The politics of complaint: a review of the literature on grievance redress mechanisms in the global South

ABSTRACT Institutionalized complaint systems are notable features of improving public programmes and government practice. This article reviews literature on formal grievance redress mechanisms in the global South to understand whether these mechanisms help the aggrieved to complain and seek redress for their grievances. In this emerging literature, the institutional and definitional boundaries of formal grievance mechanisms are slippery; systems that look like grievance systems may do little to enable complaints by those who seek to register them, and even less to enable them to achieve redress; with limited evidence on how these formal grievance systems work on the ground, and without sufficient power to act on complaints these formal grievance systems appear to be more ornamental; and where they have worked uncommonly well they have not always attracted political support to go to scale. The article concludes with a discussion of avenues for research identified through this literature review.


Introduction
"Complaint" may signal discontent and conflict in human relations, but systems of complaint have great democratic promise when applied to the public sphere.For an average persona person without great wealth, social status, or political powerto complain about official corruption or negligence implies that equality under the law is real, that governments and their bureaucracies disinterestedly serve the public, and that official power need not equal impunity.Over the last couple of decades many developing country governments have set up such complaints systems, often with the encouragement of multilateral development agencies.Systems to enable the aggrieved to complain about and seek resolution for their grievances about public services have emerged across all types of political and administrative systems, regions of the world, and levels of economic development.
This wave of new grievance redress mechanisms, or GRMs, has attracted growing attention from researchers and policy analysts (Caseley 2006;Barca 2016;Bhattacharjee and Mysoor 2016;Bhattacharjee, Mysoor, and Sivakumaramakrishnan 2014;Gauri 2013;Kotagiri and Morel 2018;Kruks-Wisner 2021;Pfeil and Agarwal 2021a;Ringold 2011;Teeffelen and Baud 2011).The definitional and institutional boundaries of GRMs vary, but GRMs are often usefully categorized into three common types: (a) within government or administrative; (b) independent redress institutions such as ombuds offices; and (c) legal redress through courts ( (Babajanian 2015;Gauri 2013;Brown et al. 2014;Rao 2014;Ringold 2011;Kruks-Wisner 2018;Allsop and Jones 2008;Warwick and Ortolano 2007;Bassett et al. 2012).This article focuses on mechanisms that exist within governments or administrative GRMs, as the phenomenon itself and the literature about them is relatively new.The article explores some of the main themes of this emerging literature as it pertains to the global South. 1 It identifies the characteristics common to a family of institutions that enable complaint and their resolution, including formal systems that combine the capacity for people to register their complaints, with (at least in theory), the capacity to enable governments to respond to those complaints.The article also sets out what is known about how and whether these mechanisms work, including the conditions under which ordinary citizens succeed in registering their complaints.Being able to exercise voice is insufficient basis for accountability however, and the article also looks at whether GRMs enable governments and their agents to grow the "teeth" they need to respond to citizens' grievances (Fox 2015).The literature shows that systems that look like GRMs may do little to enable those who most need to register their grievances to do so, and even less to enable them to achieve redress.Some appear to be little more than ornamental, the reluctantly implemented conditionalities of multilateral bank projects, or the white elephant projects of champions of governance reform that went nowhere.GRMs that appear to have worked successfully have not always attracted the political support needed to go to scale, or last beyond political transitions.
This review of the literature about GRMs in the global South is organized as follows.The next section discusses why complaint systems matter, and why it is important to understand the conditions under which they work.Section 3 situates the phenomenon of GRMs in the landscape of policies and practices designed to increase citizen engagement in governance, distinguishing between claims-making, problem-solving, feedback mechanisms, accountability, and mechanisms of grievance redress specifically.Section 4 describes the methods used for this review, while Section 5 presents its key findings.Section 6 concludes with a few key issues that merit further research.

Why do complaint systems matter?
Institutionalized complaints systems are increasingly widespread and notable features of development, both in international aid and national programmes.They have been required as part of all World Bank projects affecting indigenous peoples or involuntary resettlement since 1997; since 2018, they have been required of all investment project financing, as part of stakeholder engagement and information disclosure practice.GRMs were included in 28% of World Bank investment lending projects in 2008, and in some 69% as of September 2020 (Brown et al. 2014;World Bank 2021).Other aid agencies have also promoted the use of GRMs (DFAT, n.d.;ADB 2018).The past two decades have also seen an expansion of national grievance redress systems through legal and independent ombuds institutions and programme-level mechanisms for addressing complaints, many of which are larger in scale and scope than those of aid agencies (Randolph and Edjeta 2011;SSPS 2020;Chen 2016;Gazdar and Zuberi 2014;Gauri and Brinks 2008;Aiyar and Walton 2015).These GRMs absorb considerably more policymaker attention, administrative energy, and resources than in the past.
GRMs tend to be established with high expectations.International donors argue that they improve the quality of public investment and service delivery, and accountability to excluded and marginalized groups in particular (Post and Agarwal 2012b;Pfeil and Agarwal 2021a).For the World Bank, the rationale for GRMs "builds on the business case established for Grievance Mechanisms in the private sector where, it is recognized, they significantly reduce financial, construction, operational, reputational, and corporate risks" (Pfeil and Agarwal 2021b, 1).Project GRMs are claimed to (a) provide fast solutions to individual problems regarding projects, (b) act as an early warning system of impending problems, (c) improve public awareness of the project, (d) help prevent corruption, (e) suggest project improvements, (f) assess internal organization; (g) increase project stakeholders' involvement; and (h) improve targeting (Pfeil and Agarwal 2021b).The Australian Department for Aid and Trade notes that if complaints are credibly and accessibly handled, it increases the chances of achieving project goals, and minimize the risks of controversy, legal action, and delays (DFAT,n.d.,1).
Governments appear to make similar calculations about the costs and benefits of GRMs, seeing them as ways of accessing citizen opinion, improving projects, public services, and performance (SSPS 2020;Randolph and Edjeta 2011;Aiyar and Walton 2015).One study explained the "universal positive association between a country's educational level and the quality of its government" with reference to citizens' complaints: educated citizens complain more and … these complaints lead to better conduct by officials fearful of being punished, which in turn leads to greater accountability and a higher quality government.One attractive feature of this mechanism is that it is entirely decentralized and does not rely on any particular institution, such as democracy.(Botero, Ponce, and Shleifer 2013, 989) In China, extensive public complaints systems reflect the fact that the view that "citizen feedback helps sustain authoritarian rule" (Göbel and Li 2021, 40), by addressing sources of discontent with public authorities and engaging citizens in monitoring lower-level officials (Warwick and Ortolano 2007; see also Dimitrov 2014 on the use of complaints systems in Communist Bulgaria).
Finally, while found across regime types, complaints and their resolution are fundamental to democracy: achieving redress is an aspect of public accountability that matters in particular for the most marginalized and disempowered people.In India, advocacy by activists of national grievance redress legislation in 2011 reflected an "important analytical step in India's accountability movement in its shift from transparency to enforcement" (Aiyar and Walton 2015, 270).This analytical shift can be traced to civil society efforts to institutionalize downward accountability strategies such as jan sunwai or public hearings, statutory social audits (Aiyar and Walton 2015;Pande 2014Pande , 2021;;Vij 2013), and a sub-national Right to Hearing law from the state of Rajasthan (Agrawal and Nair 2018) in state practice.Civil society in India has expressed a longstanding demand for citizens' rights to be heard (Swamy 2020).As Kruks-Wisner 2021, 3) puts it: To complain is an act of citizenship … Through these institutionalized acts, citizens call on the authority of the state at the same time that they challenge the state in light of perceived rights violations or deficiencies in service provision.
If GRMs are able to enable citizen voice, register complaints, and facilitate the institutional response with functioning and resourced systems to assess, monitor, and follow up, they can provide the "teeth" necessary for accountability (Fox 2015).Yet GRMs also appear paradoxical: they offer an opportunity for raising voice when other avenues are closing.They have been growing and enthusiastically adopted by governments at precisely the time governments are restricting free speech and civic space.
Thus, on the one hand, GRMs have the potential to make governance more equitable and inclusive, and to shift power in favour of the powerless (UNDESA 2021; Barca 2016;Kotagiri and Morel 2018;Pfeil and Agarwal 2021a;Teeffelen and Baud 2011).On the other hand, as Sara Ahmed (2021) has convincingly argued about sexual harassment complaints in academia, the very organization of such systems can treat complainants as the problem, rather than taking complaints seriously.Institutions generally tend to protect themselves, and complainants are typically people with little power over those institutions other than their right to register complaints (Ahmed 2021).
The growing move to establish GRMs, therefore, raises questions about motivations: are they genuinely intended to offer opportunities for raising voice (especially at a time when other avenues are rapidly closing), or do they reflect a misplaced faith in bureaucratic solutions to structural differences of power?Or worse, are they driven by the need to fake responsiveness and divert and dilute citizens' anger?Existing evidence suggests a range of different reasons for establishing GRMs.Some may be motivated by technocratic considerations, taking the New Public Management view that feedback and resolution will improve public services.Assessments of such GRMs correspondingly focus on understanding how and whether they improve the efficiency and effectiveness of public investments, reduce corruption, and increase local ownership of projects.Others might be driven by progressive governments installing complaints systems to tap their democratic potential, by empowering citizens to make effective claims on recalcitrant or corrupt frontline officials.As civic space shrinks around the world, such grievance redress mechanisms (GRMs) may well be the remaining channel for citizens to engage in civic action when other, less official, channels, are narrowing or closing.
Yet other motivations may include the need to respond to growing popular discontent manifest in the large number of popular protests regarding public services and government performance in the last two decades.Official systems for gathering and resolving complaints may simultaneously provide governments with tools for citizen surveillance to defuse potential contention, as well as limit collective action by individualizing grievances and their solutions.This may help to explain why there are examples of strong GRMs in authoritarian-leaning governments.Finally, there may also be intra-governmental motivations for using citizen-generated data to monitor agencies and actors within the government, whether to improve services and political popularity, or to enable exercise greater bureaucratic control.
Situating GRMs in the citizen-engagement landscape What precisely are GRMs?While interest is growing, there is imprecision in how terms like complaints systems or grievance redress mechanisms get used.A range of activity, from claims-making advocacy that calls on government to deliver on people's expectations, to mechanisms that merely seek people's feedback, is lumped under the category of GRM (Brown et al. 2014;MSI Integrity 2016;ICJ 2019;Ringold 2011;Brewer 2007).
Complaints systems and grievance redress mechanisms can refer to a range of forms of citizen engagement in the public spherefrom protests in the street to participation through institutionalized mechanisms.Clarifying the definition of GRMs and locating it within the broader set of terms used to describe citizen engagement is critical for understanding why they are adopted, the conditions under which they might work, and what impacts can be expected from them.
To make sense of the literature, it is important to clarify the conceptual terrainhighlighting phenomena that are adjacent to GRMs, but which do not capture the full range of processes that are essential for GRMs to work.Based upon the broader literature on voice and citizen engagement in public policies, we can distinguish between feedback mechanisms, problem-solving, claims-making, and accountability, both broadly, and in terms of how they overlap with grievance redress specifically.Feedback mechanisms concern structured ways of receiving public opinions about issues, both positive and negative, and could include both general matters as well as specific grievances.Problem-solving may include unanticipated challenges in implementing a programme, which are not necessarily about misbehaviour or neglect, but about situations and processes with the potential to cause harm.Current management techniques that emphasize problem-driven iterative adaptation (PDIA), for example, are about identifying potential challenges early on, attempting to fix them, and learning about the process along the way (Andrews, Pritchett, and Woolcock 2013).These approaches can include grievance redress but are broader in the sense of aiming to improve overall systems.Claimsmaking is demand-based and involves citizens indicating that they feel they have certain rights, whether those are reflected in formal entitlements or not.The difference between claims and grievance redress being made here is that claims making is broader and can include rights that are not officially recognized; grievance redress is centrally about rights violations, exclusion and not being treated fairly, according to recognized rules.
Finally, accountability is the larger overarching relationship between powerholders and the public, in which powerholders are regularly held responsible for the exercise of power through administrative, electoral, or legal mechanisms.Social accountability, or collective, citizen-led processes in which public agencies are held to account for systemic failures through public deliberation have dominated recent discourse on improving governance through citizen participation, and the rise of GRMs has happened within this space (Joshi and Houtzager 2012).Yet, social accountability approaches often focus on the proximate causes of system failuresthe frontline providersdeflecting attention away from those with real power.GRMs may exacerbate this by individualizing complaints, so that system-wide problems remain hidden from public view.Similarly, legal empowerment, an adjacent approach which focuses on rights violations and people knowing, using, and shaping the law, can also be distinguished from GRMs: it goes beyond projects or programmes, and there are clear processes of resolution, including reparation for aggrieved parties established in law (Joshi 2017).
The processes outlined above and their relationship to GRMs can be mapped diagrammatically as shown in Figure 1.We arrived at this graphic depiction through an iterative process by juxtaposing the broader literature on accountability against the literature we found on GRMs.On the X-axis, there is a range from processes that only elicit citizen voice, including opinions, feedback, claims and grievances (left), to processes that include some measure of responsiveness from authorities (right).Further, citizen voice if individualized, relies on goodwill, rule-based behaviour, and internal incentives for the state to respond, whereas collective voice offers the potential for collective negotiations around resolution.On the Y-axis, GRMs range from taking more managerial approaches (bottom), to more political ones (top).Clearly, these all overlap at the marginse.g.grievances could be part of feedback loops, problems may arise due to felt grievances, there can be accountability for wrongdoing and complaints.GRMs sit at the centre of these closely related processes, overlapping with them in various ways.
In this review article, grievance redress mechanisms are identified as a narrower subset: they involve publicly institutionalized mechanisms of collecting complaints about rights violations or fairness from the public and providing resolution to complainants.This definition is represented by the grey area at the centre of Figure 1, which is a subset of the broader landscape of citizen engagement.What this figure does not show is that GRMs for the most part, individualize complaints rather than enabling collective action.As the term GRM indicates, grievances refer to complaints that people have about public action, when they have not been treated in line with their expectations in the public spherein terms of not receiving services, mistreated, or treated unfairly or at a fundamental level having a real or imagined cause for complaint.Moreover, GRMs must have an actual mechanism, an institutionalized process for handling complaints and doing something about them that is authorized, recognized, or embedded in public institutions under the law.This does not mean that the actions must be taken wholly or solely by governments.GRMs also require that complaints raised through them will have to have a clear process and possibility of resolution or redress (although a question emerges about what constitutes acceptable redressfull or partial satisfaction on the part of the complainant, or an explanation of how the complaint was acted upon on the part of the government).And finally, GRMs are not only a matter of technical design and information flows but also about power relations: citizens are empowered to make claims on public agents who in turn should be empowered to respond appropriately.
Defined thus, two broad strands of motivations can be distinguished in the implementation of GRMs.One, the techno-administrative strand (the blue area below the x-axis in Figure 1), sees them mainly as data collection processes that enable fine-tuning of government programmes.Drawing from New Public Management and related marketbased approaches, GRMs can be seen as means of improving "client" satisfaction.This view, popular with donors and governments, focus on the potential for better programme delivery with GRMs, and fits within a larger set of attempts at citizen engagement to improve service delivery that include technical approaches to customer feedback, social accountability, and citizen participation in programme design (Babajanian 2015;Bhattacharjee and Mysoor 2016;Randolph and Edjeta 2011;Ranganathan 2008;Rao 2014;Ringold 2011).What matters in the techno-administrative strand is that complaints are registered (sometimes as box ticking due to project requirements, or in the worst cases as a means of surveillance), and response is targeted more to programme improvements (or reprisal) rather than justice or rights.
The other, the socio-political strand (the green area above the x-axis in Figure 1), views GRMs as ways of enabling citizens to engage directly with institutions and programmes that affect their lives based on rights (and their violation).This strand emerges from a view that GRMs offer a channel to practice active citizenship by exercising voice, rendering judgements based on rights and fairness, with the potential to collectively challenge power through political engagement (van Teeffelen and Baud 2011;Chakraborty, Ahmad, and Seth 2017;Marathe et al. 2016;Caseley 2006).The potential of the socio-political strand stated above is dependent upon complaints being made visible to the broader public, to trigger collective action and negotiation and engagement with public institutions (Peixoto and Fox 2016).Grievance mechanisms located in this strand will enable the aggregation of voice on the one hand, and the institutionalization of response processes on the other, including explanations to complainants about how their grievances were dealt with.These two different starting points affect the design, use and effectiveness of GRMs, and can provide a framework for thinking about existing GRMs and their potential.

Methodology
The findings reported here are based on a review of the English language literature on grievance redress mechanisms. 2 Although we did not track the geographic focus of the articles in our initial search, a large majority of those returned in the search appeared to relate to services in the advanced economies and states of the global North.We chose to focus on grievance redress mechanisms in countries in the global South within this literature because (a) these are relatively new institutions in the developing world, which are (b) frequently promoted by aid agencies and multilateral financial institutions; and (c) unlike in the global North, there has been relatively little analysis of them to date.
To identify appropriate literature, we searched Google Scholar and Google, as we recognized that not all relevant literature was likely to be in scholarly publications.The search strategy involved the use of combinations of terms identified through trialand-error as relevant and commonly used in discussion of these institutions, including such terms and strings as grievance+redress+mechanisms & grievance+mechanisms & complaint+systems.We also included sectors or themes on which there were pre-existing reasons to believe literature might be found (e.g."social protection"; "citizenship"; "governance").We searched the citations and references in prominent articles on the phenomenon of grievance redress or complaints systems, and of synthesis reports on GRMs (which several aid agencies had commissioned or undertaken).We also undertook manual searches of organizational websites such as the World Bank, the Green Climate Fund, and the Asian Development Bank.We conducted further searches on specific programmes and service sectors as well as authors that were frequently associated with the themes of complaint and grievance redress systems.In addition, we also reached out to key informants in multilateral organizations with a track record of designing, implementing, and documenting these mechanisms to help us identify key documents and sources.
The focus was on mechanisms related to public services in the global South.This meant that it included GRMs for aid-financed projects and programmes where those are run by governments.Out of a total of 881 articles uncovered by the search, 181 documents were finally selected for closer reading on the basis that (a) they appeared to contain substantial discussions (as opposed to mentions or nominal discussion) of GRMs; and (b) described instances of such mechanisms in relation to public services in the global South.On closer reading, around half of these 181 were further excluded from the review for a range of different reasons: some described the design but not the implementation of the mechanism; some were insufficiently objective, and appeared to be intended to promote the project or programme; others focused on the technical aspects of data management but did not give a sufficiently informative account of the user experience or programmatic impact.A sub-set of 90 documents was finally included in the review.
Terms used for GRMs differ across institutions and sectors.Parts of the UN system use the term Grievance Mechanisms, while the Green Climate Fund has established a community of practice around Grievance Redress and Accountability Mechanisms (GRAM). 3Similar but not identical to GRMs are terms largely derived from private sector practice, such as "customer complaints", "complaints systems", "customer feedback" and "consumer protection" systems.Some literature on GRMs uses this language (in particular "complaints," and "feedback"), but although there is some overlap with GRMs in relation to mobile money services which are increasingly used to transfer cash as part of government social protection services, 4 these terms chiefly refer to firms' customer-engagement practices within the digital economy.

Findings
Grievance Redress Mechanisms (GRMs) as distinct mechanisms with a common label in use in relation to public policies and programmes in the global South have emerged only in the past 20 years.This emergence reflects a combination of New Public Management approaches to governance and service delivery, and the "accountability turn" in international development after the publication of the World Development Report 2003/4.A keyword search found that the use of the term "grievance redress mechanism" in published (academic and grey) literature had grown exponentially over the last two decades.Google Scholar recovered no publications using that specific term in 2000, but almost 700 pieces using that term by 2020.
The published literature features a Northern bias, focused on GRMs in the private sector as well as in public health, welfare, or other social policies and programmes.Systems for addressing complaints across transnational and/or multilateral systems, including using international legal mechanisms for human rights-related complaints, are also discussed, chiefly in institutional (rather than scholarly) publications.
Other literature examines the broader politics of "claims-making", situating GRMs within the formal structures and informal politics involved in realizing citizens' rights. 5The literature covers different political regime types (pluralist electoral democracies such as India and the Philippines and authoritarian regimes with strong governments such as China and Vietnam); weak states and strong; and those with as well as those without substantial dependence on international aid.From this overview we can conclude that GRMs close to the definition supplied above have emerged as an important institutional response internationally to the challenge of making public services work in the past 20 years.
GRMs tend to be common but understudied institutional practices (Gauri 2013; Kruks-Wisner 2018; Ringold 2011) and the scholarship is more focused on design than implementation.The small number of comprehensive studies have all noted the limited research on the role of civil society (Bhattacharjee and Mysoor 2016), and public services (van Teeffelen and Baud 2011;Chakraborty, Ahmad, and Seth 2017;Marathe et al. 2016;Caseley 2006) in assisting citizens seeking resolution or redressal.
The evidence on GRMs and their impact on the citizens, frontline providers and bureaucrats who deliver them is shaped by the following factors: . GRMs are often discussed as part of broader discussions of programme achievements or impacts without specific assessment of the role played by GRMs in those outcomes.
To date, there is no evidence that the multilateral institutions that have been promoting the use of GRMs have invested in developing an outcome-oriented business case for these mechanisms.This is in part because they are required under organizational policies, regardless of their impacts on policy or programme outcomes.Establishing GRMs is the responsibility of the client country, and thus depends on their commitment and proactiveness in putting them in place and ensuring they survive and function. .The focus of most literature that addresses GRMs directly is on formal grievancehandling mechanisms.Informal institutions and local power relations that are likely to shape how effectively citizens can complain are frequently ignored.
. Research tends to focus on responses to service failures, rather than on policy feedback regarding the causes of service failures.
The existing scholarship is largely descriptive, providing general accounts of basic principles, design, staffing, costs, or demand (civil society, citizens) and supply side (government, courts and ombuds) facets of GRMs (Gauri 2013;Randolph and Edjeta 2011;Rao 2014), or descriptions of proposed GRMs (UNDP 2017; MSI Integrity 2016; Sepúlveda 2018; Rohwerder and Rao 2015;Post andAgarwal 2012a, 2012b).There is often overlap between the three types of GRMs that are common across countries and sectors: (a) within government or administrative GRMs; (b) independent redress institutions such as Ombuds; and (c) legal redress through courts (Babajanian 2015;Gauri 2013;Brown et al. 2014;Rao 2014;Ringold 2011;Kruks-Wisner 2018;Allsop and Jones 2008;Warwick and Ortolano 2007;Bassett et al. 2012).Countries with more developed legal systems may rely more heavily on courts for redress, and countries with an active civil society may make greater use of independent third-party channels for redress.
The evidence on whether and how formal GRMs work is thin.Governmentsdespite varying motivationsappear to recognize their potential value, but the evidence suggests that formal GRMs are better at problem-solving than improving government responsiveness.Governments pay little attention to registering, tracking, and enforcing resolution.Complainants often know little about GRMs and are discouraged from filing complaints because their complaints are deemed "ineligible".Thus, by design, formal GRMs tend to raise expectations they cannot meet.Lack of attention to unequal power relations means there is limited evidence on the unintended consequences of using GRMs.Marginalized groups are at high risk of reprisals when they make complaints, and frontline service providers may end up with the blame for systemic problems.We elaborate on these findings below.

Localized problem-solving vs stimulating policy responsiveness
The most prominent impact of formal GRMs is help with local problem-solving.An Asian Development Bank supported road project in Vietnam found that GRMs made it possible to resolve many grievances that could otherwise have meant livelihood and asset losses.The GRMs registered and resolved hundreds of complaints about access to cultivable land, property damage during road-building, and unfair compensation, with formal institutions being supported by mass unions and other local organizations (ADB 2018).The resolution of these complaints both enabled the project to continue without undue hindrance, and ensured that even citizens affected by resettlement were appropriately compensated and supported through the process.
However, other projector programme-level GRMs are inundated with complaints that are about the overall policy that established the project or scheme, but which they have no capacity to address.For instance, a common complaint in social protection programmes is about exclusion from targeted schemes: several GRMs suffered from excess workloads arising from large numbers of complaints regarding exclusion (UNICEF 2015; Gazdar and Zuberi 2014;Shelley 2015).As Barca notes on the basis of their review of (both public sector and aid agency-managed) GRMs in Indonesia's social programmes, a grievance redress mechanism "cannot compensate for poorly designed or implemented programmes … eligibility criteria should be simple and there should be adequate financing to ensure all eligible target group in each province/district can be included" (Barca 2016, 6).Successful complaint systems often had strategies for tracking and "escalating" complaints, feeding their findings up to higher policy-making levels.When that occurred, widespread complaints that were about the substance of the policy rather than the operation of the programme were able to trigger policy changes.For instance, the government-run GRM for the Philippines' flagship Pantawid Pamilyang social assistance programme registered half a million complaints, many of them about exclusion.The resolution of these complaints led to the retroactive payment of USD 71 million to households who had their eligibility confirmed (World Bank 2014).

GRMs have the potential to improve public services
There is some evidence that GRMs help improve service provision where they have support from higher ups.The MajiVoice system for handling complaints about services in water utilities and contracted water service providers in Kenya found that its integrated system of (some) online and (predominantly) face-to-face or phone-based complaints registration was highly effective, with complaint numbers increasing tenfold once it was established (Belcher et al. 2016;Peixoto and Sifry 2017).Most users were satisfied with the responses received, and more felt the experience of complaining had been worthwhile.The MajiVoice system is actively monitored by the Water Services Regulatory Board, which is empowered to take legal action against failures to meet agreed service standards, thus demonstrating the value of an integrated, proactively transparent and formally empowered system (Belcher et al. 2016). 6 GRMs that work enjoy political commitment and support from senior officials.May et al (2009) observe that the complaint handling system for Asian Development Bank's Earthquake and Tsunami Emergency Support Project in Indonesia was able to identify problems early and provide resolution, largely due to high level commitment from project senior management, dedicated staff and finances for the project level GRM.High level support meant lower-level staff could not ignore complaints, and only a few complaints were elevated to higher levels.
Similarly, in India, an urban municipal water provider in a southern state reformed its complaint services (Caseley 2006).One of the reforms was an online (internally networked) system that monitored staff performance based on citizen complaints.Since the system was established, citizens who used the new service "perceived significant changes" in staff behaviour with most complaints addressed rapidly, rather than being simply ignored.Despite the improvements, complaints continue to rise, which Caseley attributes to a combination of "greater awareness among citizens, consistency in referrals made by section staff, and, most importantly, satisfaction among citizens, resulting in repeat engagement with complaints services" (Caseley 2006, 535).
Moreover, GRMs that are part of adaptively managed programmes can contribute to programme improvements that help in the resolution of complaints.As part of the "adaptive management" of Egypt's Takaful and Karama unconditional cash transfers programme, government-run GRMs were used to aggregate and analyse complaints of exclusion by people with disabilities.The GRM was a source of information that helped programme staff zoom in on key implementation bottlenecks, in this case the incorrect recording of disabilities affecting eligibility for inclusion in the programme.Of more than 270,000 complaints between 2018 and 2020, almost 100,000 complainants had their assessments revised so that they became eligible for the cash transfers (Pfeil and Agarwal 2021b, 12).

GRMs can raise expectations but often fail to provide redress
Formal GRMs often raise expectations they cannot meet, particularly when those in charge of accepting grievances have no power to redress them.In Mexico's conditional cash transfer programme Progresa (later Oportunidades), government-run citizen complaint windows provided a degree of answerability to programme participants.Before its creation, the programme was accountable "upwards" to officials but not "downwards" to participants.But programme managers of these windows were wary of making complaints public and used discretion in assessing patterns of complaints.The managers also decided whether or not a complaint was checked off as addressed, leading to fake high rates of resolution.The programme was intended to be responsive/listen to individual citizens, and addressed a felt need, evident from data on growing number of complaints.The citizen complaint window operators could fix problems involving programme implementation such as access to the programme or payment glitches because they were part of the same agency.These complaint windows, however, could not address complaints about issues related to other government department/employees like teachers or doctors; and, once a grievance was transferred to another department the commitment to anonymity was lost (increasing the risk of reprisals) (Fox 2007).
Formal GRMs are also not a guarantee for being heard or redressal.One Indian study shows that even with a legislative guarantee, in practice numerous limitations can affect the implementation of the for GRM for Right to Education such as conflict of interest between implementing and executing agencies, lack of clarity on how to deal with variety of complaints that could involve multiple actors: schools, teacher, parents; lack of clarity among complainants on who to approach; and lack of specific remedies for different violations (Bhattacharjee and Mysoor 2016;Bhattacharjee, Mysoor and Sivakumaramakrishnan 2014).Other scholars note that to enforce redressal of complaints, Indian citizens need additional spaces (van Teeffelen and Baud 2011).Citizens engage in "level hopping" (Kruks-Wisner 2021); "going up the ladder" (van Teeffelen and Baud 2011) or "level jumping" (Robinson 2014) to engage the state, draw attention to their grievances.As Kruks-Wisner (2021) notes claims-making often "fails" but GRMs are a way of "thickening" the relationship between citizen and state and provide a critical channel to marginalized groups to be heard, especially related to police-citizen relations.But complaining is "paradoxical" and can at timesdue to local government constraints deepen grievances and lead to citizen-withdrawal (Kruks-Wisner 2021).Thus, citizens demanded that India's national grievance redress legislation puts the "onus on the state" to facilitate grievance redress, allowing citizens to approach officers with powers to summon (lower-level officials) and hear out a complainant, thus increasing the chance of redress (Sabhikhi 2017).However, where there is no strong commitment from the top to resolve and address complaints, GRMs can be little more than ornamental.For example, despite legal backing for Nigeria's Public Complaints Commission, lack of power hinders how effectively it responds to complaints (Aina 2012).Thus, GRMs are only as effective as the power granted to those in charge of handling complaints.

GRMs tend to exclude the marginalised by design
Despite growing interest and widespread adoption of Internet-based GRMs, digital solutions tend to be unidirectional and opaqueinformation is taken from citizens, not given back; complainants are unable to determine whether their complaint is universal or not.More attention is paid to the use of technology than to its users.Some scholars point to a "democratic deficit" in relation to implementing new public management strategies such as electronic platform for grievance redress (van Teeffelen and Baud 2011).They find that Indian middle-class citizens are better empowered to obtain greater accountability from local government for operation and maintenance of complaints (see also Caseley 2006): When new public management procedures are introduced in local government, they are based on the assumptions of equality among the citizens they serve and the capacity of citizens to function as active consumers, protecting their own rights.In some situations, these assumptions are not realistic, that is, citizens who do not have the full set of rights usually given by local government (democratic deficit) are not able to exert their full capacity to activate those rights ( van Teeffelen and Baud 2011, 181).
Scholarship on digital solutions tend to also underplay the role of intermediaries in navigating digital complaint systems and follow up of complaints.The Indonesian government's large-scale national complaint handling system (LAPOR) is easy to use, and some argue that as a digital platform it helps citizens report on government performance by, bypassing the bureaucratic process (Mahendra, Pratiwi, and Prawesti 2014).But the government institution in charge of receiving the complaint has broad discretionary power to determine resolution which is limited to providing a response to the complaint. 7And problems of access are common.LAPOR is used primarily by men (aged 31-45), educated to degree level and who work in private sector.Such gendered effects on access to GRMs are also noted in Ethiopia.Female beneficiaries of social protection programmes, reported knowing of opportunities to raise their concerns but chose not to do so due to fear or because they did not want to "bother" the service providers (Vinci and Roelen 2018, 20).By contrast, however, beneficiaries of the Benazir Income Support programme in Pakistan claimed that being able to visit public offices to register their complaints about the programme was itself an empowering experience for women who had previously had no reason to engage with public officials (Gazdar and Zuberi 2014).Though women who were able to file grievances needed further assistance in understanding the complaint or redressal processes, and that educated and resourceful people played that mediating role (Gazdar and Zuberi 2014).
In addition to needing help from third party, civil society actors to navigate the complaint process, citizens living in remote areas may be unable to use complaint systems, particularly if access is through digital platforms.In one Indonesian province, Bojonegero, with high levels of internet connectivity, awareness about the complaint system was low and residents used the local government radio show to make complaints (Siregar et al. 2017).In Ghana, similarly, local radio was a popular channel for raising and addressing service delivery issues in the capital Accra (Selormey 2013).More broadly, deployment of advanced Information Communication Technology is no guarantee the system will be widely used.Awareness and trust building are important for reaching populations who feel they have been treated unfairly and would like their grievance heard and addressed. 8For example, digital solutions for grievance redress often ignore the crucial role of intermediaries or independent convenors that could help facilitate access by marginalized groups (Seth, Gupta, and Johri 2021;Feruglio and Misoi 2017).

Unintended consequences of formal GRMs
Despite the best motivations, GRMs can have unintended consequences for both users and those charged with implementing them.A common complaint from those tasked with administering GRMs is that this leads to excess workloads, usually by generating a large volume of complaints that are not actionable by the GRM authority (Ojha 2019), or about exclusion from social protection programmes over which they have no discretion (Gazdar and Zuberi 2014;Vinci and Roelen 2018;Barca 2016).Frontline service provider morale is an importantthough often ignoredaspect of GRMs.There is a tendency in the literature to portray public service employees in low and middle income countries as "self-interested" and "profit-maximizing" (Davis 2004).The introduction of new public management strategies could lead to reform sabotage unless complemented with measures to positively address frontline staff perceptions of their jobs.For example, one study of South Asian water and sanitation agencies found that face-to-face interactions betwen service provider staff and customers contributed to a new sense of calling, helping reorient their practices (Caseley 2006).The lack of a "human factor" in follow up for grievances is also noted as a limitation of Indonesia's national complaint handling system LAPOR (Mahendra, Pratiwi, and Prawesti 2014).According to a literature review of citizen-monitoring initiatives and corruption, if citizen-monitoring interventions provide contact between citizens and service providers, and enable citizens to monitor government performance, the potential benefits of citizenmonitoring intitiatives are likely to be higher (Molina et al. 2016).
Complaining may be an important aspect of citizen-state relations, but it can be risky for the complainant.The act of complaining, especially by marginalized populations, is an extremely powerful activity that can mobilize powerful groups who fear loss of social status or power organize to block or suppress grievances.Evidence suggests that citizens complain at great personal risk.In India, 450 users of the Right To Information law have faced violence or harassment, with 87 reported deaths of citizens legally seeking transparency on public interest matters. 9Many had filed RTI requests for information inquiring into a prior complaint or grievance related to mining, corruption, fund utilization for village drainage system, civic projects, illegal electricity connections and land (Pande 2015).In Ethiopia, female beneficiaries of two social protection programmes told researchers they do not engage in activities or discussion or raise issues because of fear of repercussions (Vinci and Roelen 2018).In Pakistan, beneficiaries of the Benazir Income Support Programme were hesitant to lodge or puruse complaints because they feared backlash or that their beneficiary status might be revoked (Gazdar and Zuberi 2014).
Research has also pointed out that designers of GRMs should bear in mind the effects of complaints on employees, frontline staff, health professionals against whom complaints are made.For example, Bangladeshi civil servants noted that people tasked with receiving complaints regarding public appointments processes were themselves corrupt: and raising questions could worsen the situation (Alom 2021).While complaining is an important activity that could lead to improved service delivery, more research is needed on the relationship between individual encounters and complaints and institutional responses to complaint systems (Gill et al. 2019).

Conclusions
The review of the literature on GRMs presented here highlights gaps in the research, as practice has outstripped evidence of their adoption, implementation, and performance.We outline several areas in which important gaps exist.First, research is needed to unearth underlying motivations of policy makers involved in their design, and the political processes through which they are adopted and implemented.We understand little about the motivations for the widespread embrace of GRMs, and our initial observations suggest that those that have been motivated by technical concerns are less likely to have the democratic-deepening effects that are possible.Moreover, GRMs that have been adopted for purely presentational purposes (e.g.due to conditionalities by aid programmes, or to be seen to be doing something) will likely remain tokenistic.At the more malevolent end, those that are set up in order to surveil activists and claimants, or simply to diffuse protest, will not be used by people as intended due to the fear of reprisal.Conversely, when civil society groups have been involved from the start in the design and adoption of GRMs, they are more likely to "fit" the needs of the constituencies they are aimed at.But these are conjectures: we simply do not know enough about why governments adopt GRMs, and what effect these motivations have on subsequent implementation and impacts.
Second, there is a need to understand the internal workings of GRMs.How and under what conditions do GRMs function effectively, in terms of successfully registering and resolving complaints?What are the technical and design features of such systems?The existing literature offers initial propositions about these issues.For example, GRMs that are proactively transparent 10 and seen to address grievances in a satisfactory fashion, are more likely to be used.People use mechanisms that they see to be working.Another proposition is that GRMs that encourage collective complaint are more likely to be used by poor and marginalized groups who are vulnerable to backlash.Similarly, there are several issues around the use of digital technologies in setting up GRMs that can affect their use and success.For example, technology enabling higher level officials to monitor the resolution of complaints by frontline workers can incentivize the resolution of complaints.At the same time digital technologies presume understanding of and access to new technologies, which might not be available to marginalized populations.
Third, this review found that despite the growth, official recognition and spread of GRMs, they have not embraced proactive disclosure.Formal GRMs are largely individualized, with opaque data on complaints, patterns, and responsiveness.Proactive disclosure about types of complaints and resolutions that can enable collective action is critical for GRMs to be useful in reshaping citizen-state relationships.But it must be done keeping in mind existing power dynamics, facilitating access to complaint systems by marginalized groups, and protecting anonymity.Proactive disclosure requires governments commit resources to register, track, and resolve complaints, and empower officials to enforce resolution and make all this information public.Without such proactive transparency about the nature, and number of complaints, and rate of resolution, widespread systemic problems will go unnoticed, and prevent governments from preventative policy measures.
Finally, an issue that deserves greater attention concerns the role of GRMs in the wider policy and governance landscape.To what extent, can GRMs be the locus of synergies between citizen capacities to complain, and state capacities for redress?Can GRMs effectively make public services pro-poor, inclusive and accountable?Does the existence of reliable GRMs lead to changes in behaviour of public officials, so that GRMs are present, but rarely used in practice?How might GRMs lead to virtuous cycles of strengthening capacities of states and citizens and increasing trust in government?More ambitiously, can GRMs lead to improvements in government policies through accurate, candid feedback that can reform design or delivery of policies higher up the policy chain?Such positive potential of GRMs is currently being claimed, but not yet empirically established.Answers to these questions then are critical given the growing importance of GRMs in programme and policy designs in developing countries.

Figure 1 .
Figure 1.Grievance redress mechanisms at the intersection of citizen voice, political and managerial action, and state responsiveness.Source: Authors own construction.