For the past 2 decades, development of the road network in Nepal has topped community and government priorities, a trend likely to continue as the country transitions to a decentralized federal government (Rankin et al., 2017). In parallel, China's new Belt and Road Initiative (BRI) offers the promise of investments in key infrastructure: expanding trunk roads, hydroelectricity, trade and development (Wire, 2017; Economist, 2017). Yet as Nepal devolves significant power to local and provincial administrations, it is uncertain whether the newly formed local administrations will rise to the challenge of establishing safeguards to ensure that promised benefits outweigh potential losses. We suggest that the issue of poor roads in Nepal is a political not a technical issue and one for which better service and less environmental damage could both be significantly addressed through improved governance. This commentary points to the need for improved road governance based on research, consultations and observations of road construction and associated landslides in Nepal and around the world (e.g., Blaikie et al., 1980; Sidle et al., 2012; Jaboyedoff et al., 2016; Rankin et al., 2017). It also highlights the need for more scientific studies on the topic as most relevant publications emanated from the grey literature, government publications or media articles.

Roads are globally accepted livelihood links for communities in rural areas. By reducing travel time on foot, opportunities are opened for quicker transportation of goods and better access to employment, education and health (Hettige, 2006; Bryceson et al., 2008). Roads generally create direct and indirect benefits to rural populations, directly through employment in constructing and maintaining them and in providing rural transportation services. Indirectly they provide opportunities for marketing goods and services, flexibility for employment and roadside businesses, and opportunities for transporting agricultural products to markets (Bryceson et al., 2008; Iimi et al., 2016). They can provide a safety net of sorts in generating alternative livelihood opportunities, especially in circumstances in which conditions for agriculture are difficult. In general, connectivity is thus positively correlated with lower poverty rates (Hettige, 2006; Iimi et al., 2016). Additionally, there are many nonmonetary benefits of roads, especially greater access for the poor to health and other public services, such as education, which can significantly reduce vulnerability and even gender inequality (Starkey et al., 2013). In Nepal, roads are also linked to the current boom in migration, facilitating easier mobility to both near and distant migration destinations (Jaquet et al., 2015; Upreti and Shrestha, 2016). Finally, a robust road infrastructure can provide vital corridors for evacuation and rescue in the aftermath of disaster.

However, benefits of roads need to be weighed alongside evidence that roads may benefit nonpoor households more, perhaps making development less even (Hettige, 2006). Furthermore, other impacts, such as increased environmental hazards, pollution, crime and unwanted cultural influences, are often overlooked (Blaikie et al., 1976; Hettige, 2006; Jaboyedoff et al., 2016; Murton, 2016). This paper builds on research and publications questioning the aspirations of the government of Nepal as early as the 1970s and 1980s. The Overseas Development Group at the University of East Anglia pioneered studies to understand short-, medium- and long-term effects of road construction on spatial and socioeconomic inequality (Blaikie et al., 1976; Rankin et al., 2017). Blaikie et al. (1980) revealed the inequalities created by road construction in Nepal, with loss of livelihoods for those without possibilities to invest and enhanced opportunities for those who could (Rankin et al., 2017).

The work presented by Blaikie et al. (1980) was conducted during the same period as the Laban (1979) benchmark inventory of landslides in Nepal to document the number of landslides and their origin as either natural or human-induced. Although roads represented a small proportion of total land area at the time, Laban warned that as the road network continued to expand, the number of landslides would “increase drastically in the near future, especially if more careful construction methods are not undertaken” (Laban, 1979, p. iv). Both research projects were widely influential and according to Rankin et al. (2017), the Blaikie et al. (1976) study may have redirected domestic budgets and foreign aid toward other rural development investments. However, this reprieve was soon to end with a greater focus on connectivity in the 10th 5-year plan (2002–2007) and the boom in foreign investments in road construction projects after 2008 and the end of the Maoist insurgency (Pokharel and Acharya, 2015). The 11th plan (2007–2010) established the ambitious goal of constructing a road network throughout the country whereby residents in the hills should have a road available within 4 h walking distance and Terai residents within 2 h (Pokharel and Acharya, 2015).

Roads in Nepal are generally classified as national roads, (i.e., strategic road network, SRN) under the jurisdiction of the Department of Roads (DOR), or local roads (i.e., local road network, LRN). The LRN is comprised of the district road core network (DRCN) and village roads (VR) under the jurisdiction of the Department of Local Infrastructure Development and Agricultural Roads (DoLIDAR) (Fig. 1). Road building started to gain momentum in Nepal with the advent of multiparty democracy in the early 1990s, intensified further after the Maoist insurgency ended in 2006 and continues to be one of the country's main priorities (DoLIDAR, 2016a; Upreti and Shrestha, 2016).

About 20 years ago, Nepal's road network was one of the smallest in the world with a road density for both SRN and LRN estimated at 13.7 km per 100 km2 in 1998 (DOR, 2002, 2017). By 2016, it had increased to 49.6 km per 100 km2 and continues to increase at a very rapid pace (DoLIDAR, 2016a). The SRN expanded rapidly from 4740 km (blacktop, gravel and earthen) in 1998 to 15 404 km in 2016 (DOLIDAR, 2016a). The LRN experienced a 1200 % increase during this period, from 4780 km in 1998 to 57 632 km in 2016, and contains the most common roads in rural areas (DoLIDAR, 2016a).

In 2007, the country spent 5.2 % of its national budget on roads, and by 2011/2012 this figure had increased to 8.0 % of the national budget or an estimated USD 491.2 million (WB-GON, 2013). The estimated investment in the LRN was about USD 245.6 million (2011/2012), of which 54 % of the rural road budget originated from donors and 20 % was soft loans to communities. Community contributions amounted to an estimated 12 % of the total budget through their own savings and remittances and earnings from community forestry (WB-GON, 2013; DoLIDAR, 2016b). This demonstrates the significance and priority given to roads and connectivity as a vector for economic development and population mobility.

Despite the budget and priority allocated to the road network, Nepal's mountain roads are in a treacherous state, subject to frequent rockfall, landslides and accidents (Singh, 2018; DOR, 2013a) (Fig. 2). According to DOR (2013a), one of the main causes of road accidents is road design, including very steep gradients, lack of safety features and poor road conditions. Local road construction or so-called “dozer roads” are most often initiated and constructed by bulldozer owners in collaboration with politicians at the request of communities, without basic grading or drainage (ITAD, 2017; Singh, 2018). The dozer roads are usually constructed or upgraded during the dry season. During the monsoon, road segments are frequently washed out because a majority of these roads lack proper engineering (WB-GON, 2013). Road failures are cleared up at high cost after the monsoon and the failure-and-clearance process is repeated for years until there is no loose soil to block roads (Leibundgut et al., 2016). Environmental impacts include destroyed irrigation schemes, buried springs and contaminated water supplies (Singh, 2018). Initial environmental examinations (IEEs) to reduce environmental impacts are usually required for local road construction but are rarely enforced (ITAD, 2017).

Such rapid and ineffective road construction throughout the country, but particularly in the middle hill and mountain areas, is placing increasing pressure on fragile ecosystems, wasting government resources, and increasing risk to road passengers and roadside dwellers (DOR, 2013a; Singh, 2018). Studies have demonstrated that roads are one of the greatest anthropogenic drivers of environmental degradation, erosion and landslides in Nepal (Petley et al., 2007; Leibundgut et al., 2016; Froude and Petley, 2018; McAdoo et al., 2018; Vuilliez et al., 2018). This situation is worsening due to the intensifying rainfall during the monsoon, largely attributed to climate change (Petley et al., 2007; Bharti et al., 2016; Devkota et al., 2018; Froude and Petley, 2018), which has led to a greater occurrence of landslides, especially in the middle hills (McAdoo et al., 2018). The possibility of an earthquake of even greater magnitude than the 2015 Gorkha earthquake (M 7.8) raises concerns about poorly designed roads increasing the likelihood of catastrophic landslides (Singh, 2018).

Nepal has a range of acts, regulations, guidelines and directives that require proper road engineering practices, various levels of environmental assessments and approval. However, while funded by government budgets, a majority of local roads do not follow established government practices (ITAD, 2017). Hence, although the legal framework for ensuring proper governance of infrastructure development is well developed with public bodies to monitor and enforce governance, the lack of political will and consensus among political leaders has undermined the impact of these bodies (WB-GoN, 2013).

As the country shifts decision-making to the provinces, it is unclear how management of roads will be affected among the main actors such as DOR, DoLIDAR (Department of Local Infrastructure and Agricultural Roads), and rural and urban municipalities. Institutional roles are shifting under ongoing reforms, with executive authority over local infrastructure development being transferred from district level authorities (district technical officer) to provincial public works departments, which are supposed to coordinate with central level ministries and departments (ITAD, 2017). At the time of printing, it is not yet clear which administrative body will have oversight of road policies and alignment of policies among provinces. The risk is that the few gains that had been achieved over the past decade, including a greater emphasis on regular maintenance of roads, become completely diluted (ITAD, 2017).

Another development which may affect the type and pace of road construction in Nepal is China's Belt and Road Initiative (BRI). In May 2017, Nepal became a signatory to the BRI with the promise of expanding several trunk roads in order to foster new trade and economic benefits (Economist, 2017). This new “Silk Road” will develop a trade and infrastructure network from China towards the west and south, including countries in Central and South Asia and eastern Europe.

The BRI has for now elicited more questions than answers, including which roads will be expanded and will they link rural mountain communities to greater economic development opportunities, better health care and education options, and increased social networks, or will the BRI trunk roads spawn more of the poorly engineered local roads with their demonstrated low cost effectiveness and high environmental impacts? Without adequate controls and support, rural villages can be expected to tie into these trunk roads by expanding the network of poorly constructed local roads, with ensuing environmental, economic and human risks associated with roadside erosion and slope failures that damage both the roads and the neighboring productive land.

Despite this bleak picture, Nepal has the governance systems in place to resolve the problem if it chooses to do so. Numerous technical manuals and departmental guidelines provide the basis for good alignment determination, careful engineering, stabilization of incipient landslides in slopes and prevention of erosion through the use of bioengineering (e.g., Deoja, 1994; DOR, 2013b). Nepal has been a world leader in the past and government agencies such as DOR and DoLIDAR all have cadres of highly trained engineers and bioengineers who could fulfill the required technical functions satisfactorily if directed properly (WB-GON, 2013; ITAD, 2017).

However, these abilities are currently ignored in the interest of political expediency and a misplaced public perception that quickly opened roads are a panacea for socioeconomic development. Institutions were established to regulate road construction. The Environmental Protection Council was formerly established under the chairmanship of the prime minister to monitor environmental impacts and to regulate the environmental and social impact assessment legal instruments (GoN, 1997) but became an ineffective facade. The Department of Roads' Geo-environmental and Social Unit is also not serving its function. Finally, political influence has overrun any efforts to instill checks and balances (ITAD, 2017), notably by the Commission for Investigation of Abuse and Authority, which was created to highlight cases of poor governance.

Yet with the revision of ministerial portfolios in 2018, the reorganized Ministry of Forests and Environment has an opportunity to ensure that statutory environmental safeguards are met by those government units that will be responsible for administering road development. Newly formed provincial administrations are now tasked with revising their master plans and have the opportunity to develop action plans to strengthen governance bodies, increase transparency and enforce regulations.

On the surface, roads are vital livelihood links for rural populations for improved access to markets, health care, education, employment and migration. Mobility is increased, rural populations can develop greater resilience to harsh environmental conditions and there are possibilities of new economic opportunities, ultimately reducing economic vulnerability. However, mountain roads, especially when poorly constructed, present particular challenges to sustainability, risk and governance (Sidle and Ziegler, 2012). Hence, the full benefits of such roads in mountainous areas should be questioned.

Finally, the issue of poorly designed and risk-filled roads in Nepal is a political rather than technical issue. As Nepal moves towards greater decentralization of power, there is considerable opportunity for its local and national administrations to turn the tide toward safer and more sustainable road development. The two new major drivers of road development in Nepal – decentralization of power and the BRI – could be harnessed to change road construction from the current trajectory toward environmental disaster to development. The high environmental and maintenance costs of haphazard dozer roads could be significantly reduced if government policies were enforced to achieve well-established road engineering designs, including basic standards of road grading, alignment, drainage and bioengineering. Nepal is at a new crossroads with fresh opportunities to rein in the dozer road constructors, but this will require concerted effort and considerably more political willpower than has been demonstrated over the last decade.