Negotiating expertise in Nepal’s post-earthquake disaster reconstruction

1. INTRODUCTION

One characteristic feature of post-earthquake recovery in Nepal after the 2015 earthquakes was an owner-driven approach to private housing reconstruction in the most devastated rural regions. Through this approach, the Government of Nepal and its donors aimed to combine a commitment to appropriate technology and building safety. To disseminate the model, engineering expertise was deployed throughout the affected areas in the person of over 3000 engineers tasked with implementing the National Reconstruction Authority’s (NRA) Private Housing Reconstruction Program, which eventually included almost one million households. As they carried out their responsibilities, the NRA field engineers encountered local administrative systems and everyday life-worlds.¹ Some of the complexities at the interface of rural publics and NRA engineers became evident in scenes the authors witnessed during the ethnographic field research.

Private housing reconstruction grants for a total of NPR300,000 (Nepali rupees²) were broken into three tranches for each beneficiary household. In order to gain approval for the second tranche of government funding, householders had to demonstrate compliance with a requirement to strengthen new stone and mortar with wooden or reinforced concrete bands at two-foot intervals. The second (and largest) tranche release (NPR150,000) was contingent on completion of the house’s foundation, which the NRA engineers assessed by inspecting the plinth serving as a damp proof course (DPC), required to be built between the foundation and the house itself. If in urban settings the DPC would typically be made of cement, the NRA house designs allowed for timber in rural areas, as long as it were treated with a ‘black paint’ made from pine tar to prevent it from rotting or being attacked by termites. In practice, householders understood the technical importance of this procedure, but either ignored it or painted only that part of the plinth that would be visible to the NRA engineer for the purpose of passing the inspection. In fact, in many cases, people had already rebuilt their ‘own houses’ damaged by the earthquake as much as a year before the reconstruction programme arrived in their villages (Limbu et al. 2019).³ Many did not need the ‘earthquake house’, whose designs they found inadequate for their purposes, and were long accustomed to designing and building their own houses. But they did need the funding tranches, having already expended far more significant resources to rebuild their ‘own houses’. Other actors were also involved: newly elected local politicians to whom the engineers reported; district- and central-level officials developing house designs and technical trainings; and ‘social mobilisers’ dispatched by the government and non-governmental organisations (NGOs) to facilitate encounters across multiple vectors of expertise.

The present paper examines these encounters through ethnographic research in two rural settings in Nepal. The aim is to convey what happened in practice as diverse actors in the reconstruction process encountered each other’s disparate subjectivities and expertise. A wide range of participants were interviewed in post-earthquake reconstruction ‘on-the-ground’ situations, and various aspects of the process were also observed. ‘Knowledge transfer’ is conceptualised as a multidimensional process in which local community members shared relevant expertise with those deemed to be ‘experts’, at the same time as the latter imparted technical information. These experiences of mutual learning and implementation constituted the fraught field of reconstruction in practice. On that field, the concept of ‘cultural safety’ emerged as a dimension of reconstruction that proved critical to the wellbeing of both engineers and householders.

When reconstruction is approached in this holistic manner, neither of two prevailing explanatory paradigms commonly engaged to interpret reconstruction in the social sciences—‘Build Back Better’ (BBB) or ‘disaster capitalism’—proved adequate for analysis (Cheek & Chmutina 2022; Klein 2008). By investigating how engineering expertise articulates local knowledge in practice, this paper challenges such binaries and instead explores three epistemological questions that centre the cultural politics of disaster reconstruction:

• Who is an engineer?

  This question elaborates engineer subjectivities and knowledges in practice.

• What is a house?

  This considers how residents negotiated engineering expertise in owner-built reconstruction of their houses.

• What is safety?

  This explores engineering rationalities in relation to expanded conceptions of safety rooted in culture and everyday life.

These queries relate to the central problematic of negotiating expertise, allowing for not just multidisciplinary frames but also ontological and epistemological complexities, which the paper argues are necessary to create opportunities for approaching disaster preparedness from ground-level anticipatory frameworks.

The Gorkha Earthquake struck Nepal in April 2015, measuring 7.8 on the Richter scale and triggering several significant aftershocks (hence the plural ‘2015 earthquakes’ throughout). A total of 31 of 75 districts were affected, 14 of which were declared ‘severely affected’, all in central Nepal where this research was conducted. Over 9000 lives were lost (NPC 2015). The present enquiry on negotiating expertise in disaster reconstruction was part of a larger collaborative international research partnership concerned with grassroots perspectives on the domains of construction, law and finance in earthquake reconstruction.⁴ Like the overall partnership, the research presented here involved a wide range of researchers from multiple disciplines and vocations—anthropology, development studies, engineering, geography and political science—at different stages in their careers, based (and trained) variously in North America, Europe and Asia.

This paper prioritises perspectives on reconstruction encountered ‘on the ground’, with an emphasis placed on distinctive forms of expertise. It draws on ethnographic research in two of the districts most severely hit by the earthquakes, namely Dhading and Sindhupalchowk, located to the north-west and north-east of the capital Kathmandu, respectively (Figure 1). In Dhading, the research centred on the relatively remote village of Borang, populated primarily by subsistence farmers from the Indigenous Tamang community. All 298 houses in Borang were categorised as ‘fully damaged’.⁵ Borang was chosen because it is a rural location in the Himalayan belt without easy road access, and it was receiving technical assistance from the Baliyo Ghar (Strong House) Program of the National Society for Earthquake Technology—Nepal (NSET-Nepal, a Nepali NGO). In Sindhupalchowk, Kartike Bazaar was selected for comparative purposes, as an ethnically diverse market town with road access in an otherwise rural area, and where the Baliyo Ghar Program was not implemented.⁶

Figure 1

Research proceeded in three phases, in spring and winter 2018 through 10–12 days of site visits, and for a shorter follow-up visit in winter 2020. This site-based research was supplemented by archival research, key-informant interviews and observations at district headquarters. In Borang and Kartike, research included in-depth interviews with householders, observations of house reconstruction and associated local economies, as well as key-informant interviews with individuals in local government offices and NSET-Nepal. In total, across all research sites, 180 interviews were conducted in Nepali, which were then translated into English.⁷ In the district headquarters the observations centred on interactions in and around banks, government offices and public infrastructure. The academic and applied engineering expertise of co-author Bishnu Pandey (e.g. Pandey & Ventura 2019; Pandey et al. 2020), a Professor of Civil Engineering in Canada who has long-time affiliations with NSET-Nepal, was also drawn upon.

The paper is structured as follows. Next, the scholarly debates about disaster reconstruction and theories of expertise are presented. Context of the official institutional landscape of reconstruction is then explained. The subsequent three sections delve ethnographically into the three key questions. The conclusions consider the overarching question: ‘what is reconstruction?’, which reflects on insights and themes that could be relevant to planners and engineers in other contexts and conjunctures.

2. PARADIGMS OF DISASTER RECONSTRUCTION AND EXPERTISE

The prevailing trope for interpreting post-disaster reconstruction resides in liberal narratives of ‘Build Back Better’ (BBB), which references the possibility that damaged infrastructure can be reconstructed to be more structurally resilient, as well as more socially and environmentally sustainable, than before the disaster. The slogan gained traction internationally after the 2004 Indian Ocean Tsunami, when development agencies sought to engage disaster response as an optimal occasion for introducing new ideas, technologies and methods in construction (Lyons 2009). Later BBB was formally integrated by United Nations into the Sendai Framework for Disaster Risk Reduction (2015–30), to which Nepal is also a signatory.⁸ Post-2015, the Nepali state and its donor agencies officially adopted BBB principles, with the former assuming a particularly nationalist tone emphasising the self-respect and dignity of people who had been portrayed around the world as poor, helpless victims (NPC 2015)—a common trope for presenting disaster-affected people in the Global South (Bankoff 2001/02). The formulation has also been widely critiqued as enabling neoliberal policies that ultimately produce inequality through the universalising application of a presumed shared definition of ‘better’ (Cheek & Chmutina 2022).

A second interpretive trope resides in a blend of Marxist and Foucauldian analysis popularised by Naomi Klein’s trenchant critique of ‘disaster capitalism’. In The Shock Doctrine (2008), Klein builds on Marxist insights about alienation to suggest that capital utilises disasters as fertile ground for its own reproduction. More specifically, surplus capital preys upon local means of production and subjugates hapless disaster victims to processes of accumulation by dispossession (see also Loewenstein 2015). It does so with the aid of neoliberal governmentality, a Foucauldian concept, as state regulation imposes the ‘shock therapy’ of conventional supply-side adjustments, such as privatisation and deregulation.

The authors recognise the unmistakable imprint of disaster capitalism in the vigorous lobby of contractors and the Federation of Nepalese Chambers of Commerce and Industry (FNCCI), which aimed to consolidate an official reconstruction modality rooted in contracts to private contractors for house reconstruction. The lobby claimed that contractors could complete safe house reconstruction within two years on a mass scale, in a cost-efficient manner, by building prefabricated houses off-site, then ‘connecting’ them to specific sites of reconstruction.⁹ Ultimately, however, the Nepali state rejected that approach and instead sought to directly manage the circulation of engineering expertise and, in so doing, keep markets at bay.¹⁰ NRA engineers were to forge direct lines of contact with Nepali publics, mediated only, in some instances, by NGOs, themselves operating under the auspices of a single state-led, owner-built housing reconstruction programme.

A primary concern with construction as a practice and domain of expertise requires paying attention to the subjectivity of the engineer. A disaster-capitalism (or ‘disaster statism?’) framework might characterise the engineer as a villain, whose task is to ‘render technical’ the hegemonic rationalities of post-disaster reconstruction, and disregard the specificities of local knowledge, much like the planners, architects and engineers in James Scott’s Seeing Like a State (1998).¹¹ BBB, meanwhile, evokes the trope of engineer as benign fixer. The present authors seek a middle way between these two approaches by exploring how differently positioned people deploying technical knowledge became interpolated in state-led reconstruction, and exercised expertise as they interpreted, adapted and, in some cases, subverted official engineering paradigms. Treating expertise either as solely an instrument of governmentality or as a liberal vector of modernity, both leave something to be desired. In fact, both manifest a fundamentally top-down vision of power in different ways. The emphasis here is on the engineers as lower-level state actors who actually produced state practice through their everyday actions. This reflects a Gramscian vision of social reproduction that emphasises how ideology and practice are intertwined through multidimensional practices of negotiation. In lieu of a focus on discourse or disembodied knowledge, the paper engages engineering subjectivity as an everyday practice through which people experience and in turn express themselves as subjects engaged in the work of reconstruction.

The concept of ‘negotiating expertise’ aptly characterises these dynamics. The 3000 NRA engineers deployed by the Nepali state across the earthquake-affected areas engaged the fluid, complex settings in which they were required to communicate at once with people and implement reconstruction programmes. Earthquake victims and programme beneficiaries, too, navigated a dynamic socio-cultural, political and post-disaster terrain and engaged NRA engineers in varied ways as a local state presence. Outcomes are contingent on such dynamics, as anthropologist David Mosse emphasises in an article on relational knowledge in international development:

The present research builds on this assertion (widely referenced in critical development studies) by emphasising ‘negotiating expertise’, which similarly rejects a binary interpretation of knowledge—as expert versus local—to underscore instead its relational character. Differently situated actors bring multiple and context-specific sources of knowledge to bear on their own practices of re-construction.¹² The lived outcomes in any given reconstruction setting reflect a balance of these perspectives. Foregrounding negotiation allows consideration for local social potential and future imaginaries; it points to diverse expressions of political agency that could in turn inform equitable approaches to disaster preparedness.

Social scientific scholarship on engineering expertise in particular has tended to focus on its role in advancing modernist discourse and statecraft. A study on the mediating role of engineers in advancing satellite technology across Indonesia in the 1990s is instructive in this regard. Anthropologist Joshua Barker chronicles how Indonesian engineers travelled the country constructing an ideology of industrial and technocratic nationalism through controlled nationalist messages about the transformative effect of satellite technology (Barker 2005).

Ethnographic observations in the context of earthquake reconstruction in Nepal paint a very different picture in which official (NRA) engineers in fact lacked technical expertise, formal power and a willing constituency—and were not in control of ideology. They resemble more the engineers in a study of road construction in Peru by anthropologists Penny Harvey and Hannah Knox (Harvey & Knox 2015: 8). Here, engineers are depicted more ambivalently as holding in tension an ambition for order and stability with a:

Probing this ambivalence draws out the ‘anxieties and internal critiques that have always been integral to modernist thinking’. It also points to a craft dimension of engineering practice, how ‘craftiness informs the ingenuity of the engineer’ (Harvey & Knox 2015: 8).

The analytical lens of negotiation aptly captures engineering craft, as it encompasses encounters across multiple knowledge systems. The aim is thus not just to challenge the isometry of engineering subjectivity and technical knowledge, as Harvey & Knox (2015) advocate. The intention is also to explore how engineering craft is pursued by village residents who negotiate multiple, contradictory relations in earthquake reconstruction—and thus how local knowledge informs engineering subjectivity. Sections 4–6 below present ethnographic findings on overlapping knowledges of NRA engineers and village residents involved in reconstructing their homes. In so doing, the binary frames of engineering versus local knowledge can be expanded to highlight the relational negotiation of expertise through which reconstruction unfolds.

3. INSTITUTIONAL LANDSCAPE OF RECONSTRUCTION

When the earthquakes struck, a protracted process of post-conflict restructuring was still underway after a 2006 Comprehensive Peace Agreement ended Nepal’s decade-long civil conflict between Maoist and state forces, and mandated a transition from unitary monarchy to federal republic. Owing to donor pressure, a constitution was fast-tracked as a condition of accessing funds pledged during an international conference on Nepal’s reconstruction in June 2015 (in the amount of US$4.4 billion). Subsequently the NRA was established, a full eight months after the earthquakes. A National Reconstruction Act (2015) (Government of Nepal 2015) had given the NRA extraordinary jurisdiction to oversee the Housing Reconstruction Program (and its urban counterpart). NRA responsibilities included issuing house design catalogues, training engineers to provide technical assistance to house owners, assessing owner compliance to ensure the safety and quality of reconstruction, disseminating house reconstruction grants, providing masonry training to support local reconstruction capacity, and managing the redress of grievances.

Institutionally, the NRA coordinated grant disbursement and grievance redress with the then Ministry of Foreign Affairs and Local Development (MoFALD)¹³ through its Central Level Project Implementation Unit (CLPIU) and District Level Project Implementation Units (DLPIUs). Technical assistance and inspection of housing reconstruction was coordinated with the Department of Urban Development and Building Construction (DUDBC) (under the Ministry of Urban Development—MoUD), which also operated through the CLPIU and DLPIUs. Despite the heavy donor involvement in funding reconstruction, and eagerness to engage directly in reconstruction activities, the NRA maintained tight control in order to maintain the integrity of its owner-driven model of house reconstruction within a BBB framework. To do so, the Government of Nepal instituted a one-window policy through which all funding had to be channelled, as well as retaining sole responsibility for monitoring, quality control and disbursing house reconstruction grants (NPC 2015; NRA 2016). International/non-governmental organisations (I/NGOs) were permitted to operate in supportive capacities in jurisdictions approved by the NRA. Sindhupalchowk District was an early focus of I/NGO activity at the relief stages, given its heavy losses and relative accessibility from Kathmandu, with a more extensive road network than many other affected districts. The site in Borang, Dhading District, was selected because of the presence of NSET-Nepal, an NGO focused specifically on reconstruction, specialising in earthquake engineering and preparedness. With funding from USAID, NSET had been assigned to implement its Baliyo Ghar Program in Dhading and three other districts, with the mandate to provide technical assistance to beneficiary households and provide masonry training (to skilled and lay local builders). NSET’s technical assistance teams included ‘social mobilisers’ tasked with facilitating communications among beneficiaries, NSET technicians, NRA engineers and local government officials.¹⁴

Among the unanticipated impacts of the 2015 earthquakes has been the circulation of new forms of expertise and large numbers of experts in rural areas of Nepal. Much of this flurry of activity centred on the NRA’s deployment of the 3000 ‘trained engineers’ to oversee affected households as they built ‘earthquake-resilient’ houses according to prescribed design catalogues.¹⁵ The NRA engineers’ role was to inspect house reconstruction as a condition for approval of subsequent tranches of housing reconstruction grants. All households were assessed for damage. In order to access the housing reconstruction grant to which all earthquake-affected households were entitled, homeowners had to enter into a Participation Agreement with the local government, stipulating their commitment to rebuild their houses as per prescribed earthquake-resistant designs. Through this process, homeowners were issued a colour-coded card indicating the extent of damage. All households in Borang and most in Kartike received red cards indicating ‘fully damaged’ and warranting full grants of NPR300,000. The first tranche of the reconstruction grant (NPR50,000) was provided after the Participation Agreement was signed. To receive the largest, second tranche of NPR150,000, a household had to demonstrate the completion of a four-foot foundation, which is when the plinth in the opening narrative was checked. The third tranche (NPR100,000) required approval of the roof and installation of a toilet or an alternative energy source (Limbu et al. 2022). Specific conditions for approval as well as deadlines for completion continually evolved as the NRA confronted delays and the challenges of owner-driven reconstruction.

Two key points must be made about the institutional landscape before proceeding to ethnographic findings. First, the simultaneity of earthquake reconstruction and devolution to new local governments in 2017 as part of the process of post-conflict federal state restructuring created opportunities for greater local control appropriate to the vision of ‘owner-driven’ reconstruction. But it also, as Shneiderman et al. (2023: 504) have pointed out:

Second, the deployment of NRA engineers into the countryside marked a profound shift from an unregulated to a regulated space of rural construction. Both these tensions manifested in encounters among distinct building and design knowledges—especially between NRA engineers and homeowners.

4. WHO IS AN ENGINEER?

This section elaborates engineering subjectivities and knowledges in practice. Official engineering paradigms are considered, ranging from university curricula to the experience of NRA engineers, and to the traditional engineering expertise of homeowners.

In general, Nepali society may assume that university-educated engineers are well-qualified to make professional judgments and design decisions for earthquake-resistant construction in all contexts. This premise is belied by a quick look at engineering curricula.¹⁶ A four-year civil engineering degree is comprised of mathematics, engineering science and design courses covering a range of subdisciplines, including structural engineering, hydraulic design, pipe and sewer lines, transportation engineering, irrigation and soil engineering. The curriculum gives priority to urban infrastructure and concrete-based structures. For example, fourth-year elective courses on ‘The Earthquake Resistance Design of Structure’ (CE 76501) and ‘Vulnerability Assessment and Retrofitting Techniques’ (CE 76504), along with a third-year core course on ‘Concrete Technology and Masonry Structure’ (CE 603), all provide theoretical and practical knowledge on reinforced concrete and steel design characteristic of urban structures. None of these courses, or others in the structural engineering subdiscipline, covers traditional, rural mud-mortar housing structures (IoE 2015). They also cover only the analysis and design of buildings for normal loading conditions that do not include earthquake loading, which requires advanced concepts and design tools. No significant curricular changes have been made since the 2015 earthquakes. Only two electives relate engineering to ‘society’ and ‘community development’ (CE 753 and CE 72513).

These details are mentioned for two reasons. First, to indicate the ethos of the official engineering expertise informing reconstruction policy; and second, to indicate the limited scope of official engineering expertise in the person of most NRA engineers. Regarding the former, official engineering expertise dominated the formulation of reconstruction policy. The DUDBC, a permanent technical department of the MoUD, drove the specification of standards and guidelines for house reconstruction. It is composed primarily of engineers: 11 of 16 personnel, with the remaining being remote sensing, geographical information system (GIS) and statistics experts. The NRA also struck a technical committee to review the design catalogues and manuals issued by the DUDBC, and to develop ‘exception’ and ‘correction’ manuals in response to chronic non-compliance with prescribed house designs in the field (NRA 2017a, 2017b). Its members were drawn from the formal engineering community, including the DUDBC, NGOs such as NSET and private engineering consulting firms. In other words, official engineering knowledge about rural house reconstruction was exclusively formulated at the centre for delivery to affected communities.

NRA engineers were typically fresh graduates of four-year undergraduate B.Eng. programmes.¹⁷ They did not have official tools until November 2016 when the NRA published its ‘inspection guidelines’ (NRA 2016). Even then, the interlocutors in Borang shared that the NRA engineers did not visit regularly until after local elections, in April 2017. By this time, there were locally elected ward chairs (with the ward being the smallest unit of local governance within each rural municipality) in place to verify attendance—indicating that many grants were not distributed until this time. Being more accessible by road, Kartike was more closely subject to the official schedule.

Before being deployed, engineers were given three days of training. As earthquake engineering requires far more extensive exposure to specialised knowledge and advanced concepts, the training implicitly reduced the role of the NRA engineers to checking boxes in the field. They were simply not equipped to handle site-specific problems or advise homeowners about the technicalities of their house reconstruction. As the reconstruction programme unfolded and more engineers were deployed, some engineers went to the field without the three-day training, as recounted by DLPIU officials in both Dhading and Sindhupalchowk. They faced difficulties even with simple bookkeeping tasks and communication with homeowners about securing approval for subsequent tranches of the house reconstruction grant. There were in fact so many NRA engineers in the field that it became difficult for the DPLIU to keep track of their practices, as indicated by the comments of a DPLIU official in Sindhupalchowk regarding criteria for the third tranche of funding:

This DLPIU official meant that in practice NRA engineers had considerable discretion to manage and articulate their own engineering expertise in relation to the local setting. In some cases, this may have benefited owners reconstructing their houses, while in others it may have created additional barriers for beneficiaries.

Subjectivities expressed by NRA engineers in Borang and Kartike were hardly limited to the technical dimensions of practice. On the contrary, the engineers interviewed expressed frustration about the untenable positions in which they had been put vis-à-vis homeowners. Their limited role as inspectors of owner-driven housing reconstruction was compromised by the updates and changes to house design guidelines made by the CLPIU. Most impactful for the NRA engineers in Borang and Sindhupalchowk was the Corrections/Exceptions Manual for Masonry Structures, released in mid-2017 (NRA 2017a), which allowed more rural houses to qualify for grants. Reconstruction progress had stalled because many households were not meeting minimum technical requirements for the second and third tranche disbursements. The corrections manual identified corrective measures for some structural non-compliances, after which NRA engineers could waive certain details in order to approve the next disbursement. In fact, several ‘corrections manuals’ were issued aiming to expedite reconstruction.

For NRA engineers, the cycle of revisions challenged their relationships with house owners, for whom they were supposed to provide support. Engineers expressed that communicating about the changes was ‘difficult’ and ‘created mistrust’. In Borang, as elaborated below, most householders opted for the least costly and simplest one-room baliyo ghar design promoted by NSET. However, in Sindhupalchowk, where a wider range of materials was available, along with more labour capacity, there was more variability in the earthquake-resilient designs homeowners adopted. In this context, the NRA engineers had required some homeowners to conform to design requirements at great cost, labour time and sacrifice of housing preferences, which were subsequently relaxed through the corrections manual. An NRA engineer in Kartike explained:

Engineers expressed remorse and humility at the profound hardship caused by changes in the design guidelines, and their complicity in the unpredictability of the state. As another engineer put it in an interview in Kartike:

In Sindhupalchowk, some NRA engineers reported feeling threatened by homeowners’ anger. Their primary task was to assess house construction, and they were to disqualify those structures failing to meet government-prescribed designs. Of course, homeowners would have invested significant resources and sometimes did not respond generously to being disqualified. As one NRA engineer in Kartike put it, ‘such insults and frustrations, we get a daily dose of it’ (Limbu et al. 2022: 19). Some engineers reported that fear of violence and insult were more intense as the house reconstruction programme was first being rolled out, when people did not understand the key role of the NRA engineers in a national effort to get earthquake resilient houses built by owners. Interactions sometimes turned not only threatening but also coercive and violent. A woman from a village near Kartike recounted villagers’ hostility if engineers did not oblige their demands:

An NRA DLPIU official at Dhading Besi, the district headquarters, also reiterated concerns about the safety of their field engineers:

The NRA engineers in Kartike acknowledged that they sought protection from local political leaders against their fears and threats of violence. For example, the ward chairperson would accompany the engineers for house inspections and helped rebuff the pressure or threats from community members (Limbu et al. 2022: 22). Engineers were reported to stay in the homes of political leaders during their fieldwork. It was evident that these arrangements also benefited politicians seeking to build or maintain their political networks. As the DLPIU official indicated:

Thus, engineers became involved in the dynamics of political patronage and influence. Given the demands, combined with the lack of training and experience (both technical and socio-political), reports of a high turnover rate among NRA engineers came as no surprise.

The NRA engineers, inadequately supported by skill sets acquired in their formal engineering education, had to develop expertise outside the scope of technical realms in order to negotiate with homeowners, politicians and officials based in the district centre. They routinely found themselves in situations outside the realm of their training and experience. In this sense, their experience echoed that of homeowners, themselves navigating novel, frustrating and frightening situations as they sought to reestablish their homes in negotiation with unfamiliar forms of engineering expertise. The engineering expertise of homeowners was rooted in traditional knowledge of the available resources and requirements of agrarian and mixed-livelihood strategies in rural Nepal. After all, before the earthquake, houses had been constructed by their inhabitants, using locally available stone, wood and mud mortar for the most part (the few reinforced concrete buildings in Kartike would have required specialised labour).¹⁸ Within these parameters, people adopted house designs to meet the needs and preferences of the occupants, including family size, ethnicity, conceptions of kinship and religion (see Shneiderman et al. 2023 for an extended reflection of conceptions of house, home and household in the research sites). In both Kartike and Dhading, houses were typically built to two to three storeys to accommodate agricultural livelihoods (e.g. grain storage and processing, sometimes livestock on different floors), and in the case of Kartike, sometimes accommodating space for a storefront or other business on the ground floor. Construction knowledge was shared, passed down over generations and commonly put into practice through collective labour exchanges across households.

After the earthquakes, households repaired their own houses or built temporary structures depending on the extent of the damage. Homeowners were initially unaware of support from the government. They designed the repair, performed their own construction labour, and often had help from neighbours (in the form of loans) and family (remittances) long before the government’s formal reconstruction programme launched or NRA engineers visited their villages. Given the predominantly agrarian livelihoods, living in a temporary shelter was not feasible as space was needed above the hearth to store harvested foodgrains. Repairs thus had to be completed within one agricultural cycle after the earthquake. Many families repaired their houses using stone walls on the ground floor, lighter materials such as wood or corrugated galvanised iron (CGI) sheets on the upper level, and CGI for roofing. These self-repaired houses were usually smaller than multistorey pre-earthquake houses, but larger than house designs that were later prescribed by the NRA. They retained features that suited villagers’ specific lifestyles and family configurations, such as a storefront shutter, separate rooms to accommodate family members, grain storage and space for altars for worship. This kind of knowledge, which is recognised here as expertise, certainly did not inform NRA specifications for what qualified as a ‘house’.

5. WHAT IS A HOUSE?

By official Housing Reconstruction Program metrics, a house foremost has a household head, gharmuli. A household became eligible for the reconstruction grant only upon presenting legal certificates of land ownership. This provision makes sense from an administrative perspective concerned with accountability, but poorly suits lived realities in rural areas. In practice, many households had not officially divided their property among brothers residing in separate houses on land titled to their father or even grandfather, even though those separate residences were treated as the legitimate property of their occupants, according to patrilineal and patrilocal customs. In Borang, many households thus had to go through the onerous process of registering land ownership, which entailed first securing seven witnesses who already had a beneficiary card in the same ward, and then making the two- to three-day walk to the District Land Registration Office (DLRO) to initiate a multi-step process (often requiring repeat visits) which would legally align house ownership with property ownership. This was all before making the journey to a bank branch—also at the district centre—to receive grant disbursements after families had already repaired and resettled in their houses.

In Sindhupalchowk, although the journey was shorter for Kartike residents, the DLRO was reportedly flooded with people seeking land partition and retitling. Multiple trips were required to navigate the months of overcrowding. Here, in addition to those families seeking to align their legal and everyday lived realities, others sought to ‘game the system’ by fragmenting their families ‘on paper’ in order to become eligible for multiple grants (Limbu et al. 2022). Opportunities to benefit in this way seemed to be mediated by a particular family’s connections to political leaders, who could facilitate family partition in exchange for various patronage obligations.

In official terms, a ‘house’ also conforms to a design specified in the Catalogues and Correction/Exception manuals issued by the NRA. As these were revised, over two volumes and the series of corrections manuals, the scope of what could constitute an earthquake resilient house expanded. It must be acknowledged that in addition to resisting the pressure to engage a contractor-driven modality for house reconstruction, the NRA, for all its disciplinary engineering centrism, responded iteratively to realities on the ground. The practices of owners—reflecting their capacity and interests in relation to the available resources—indirectly informed official house designs, most notably in expanding variability and acceptable alternatives. However, as shown, from the perspective of owners, the rounds of revisions were not experienced as a feedback loop engaging local knowledge, but rather as an arbitrary, inflexible and disciplinary device unmoored from village lifeworlds. Engineers critical of the reconstruction process, including co-author Pandey, might rather characterise the official revisions as a pragmatic move to advance the reconstruction programme and facilitate the role of NRA engineers in compelling homeowners to follow the approved designs.

How did house owners negotiate NRA engineering expertise and designs? The research team observed strained relationships where local people largely viewed the engineers with mistrust and suspicion as outsiders. The latter were reported to be mostly absent from the villages and only conducted short visits, focused on checklists and inspections (Rawal et al. 2021: 19). Where the engineers continued engagement with the community, a sense of familiarity and relationship developed that benefitted both parties in the reconstruction process (Limbu et al. 2019). Where there was lack of familiarity, people usually found communication with the NRA engineers confusing and inconsiderate for their lived realities, and the series of design revisions disrespectful of their efforts to conform to a process already experienced as constraining during a time of immense hardship. Despite the revisions, the overall framework for the designs had neglected to account for the specificities of agrarian lifestyles and rural economies, and attempts to accommodate the material manifestations of these realities came too late to be meaningful.

In Borang, NSET-Nepal promoted the one-room house design built with stones and mud-mortar, which households found utterly inadequate for their lifestyle needs. Several factors contributed to households selecting this model, as was explained by an NRA engineer in Borang, demonstrating compassion for and understanding of residents’ predicament (Rawal et al. 2021): the greater cost of building a larger house from the catalogues; the relatively simple design and locally available materials; the lack of availability of trained masons to support construction; and, foremost, a condition that the release of the second grant tranche was contingent on owners completing the foundation, one window and one door of their earthquake-resilient house.

A similar process was repeated at the disbursement of the third tranche. Once the house reconstruction was completed, since people were already living in their repaired pre-earthquake homes, many households in Borang opted to lock up the new one-room ‘earthquake house’. As noted by Rawal et al. (2021: 17):

The second tranche disbursement was also when NRA engineers checked the first wooden band at the house’s plinth. Homeowners reported understanding that untreated utis (locally available alder) wood generally used for firewood would quickly decay when it came into contact with water, especially at the foundation level, but were not motivated to apply the required pine tar consistently because their main purpose in building the one-room house had been to secure the second tranche (Rawal et al. 2021: 17).

In Kartike, Sindhupalchowk, the same dynamic was evident—of householders doing the minimum construction of earthquake-resilient houses to qualify for the second and third tranches. Built form differed in that people constructed the new houses according to prescribed designs right alongside the reconstructed old houses. They also adapted the designs discretely to build hybrid houses, combining stones with brick–masonry, cement mortar and iron-truss frames (Social Science Baha 2022)—resulting in a similarly ‘patchwork built environment’ (Shneiderman et al. 2023: 520). Houseowners explained that in Kartike and surrounding villages the significant impediment to grant disbursement approval came at the third tranche, when toilet construction was an option for meeting the funding criteria.¹⁹ This criterion irritated many people who already had a functioning toilet, outhouses usually, that had survived the earthquakes, or who were simply using the toilets of their neighbours or relatives if their own had been damaged. Homeowners reported tearing down old toilets that were categorically deemed inadequate by the NRA engineers to build the prescribed new ones for the sole purpose of accessing the third tranche. Many indicated plans to consign their new toilets for storage space. As noted in the project report on Sindhupalchowk District:

A key strategy deployed by some homeowners in securing their grant instalments was to leverage political patronage. Those with connections to political parties might have their old toilet passed if local politicians accompanied the NRA engineers on their inspections. The DLPIU official in Sindhupalchowk acknowledged that housing guidelines could be relaxed to curry political favour. Referring to newly elected local government officials, he stated:

To answer the question ‘What is a house?’ therefore requires recognising how people in Dhading and Sindhupalchowk distinguished between the new earthquake-resilient houses—to which they referred as ‘earthquake house’ (bhukumpa ghar) in Dhading and ‘grant house’ (anudanko ghar) or ‘government house’ (sarkarko ghar) in Sindhupalchowk—and their ‘own house’ (aphno ghar), which might translate more aptly as ‘home’, the organically repaired houses damaged by the earthquakes with no intervention from the NRA. As Shneiderman et al. (2023: 514) have argued, a home denotes not only a material structure but also an ‘affective site of belonging’ consistent with local livelihoods and world views. People sought to negotiate NRA engineering expertise and designs in a way that would preserve the integrity of their own houses, while putting minimum effort and cost into building ‘earthquake houses’ that would secure their right to financial support. In this sense, the ethos of ‘owner-driven’ house reconstruction seemed to have ‘lost its essence’ (Shneiderman et al. 2023).

6. WHAT IS SAFETY?

This distinction between house and home informs the discussion of safety. An official engineering rationality would posit, ‘if it is not safe, it is not a house’, denying the category ‘house’ to any structure built with people’s own traditional knowledge before the earthquakes. The latter were categorically deemed vulnerable and unsafe, and houses built or repaired by people without consultation with the NRA engineers were automatically disqualified, compelling them to build new ‘earthquake houses’ to conform to official standards of safety. Many residents acknowledged, along with official engineers, that the compact one-room house is the most stable structure, and the most apt to secure physical safety. But they also recognised that engineers had failed to provide designs based on functional requirements in rural areas, or suitable for cultural and social needs. Engineers were not adequately trained, did not have the authority and lacked any incentive to develop alternative solutions based on the full range of homeowners’ needs. The single lens of structural resilience informed engineers’ conception of safety.

Alternative conceptions of safety have been articulated in the literature on health equity, with a particular focus on ‘cultural safety’ needed to redress Indigenous and ethnic health inequities (Curtis et al. 2019, 2025; McGough et al. 2022). Critical health scholars and practitioners formulated the concept of cultural safety in the 2010s as a counter to the prevailing concept of cultural competency, in order to emphasise critiques of structural racism and normalised power structures within the health systems in the US, Canada, Australia and New Zealand. If cultural competency requires becoming ‘competent’ in the cultures of others, ‘cultural safety’ by contrast emphasises self-reflection and self-awareness on the part of health practitioners as a basis for challenging the racist assumptions embedded in their own cultural frameworks (Curtis et al. 2025). Australia and New Zealand have been particularly proactive at embedding cultural safety into educational curricula and regulatory standards for health-related professions, particularly nursing and midwifery.²⁰

The focus in such conceptions of cultural safety is procedural, seeking to decolonise the health profession by building a critique of power dynamics endemic within settler societies. The study of earthquake reconstruction aligns with this impulse and also contributes more substantive considerations, mostly having to do with expanding the frame of engineering expertise to encompass local knowledge and the experience of front-line engineers. It is contended here that local knowledge encompasses not just engineering expertise by which people in rural Nepal have constructed their houses for generations, but also the affective ties they build over time and over generations with their homes. Such affective ties can be framed as a sense of ownership or belonging rooted in the role of a home in forging identity (Besky 2017, cited in Shneiderman et al. 2023: 508). People did not want to abandon their damaged traditional homes because they had evolved in relation to livelihood strategies, kinship structures, cosmologies and the material resources of the families who inhabited them (Shneiderman et al. 2023: 504). Pre-earthquake homes carried memories and suited residents’ housing practicalities, which were valued more by many residents than the physical safety promised in the designs of the one-room baliyo ghar and other earthquake-resilient designs. In Borang, residents who had substantial income and resources to spare reported spending NPR400,000–800,000 on repairing and rebuilding their homes, compared with the NPR300,000 reconstruction grant intended to finance the one-room house.

Prescribed house designs had the effect of weakening affective ownership by redirecting people’s scarce time, resources and energy toward building houses lacking qualities necessary for becoming their day-to-day home. Thus, a conception of cultural safety in disaster reconstruction is an important ingredient that encompasses fundamental cultural competencies about people’s subjective experience of ownership, belonging and citizenship. For the case of housing, that would require understanding what it takes to make an earthquake house a home (Shneiderman et al. 2023). The presented ethnographic work points to the imperative of ample space to accommodate the appropriate configurations of household members and visiting family, and appropriate space for the agricultural livelihoods and small-scale enterprises characteristic of rural areas.

Cultural safety must protect not only residents but also front-line engineers and other officials tasked with implementing reconstruction programmes. If the protocols official engineers are deployed to enforce are themselves culturally unsafe, engineers too are put at risk. The authors witnessed and learned about NRA engineers being sent to new localities with inadequate experience, knowledge and training; encountering unruly publics; and navigating the confusion and chaos through frustration and fear. For both engineers and residents, negotiations across official and local knowledges generated new forms of knowledge, which informed subjective experiences of expertise and ownership going forward. These negotiations, however, sit in ironic relation to the moniker of ‘owner-driven’ reconstruction.

7. CONCLUSIONS: WHAT IS RECONSTRUCTION?

Framing reconstruction as a negotiation across forms of knowledge and expertise circulating within the pragmatics of house construction has furnished an analytical frame for rethinking ‘what is a house?’, ‘who is an engineer?’ and ‘what is safety?’ Querying these often taken-for-granted terrains of reconstruction in turn reveals the relational character of knowledge, rooted in the conditions of its production and evolving through encounters with other knowledge systems during processes of disaster reconstruction. In the context of a reconstruction programme that faced many challenges in achieving the laudable goal of ‘owner-driven’ reconstruction, the aim is to eschew polarising judgments and situate disaster reconstruction within wider political–economic and institutional contexts from which intersubjective negotiation of expertise unfolds. This will complicate prevailing narratives that a narrowly materialist orientation to reconstruction can ‘Build Back Better’, or that it merely furnishes a ground for accumulation through ‘disaster capitalism’. An expansive understanding of expertise acknowledges a house can be both structure and affective belonging, and safety can encompass a cultural dimension that accounts for a sense of belonging mediated through kinship, livelihood and cosmology. As such, expertise manifests in local knowledge as much as in official engineering knowledge, which in practice morphs on the ground in relation to situated contexts of disaster reconstruction.

In tracing these negotiations across multiple systems of knowledge, this approach aligns with that of Harvey & Knox (2015) and Shneiderman et al. (2023), who emphasise the openness and possibility forged through negotiation of the spaces and subjectivities emerging from pragmatic accommodations in construction.

Normative points about earthquake reconstruction can be drawn as reflections that could be relevant to planners and engineers in other post-disaster conjunctures. Foremost, those engaged in formulating reconstruction policy could actively anticipate the kinds of questions posed here about engineering knowledge and subjectivity as a key dimension of mitigating the negative effects of future disasters for those most affected.

More concretely, a more open stance toward expertise could point to the imperative to consult locals about building design rooted in principles of both structural and affective integrity. Procedurally, an open stance might take the form of reflecting on the limits of official engineering expertise (as in the cultural safety articulated in health professions; Curtis et al. 2025); expanding the flow of ‘bottom-up’ communications beyond grievances to more open-ended venues for sharing knowledge and reflecting on the root causes of problems in implementing official policy; and developing protocols for ensuring that genuine feedback loops are integrated into processes of policy formation.

Substantively, an open stance points in the direction of cultural competency—having reasonable knowledge of the affective dimensions that make up a home in any given context, in addition to the structural dimensions of physical safety. Official building codes and housing designs should be flexible enough to accommodate:

Creating a system to track peoples’ design preferences and availability of local labour, materials and skills could go a long way toward building this kind of cultural competency. This can also be done in the ‘disaster interval’ between major events, so that planning authorities already have access to the relevant information when new disaster strikes (Nihei 2012).

A commitment to cultural competency also requires revising curricula in engineering education. Undergraduate and graduate engineering programmes can include courses on cultural safety, including cultural competency, in their core curricula, supported with full-time and contractually secure appointments of faculty with training in the social sciences. The training of front-line engineers tasked with implementing reconstruction policy can encompass exposure to evolving ‘findings’ from iterative knowledge-sharing between locals and government agencies—if not enrolment of trainees in implementing feedback loops that recognise local engineering expertise and dimensions of affective belonging. A corollary commitment might be to reconsider how reconstruction programmes have forged boundaries between the labour of official engineers and that of social mobilisers charged with the more nuanced dimensions of communication and coordination at the local scale. Culturally competent engineers might be expected to adopt more expansive roles.

The questions involving how expertise is interpreted fundamentally drive material and social processes of change originating in moments of disaster. The stakes may extend far beyond that immediate temporality, to reshaping communities in ways that weaken affective ties to tradition while also mitigating opportunities for transforming state–society relations in the wake of political restructuring. An expansive interpretation of expertise not only elevates the significance of local social potential in disaster response, but also more generally builds toward procedurally and substantially just modes of planning and development.

DATA ACCESSIBILITY

Preliminary reports and related publications are available at: https://elmnr.arts.ubc.ca/. Interview transcripts and observation notes are not publicly available, in accordance with institutional ethics board policies.

COMPETING INTERESTS

The authors have no competing interests to declare.

ETHICAL APPROVAL

The research design for ‘Expertise, Labour and Mobility in Nepal’s Post-Conflict, Post-Disaster Reconstruction was approved by The University of British Columbia Behavioural Research Ethics Board (reference number H17-03029).