (5) the cost categorization is based on an implicit stage logic.

Learning and Compliance Costs Conflate State Actions with Onerous Experience

Herd and Moynihan (2018, 22) argued that the three cost categories are useful because they distinguish between state actions and individual experiences. However, I argue the conceptualization of learning and compliance costs is imprecise and inadequate. It conflates the sources of burdens (i.e. state actions) with the burdens themselves (i.e. onerous experience). The characterization of Herd and Moynihan (2018, 23) suggest that learning and compliance costs are defined more by the tasks or activities individuals perform (e.g. searching for information about programs, queuing, traveling to government offices, filling out forms) than by the costs generated by these tasks and activities. Moreover, time, money, and effort are not fully articulated in their definition. Although learning costs mention time and effort, they overlook monetary aspects. For instance, learning about programs and eligibility rules may entail financial costs, such as hiring professionals for assistance with immigration programs. Conversely, compliance costs are described solely in terms of money, disregarding the substantial time and effort often required to fulfill compliance tasks.

The conceptual distinction between state actions and onerous experiences matters theoretically and empirically. As Madsen and Mikkelsen (2022) explain:

…[W]hile learning and compliance costs will often correlate with actual burdensome rules and demands, an important contribution of the concept is shifting the focus from rules to citizens’ experiences of rules (Herd and Moynihan 2018; Madsen et al. Forthcoming). […] This inserts a conceptual distinction between the state enacted policy and the individual’s costs from facing this policy. From an administrative burden standpoint, the same rule may impose significant learning or compliance costs only on some citizens, for instance owing to their stock of human capital (Christensen et al. 2020). (242)

For example, social workers, due to their professional expertise, may experience fewer burdens when gathering information about social benefits online compared to other citizens (Madsen, Lindgren, and Melin 2022). Furthermore, conflating state actions with onerous experiences can impede effective communication among scholars. Ray, Herd, and Moynihan (2022, 139) defined “racialized burdens” in terms of administrative practices rather than the individual costs incurred. Thus, focusing exclusively on state actions overlooks the variations in subjective experiences that arise from these actions. This is not a moot point: the added value of the administrative burden framework for scholars lies in its ability to highlight the distributive implications of state actions. Disadvantaged individuals are more likely to interact with the state in order to access public services and social benefits targeted at them, which, by design, tend to be more burdensome. Clients of these programs may, on average, be less equipped to handle these costs (Daigneault 2023; Herd et al. 2023). This, in turn, may reinforce inequality, creating a “catch-22” where those who need public services and social benefits the most are the least likely to access them (Christensen et al. 2020; Herd et al. 2023).

The Proxy Approach

The conceptual confusion between state actions and onerous experience discussed above also matters for empirical research. Again, many scholars measure administrative burdens by focusing on state actions rather than onerous individual experiences (Baekgaard and Tankink 2021; Halling and Bækgaard 2024). I call this indirect measurement strategy the “proxy approach.” Whereas state actions generally incur costs for citizens, the proxy approach generates construct validity problems (Adcock and Collier 2001; Saylor 2013). There is a gap between what scholars claim to study and what they measure in practice, with a case in point being Nicholson-Crotty, Miller, and Keiser (2021) study. Nicholson-Crotty, Miller, and Keiser (2021) defined administrative burden experientially, yet their indicators focused on state actions. Specifically, they examined the rules and requirements imposed by public authorities and how individuals complied with these (e.g. providing documents). Similarly, Fox, Feng, and Reynolds (2023) assessed administrative burdens associated with major safety-net programs by focusing on the rules and requirements imposed on citizens, thereby overlooking individual experiences of burden. To their credit, the authors acknowledged the inconsistency between the experiential definition of administrative burden and their indirect indicators, providing justification for their approach.

The proxy approach may be necessary because scholars do not have access to experiential data, but it remains suboptimal. This approach ignores individuals’ subjective interpretations of their experiences, thereby undermining the analytical traction of the administrative burden framework. It may also compromise empirical research on the relationship between individual characteristics (e.g. administrative literacy, political ideology, prior experience with the program, personality traits) and administrative burden because it does not measure the actual experience of individuals (see Aarøe et al. 2021; Baekgaard et al. 2020; Döring 2021). By measuring state actions rather than an individual’s onerous experience, we miss relevant variations on the dependent variable and potentially underestimate certain causal relationships. This underscores the importance of a measurement strategy that recognizes the experiential and distributive nature of administrative burdens.

Psychological Costs Overlap with Other Cost Categories

Psychological costs conceptually overlap with learning and compliance costs, a point that echoes previous criticisms of the framework (Baekgaard and Tankink 2021; Nisar and Masood 2023). The literature on psychological costs provides minimal guidance on their nature—specifically, how many exist and how they interrelate. However, stress, anxiety, confusion, anger, stigma, frustration, and lack of autonomy are identified as direct costs experienced by individuals (see Bækgaard et al. 2021; Hattke, Hensel, and Kalucza 2020; Herd and Moynihan 2018). Psychological costs, therefore, align with the experiential nature of administrative burden. Learning and compliance costs, by contrast, are higher-order categories situated at a different level of analysis than psychological costs. They represent more general conceptual categories that conflate specific activities associated with citizen-state interactions (e.g. searching for information, filling out forms) with the direct costs arising from those activities (i.e. time, money, effort, and psychological costs). Thus, learning and compliance costs are considered burdens indirectly, examining them through cost-generating tasks. To be clear, this point is not premised on the empirical association between these categories found in the literature (e.g. Holler and Tarshish 2022; Moynihan, Herd, and Harvey 2015; Ritzel et al. 2020). The argument is, rather, that psychological costs are constitutive of the higher-order categories of learning and compliance costs, reinforcing the need to reconsider their use as dimensions of administrative burden.

Redemption and Correction Costs are Also Problematic

Redemption and correction costs are also not aligned with the experiential nature of administrative burden. They represent higher-order categories focused on the tasks and activities that lead to onerous experiences rather than directly capturing the experiences themselves. Moreover, they introduce conceptual overlap between different cost categories. Indeed, Barnes (2020) defined redemption costs as a subset of learning costs. I argue that they also overlap with compliance costs. Individuals face significant expenses when attempting to comply with program rules, even when they are familiar with these rules and requirements. For example, due to limits on the quantity of food purchasable with vouchers, individuals must calculate purchases carefully (Barnes 2020, 361). This illustrates the costs associated with meeting administrative rules, such as completing forms or providing documentation to demonstrate eligibility. Similarly, travel expenses incurred by some program participants to redeem vouchers in a neighboring town can also be considered compliance costs. Similarly, the Dutch woman whose car was stolen experienced learning, compliance, and psychological costs in her attempt to correct a series of administrative errors (Widlak and Peeters 2020). This illustrates how these categories conflate the costs individuals experience with specific types of interaction they have with the state (i.e. redeeming benefits and correcting errors).

Cost Categories are Premised on an Implicit Stage Logic

In exploring administrative burden, we struggle with cost categories because we implicitly interpret citizen-state interactions through a “stage logic”—breaking down the interaction process into distinct stages to better understand its components and the relationships between them. This focus on various tasks and activities inadvertently shifts attention away from onerous experiences. While psychological costs directly reflect these experiences, other cost categories merely capture different stages of interactions between citizens and the state. Drawing on van Oorschot’s (1998) dynamic model of non-take-up and Daigneault’s (2023) application to administrative burden, this implicit stage logic is apparent. Initially, individuals learn about programs and services and assess their relevance (learning costs). They then apply for them and fulfill all requirements (compliance costs), utilize them and receive benefits (redemption costs), and if necessary, attempt to rectify administrative errors (correction costs). It is worth noting, however, that the administrative burden extends beyond accessing benefits. It encompasses various citizen-state interactions such as voting, requesting building permit exemptions, filing income tax returns, and contesting parking fines (Halling and Bækgaard 2024; Linos, Quan, and Kirkman 2020; Martin, Liam, and Orla 2023). While dividing citizen-states interactions into distinct stages or activities can be worthwhile analytically, it should not compromise conceptual clarity. Administrative burden refers to costs, not stages or activities.

Putting the Pieces Together

To improve the conceptualization of administrative burden, several steps are essential. First, it is crucial to reaffirm that psychological costs are an appropriate dimension of the concept. Indeed, the negative emotions and mental states represent significant costs arising from citizen-state interactions, making them directly experiential. Second, we should avoid expanding the cost categories, as Baekgaard and Tankink (2021) argued, or developing valid experiential indicators for flawed dimensions, as Jilke et al. (2024) attempted. Instead, we should eliminate learning, compliance, redemption, and correction costs entirely. These should be replaced with dimensions that more accurately reflect the true nature of administrative burden. The new categories should therefore be directly aligned with the experiential nature of administrative burden (i.e. the costs that citizens face, not the tasks that generate these costs). These dimensions should also be conceptually distinct and cover the entire theoretical scope of the concept, encompassing all types of onerous experience. So what should this reconceptualization look like?

An Alternative Conceptualization of Administrative Burden

The Starting Point: Nisar and Masood (2023) Reconceptualization

Nisar and Masood (2023, 8–9) “operationalization” of administrative burden around financial, time, and psychological costs represents a significant step forward. They define financial costs as the monetary expenses incurred by individuals affected by state actions. These include admission fees, travel expenses to government offices, costs associated with obtaining necessary documentation (including fees for obtaining such documents), and expenses for hiring legal assistance to navigate program rules. Time costs refer to the duration of citizen-state interactions, including time spent queuing, filling out forms, and awaiting application processing. Nisar and Masood (2023) clarified that these time costs can also be quantified as financial costs by calculating the opportunity cost of time spent away from employment, based on individual salaries or national income data. Psychological costs arise from direct and indirect stressors, as well as the stigma associated with program participation. Expanding on the original definition, Nisar and Masood (2023)cognitive costs, that is, the “mental effort required to access or implement a particular policy and includes loss of knowledge resulting from failure of policy implementers and beneficiaries understanding each other” (p. 9).

Nisar and Masood (2023, 8–9) have contributed valuable insights for reconstructing the administrative burden on stronger foundations. Unlike the original conceptualization, which predominantly views administrative burden through the actions of the state, Nisar and Masood (2023) introduce new cost categories. These categories—time and money—focus directly and explicitly on the individual’s onerous experiences, while exhibiting less conceptual overlap. Furthermore, the addition of cognitive costs represents a significant enhancement to the theoretical coverage of the administrative burden concept. Dealing with public authorities and navigating administrative processes require considerable mental effort, generating cognitive strain for individuals. As Nisar and Masood (2023) argue, their new categories directly address how citizens perceive and evaluate burdens in everyday life. Individuals often engage in trade-offs between different types of burdens, such as hiring tax professionals to reduce the time and psychological costs associated with filing taxes.

Amending Nisar and Masood (2023) Reconceptualization

Nisar and Masood (2023) reconceptualization requires revisions to establish a robust conceptual foundation for administrative burden. A substantial amendment is the recognition of effort as a distinct dimension within the concept. Nisar and Masood (2023) include cognitive costs (i.e. mental effort) within the psychological costs category, mental effort is qualitatively different from negative emotional states such as stress, stigma, frustration, or feelings of powerlessness. Mental strain represents a burdensome experience, yet it lacks the affective or interpretive dimensions characteristic of psychological costs. Furthermore, Nisar and Masood (2023) reconceptualization overlooks physical effort, a significant cost endured in numerous citizen-state interactions. The dimension of effort should encompass both cognitive strain and physical exertion. For instance, voters in developing countries often trek for hours through challenging terrain to reach polling stations or government offices. Thus, focusing solely on time and monetary expenses fails to fully capture the burdens experienced by certain individuals. In both developed and developing countries, accessing social benefits frequently involves manual labor. This is exemplified by India’s Mahatma Gandhi National Rural Employment Guarantee Program, where participants engage in public works for wages (Ehmke 2016). Similarly, in developed countries such as Denmark, labor market activation requirements include active job searching, maintaining an online CV, and attending courses (Mikkelsen, Madsen, and Baekgaard 2023).

I also suggest two revisions to refine the time and psychological cost dimensions proposed by Nisar and Masood (2023). First, we should exclude “loss of knowledge resulting from the failure of policy implementers and beneficiaries understanding each other” (Nisar and Masood 2023, 9) from the psychological costs category. This exclusion is warranted because loss of knowledge does not directly constitute an onerous experience; rather, it contributes to conceptual confusion. Instead, the costs incurred stem from communication failures and missed opportunities to access public services and programs, encompassing time, money, effort, and psychological impacts. I also propose that time costs should exclude application processing delays, referred to as “passive waiting”—the period between application submission and service delivery. Passive waiting differs qualitatively from actively waiting in line or completing administrative tasks. While passive waiting may lead to financial consequences (e.g. diminished purchasing power, bankruptcy) or psychological impacts (e.g. stress, alienation) for individuals, these costs are already addressed by the other dimensions of administrative burden. Essentially, passive waiting involves no trade-offs or opportunity costs. In contrast, active waiting entails trade-offs between leisure and work time, as well as time dedicated to administrative tasks. When individuals have 60 min at their disposal, they must choose between working or waiting in line, but cannot do both—the essence of opportunity costs (Buchanan 1991). Hence, time costs should be defined solely in terms of the time directly and actively spent interacting with the state (I elaborate on this point below). This minor revision diverges from Nisar and Masood (2023) perspective but aligns with their acknowledgment of citizens’ trade-offs among different types of costs.

Reconceptualizing Administrative Burden Around Four Dimensions

Building on my amendments to Nisar and Masood (2023) categories, I propose a reconceptualization of administrative burden structured around four dimensions: (1) time, (2) money, (3) effort, and (4) psychological costs. These categories emphasize the direct experiential costs of citizen-state interactions. Collectively, they cover the entire theoretical scope of the concept, ensuring it is both directly experiential and theoretically exhaustive. Furthermore, existing administrative burden scholarship frequently addresses, albeit often implicitly, the dimensions of time, money, and effort (e.g. Bell et al. 2023; Chudnovsky and Peeters 2021; Martin, Liam, and Orla 2023; Ritzel et al. 2020). The distinctiveness and non-overlapping nature of these four cost categories remain intact, despite potential empirical correlations or individual trade-offs between them, as acknowledged by Nisar and Masood (2023). While time, money, effort, and psychological costs represent distinct phenomena, they are “fungible” in that they can be exchanged for one another. For instance, the effort, stress, and frustration involved in administrative tasks like filing income tax returns can be mitigated to some extent by hiring tax professionals. Individuals possess both money and time and can convert one resource into the other, but this exchange involves opportunity costs—deciding between alternatives with different implications. For example, an individual could invest three hours in completing their income tax return (thereby saving money) or hire tax professionals for $300 to handle the task (thus saving time), but not both simultaneously. This example highlights how opportunity costs underscore trade-offs between time and money.

I now present the alternative cost categorization and discuss the concept structure in figure 2. As mentioned, the ontological-semantic approach emphasizes the logical relationships between dimensions rather than their empirical correlations. This approach utilizes conceptual maps instead of quantitative methods like factor analysis to uncover underlying assumptions and conceptual challenges in dimensionality and measurement. How do the constitutive dimensions of time, money, effort, and psychological costs interact to shape the fundamental concept of administrative burden? At this stage, it is premature to definitively select among the three main aggregation options discussed earlier. Following Goertz (2020), question marks “?” in figure 2 highlights this uncertainty.

I illustrate each aggregation option with hypothetical examples using Likert scales to measure the administrative burden and specific costs associated with renewing a driver’s license. In this scenario, where “1—strongly disagree” indicates no burden/cost and “5—strongly agree” indicates the heaviest burden/costs, we can consider scores from a citizen survey: time = 5, money = 3, effort = 3, and psychological costs = 1. Under the logical “AND,” the overall administrative burden score is determined by the minimum score among the dimensions, which is “1” in this case. Conversely, the logical “OR” aggregates the maximum score across dimensions, resulting in an overall burden score of “5.” The mean, averaging the scores across dimensions, yields an overall score of “3.” The logical “AND” approach is the most stringent because each dimension is necessary and cannot substitute for another (i.e. a high score on one dimension cannot compensate for a low score on another). In contrast, the logical “OR” and mean allow for partial or complete substitution between dimensions. I argue against the logical “AND” approach, as it contradicts our understanding of administrative burden. Most scholars would likely disagree with a scoring system where the absence of one type of cost results in no administrative burden, despite the presence of other costs. In the hypothetical example discussed above, the logical “AND” implies that there is no administrative burden associated with renewing one’s driver’s license, as psychological costs are nonexistent.

Moving to the indicator-data level, indicators for each fundamental dimension of administrative burden should accurately reflect burdensome individual experiences. However, the determination of the number of indicators and their specific content should be guided by the study’s objectives and context, such as a particular public service or clientele. Given the necessity for both substitutability and redundancy among indicators (Goertz 2020), a definitive list cannot be provided at this stage. Nonetheless, I have compiled a tentative list of general examples of indicators as a starting point (see table 1).

One approach involves using objective indicators to quantify the burdensome experiences arising from citizen-state interactions. This approach includes measuring time (e.g. minutes), money (e.g. dollars), effort (e.g. calories burned, heart rate increase, neural activity), and psychological costs (e.g. hormone levels, facial expressions). For instance, Hattke et al.“s (2020) study on red tape and administrative burden utilized automatic facial recognition software to assess participants” negative emotional responses, alongside measures of electrodermal activity and heart rate. Objective indicators can be obtained through self-reported data or objective sources, such as on-site observers calculating waiting times. Alternatively, a second approach relies on subjective (i.e. perceptual) indicators, where participants evaluate their own perceptions of burdens, including cognitive strain, stigma, stress, and loss of autonomy (e.g. Baekgaard et al. 2021).

The logical “AND” approach is uncommon at the indicator-data level, meaning that multiple suitable indicators are available and can often substitute for one another. I argue this flexibility applies to administrative burden, indicated by the subscript “i” in figure 2 to denote variability. For example, time costs can be assessed by measuring the duration of tasks or activities (e.g. minutes spent completing application forms) or by evaluating perceptions of time burdens using agreement ratings with specific statements (e.g. “completing the application form took too long”). This variability in indicators allows researchers to capture the multifaceted nature of administrative burden across different contexts and study designs. For unidimensional dimensions, the default aggregation option for indicators is typically the mean, as they measure the same object/phenomenon. The logical “OR” is also viable, where each indicator can perfectly substitute for the others (Goertz 2020). The mean or logical “OR” approach can effectively capture indicators for attributes like time and money. However, effort and psychological costs are multidimensional. Effort encompasses both physical and mental components, while psychological costs include various negative states such as stigma and stress, each representing distinct experiences. Understanding the nature of these psychological costs requires further exploration. Moreover, there is a need for more research into positive experiences arising from citizen-state interactions that could mitigate administrative burdens (see Baekgaard and Tankink 2021; Nisar and Masood 2023). This exploration could provide insights into how positive interactions might counterbalance or alleviate the negative aspects of administrative burden.

The reconceptualization of administrative burden into four distinct dimensions—time, money, effort, and psychological costs—provides a comprehensive framework for measurement. I have also illustrated how different aggregation structures—logical “AND,” “OR,” and mean—affect the overall burden score. This reconceptualization offers a more nuanced understanding of the trade-offs between dimensions in capturing the costs of citizen-state interactions.

Having defined “burden” within “administrative burden,” the next step is to clarify the scope of “administrative.” What constitutes the application of the new cost categories? In essence, which citizen-state interactions should be considered when measuring time, money, effort, and psychological costs?

Operationalizing Citizen-State Interactions

I argue we should adopt a broad definition of citizen-state interactions to facilitate the comprehensive study of administrative burdens across diverse settings. First, individuals interact with the state in various ways that extend beyond accessing social benefits. These interactions encompass activities such as voting, filing income taxes, applying for building permit exemptions, contesting parking fines, dealing with child protection services or courts, and requesting visas, documents, usernames, or passwords (Halling and Bækgaard 2023; Herd and Moynihan 2018; Linos, Quan, and Kirkman 2020; Martin, Liam, and Orla 2023). This inclusive approach ensures that administrative burdens are examined in their full complexity and context.

Second, we should broaden our temporal focus to consider tasks and activities that create burdensome experiences before, during, and after official encounters with the state (see table 2). These include anticipating compulsory welfare meetings (Bækgaard and Madsen 2023) or tasks preceding direct interactions, such as assembling documents or traveling to government offices (Chudnovsky and Peeters 2022). Individuals can also experience intense psychological costs after stressful and even traumatic experiences with the state (see, e.g. Widlak and Peeters 2020).

Third, various third-party actors may play a crucial role in citizen-state interactions. For instance, landlords and businesses can be directly involved in program delivery, thereby creating costs for citizens (Barnes 2020). Conversely, nonprofit third parties may contribute to alleviating burdens for citizens (Rode 2024; Tiggelaar and George 2023).

Fourth, citizen-state interactions can be multilevel, nested, and/or sequential. For example, in Argentina, eligibility for the Argentina Universal Child Allowance or AUH (a nationwide program) requires two official documents: a birth certificate issued by local governments and the National Identity Document (DNI) issued by provincial governments (Chudnovsky and Peeters 2021, 2022). However, some individuals, particularly poorer residents, may lack these documents. “Passported benefits,” where eligibility for a specific benefit depends on receiving another benefit (Tarshish et al. 2023), is another compelling example.

Thus, a broader perspective of citizen-state interactions is needed. This perspective should encompass various types of interactions, including those involving third parties, the prelude and sequel of these interactions, as well as their complexity and multilevel, sequential nature. It is crucial to study the cumulative costs arising from these intricate citizen-state interactions (see Martin, Liam, and Orla 2023), ensuring a complete understanding of the cumulative costs individuals face.

Discussion

I have proposed an alternative conceptualization of administrative burden that directly addresses the costs that individuals face when they interact with the state: time, money, effort, and psychological costs. These categories align with the ontology of administrative burden, reflecting individuals’ onerous experiences arising from citizen-state interactions. The four dimensions are conceptually distinct and collectively exhaustive. Now, I address three questions regarding the theoretical and empirical value of the alternative conceptualization.

First, are cost categories truly independent? Conceptually, yes: time, money, effort, and psychological costs represent distinct phenomena. Thus, according to the ontological-semantic approach, we should measure these costs independently from each other. However, in practice, individuals often make trade-offs between these costs (Nisar and Masood 2023). For example, individuals may reduce the burden of waiting in line by purchasing someone else’s place in the queue, thus mitigating time, effort, and frustration simultaneously. Therefore, the subjective evaluations individuals make of these costs are likely not entirely independent within specific contexts. This introduces the risk of underestimating or overestimating administrative burden if we measure only specific dimensions (e.g. time). Hence, it is essential to study all four types of costs together. Moreover, converting certain costs, such as the time spent on administrative processes, into monetary terms (e.g. forgone income), may be appropriate. Using a single metric (e.g. dollars) facilitatescomparison of burdensome programs or costs across various subpopulations. Empirically, the cost categories exhibit correlations and, in some cases, causal connections. For instance, in their study of Swiss farmers, Ritzel et al. (2020) found that compliance costs, measured using time and effort indicators, increase psychological costs (see also Baekgaard et al. 2021). Similarly, correlations may exist between effort and time spent on administrative tasks, or between monetary costs and psychological impacts. While these relationships are valuable for further study, they fall outside the scope of this article, which focuses on reconceptualizing administrative burdens.

Second, can the alternative dimensions and proposed indicators adequately capture the distributive nature of administrative burden? I argue they can, as both the objective and subjective indicators I have proposed are sensitive to variations in individual experience. Social assistance and other means-tested programs, by design, exhibit complex procedures and burdensome requirements (Fox, Feng, and Reynolds 2023; Herd and Moynihan 2018), leading to higher time, money, effort, and psychological costs for their recipients. These costs disproportionately affect disadvantaged groups and individuals, who are more likely to rely on such programs. Moreover, these individuals often face socioeconomic challenges such as lower literacy levels, poor health, or transportation difficulties (see, e.g. Carey, Malbon, and Blackwell 2021; Daigneault and Macé 2020). This more burdensome experience will be captured by objective and subjective indicators of time, money, effort, and psychological costs. For example, a $300 fee to renew a driving license affects individuals differently based on their income level, with less affluent individuals bearing heavier financial burdens. Objective indicators, such as the proportion of income spent on this expense, reveal the disproportionate impact on economically disadvantaged individuals. Subjective indicators, such as the perception of the financial burden, will also capture the greater costs supported by disadvantaged individuals in this hypothetical example. However, there is a caveat. More affluent individuals (e.g. high-income professionals with young children) are likely to value their time more than unemployed individuals or retired pensioners without dependents. Subjective indicators could therefore indicate higher time burdens for these individuals, although this may not necessarily translate into a greater overall level of administrative burden. The key takeaway is that the alternative dimensions and proposed indicators can effectively capture objective and subjective variations in onerous experiences.

Third, how can we empirically validate the inferences derived from the proposed indicators? Empirical validation follows conceptualization, aiming to ensure that we accurately measure the intended concept (Adcock and Collier 2001). We must first check that developed indicators cover all key dimensions of the concept (content validation). While I have argued that time, money, effort, and psychological costs encompass the entire scope of the administrative burden concept, research is required to substantiate this claim. Qualitative studies exploring the subjective perceptions of individuals engaged in citizen-state interactions could prove particularly valuable in that regard. We would then need to assess convergent and discriminant validity for the alternative conceptualization (Adcock and Collier 2001; Pandey and Scott 2002). This involves corroborating that the scores for administrative burden align with those of closely related concepts (convergent validation) while remaining distinct from scores of unrelated concepts (discriminant validation). To gauge convergent validation, I suggest using the three-item scale proposed by Jilke et al. (2024) to measure experiences of burden. Despite being based on flawed cost categories, the scale’s indicators stem from a valuable effort to develop and validate indicators that are directly experiential. Therefore, strong correlations between this scale and the alternative conceptualization would indicate that the latter effectively measure onerous experiences.

For discriminant validation, I recommend using the US Federal Customer Experience (CX) scale, as utilized by Jilke et al. (2024) in their study. Their correlation analysis demonstrated that administrative burden and customer experience are distinct concepts, confirming that their scale effectively measures the intended concept. Ultimately, we must test whether the scores of administrative burden indicators align with those of other concepts considered its causes or consequences (i.e. nomological/construct validation). For instance, if lower administrative literacy leads to higher administrative burdens for individuals (Döring 2021), it would support the validity of the alternative conceptualization.

In summary, the proposed reconceptualization of administrative burden—structured around four distinct categories: time, money, effort, and psychological costs—provides a comprehensive framework for understanding the costs of citizen-state interactions. Measuring all dimensions together is essential, as individuals often make trade-offs between these costs. The alternative indicators are particularly useful in capturing variations in individual experiences, especially for those facing disproportionate burdens. Finally, the proposed conceptualization highlights the need for empirical validation to ensure that the indicators accurately measure administrative burden.

Conclusion

This article has addressed enduring issues with the conceptualization and measurement of administrative burden. Building on the insights of Nisar and Masood (2023), I have reconstructed the concept on stronger foundations. I have argued that administrative burden encompasses four essential dimensions: time, money, effort, and psychological costs. Unlike the cost categories of the original framework, these dimensions are inherently experiential. Therefore, they directly focus on the onerous experiences of individuals who interact with the state rather than on state actions. Moreover, the cost categories cover the entire scope of the concept without redundancy. To operationalize these dimensions, I have proposed a combination of objective and subjective indicators. Finally, I have specified a list of tasks and activities that constitute citizen-state interactions, highlighting the diverse range of contexts where administrative burden manifests.

The reconceptualization proposed in this article offers significant contributions to enhancing our understanding of administrative burden. This alternative framework aims to diminish ambiguity and foster effective scholarly communication. By emphasizing the experiential dimensions of administrative burden, it shifts focus from state actions to the onerous experiences of individuals. The inclusion of time, money, effort, and psychological costs provides a more nuanced perspective on these experiences, enabling scholars to explore a broader range of facets compared to the original framework. Clear conceptual boundaries among the four costs also facilitate precise and valid measurement, eliminating overlap. The proposed conceptualization also supports the study of variations among individuals and causal questions, such as the relationship between administrative literacy and administrative burden. Moreover, it enables scholars to better address questions regarding which costs are significant in particular contexts and for different individuals. Thus, the alternative conceptualization enhances the coherence, rigor, and analytical utility of the administrative burden framework.

Acknowledgements

I am extremely grateful to Martin Bækgaard for hosting me at Aarhus University during my sabbatical leave and for organizing the Administrative Burden Workshop on September 29, 2022. I thank the workshop participants for their comments on my presentation. Special thanks go to Kim Sass Mikkelsen, Matthias Döring, Martin Bækgaard, and Jonas Krogh Madsen for their insightful comments on a preliminary version of this article. I also thank Christophe Tav for editing the figures and Rachel Hatcher and Ashely Rhéaume for copyediting previousversions of this article, respectively.

Funding

This study was funded by an Insight Grant (435-2021-0597) from the Social Science and Humanities Research Council of Canada (SSHRC). I also wish to acknowledge funding from the Aarhus University Research Foundation’s Visiting Researcher Program (AUFF-E-2021-6-20).

References

Footnotes