Decentralisation and urban primary health services: a case study of Delhi’s Mohalla Clinics

The Indian political party Aam Aadmi, which assumed power in the city-state of Delhi in 2015, introduced Mohalla Clinics (i.e. neighbourhood clinics) to provide free primary health services for all, as a response to the rising inaccessibility of primary healthcare facilities for the urban poor. These clinics were to be governed through Mohalla Sabhas (i.e. neighbourhood committees), which are instruments of participatory governance within the neighbourhood. The research compares promises and practice for Mohalla Clinics, especially focusing on governance and the politics surrounding it. The authors find that in their current form Mohalla Clinics are limited to providing primary curative healthcare and have shown limited success, although Mohalla Clinic users do save time and expenditure on primary healthcare, and the clinics have led to a more comprehensive form of primary healthcare than in the past. However, Mohalla Clinics are governed in a top-down fashion by the Government of the National Capital Territory-Delhi, and not by urban local bodies or the envisaged neighbourhood committees. As a result, they face problems that may inhibit their functioning in the long term.


Introduction
Cities arguably have a wider range of health and social infrastructure than rural areas. However, for countries in the global south, access to these services for the urban poor may be restricted by their inability to pay (e.g. for medication and supplies, even if health services are free), inconvenient locations or times of operation, or poor quality of healthcare services. All of these can result in low utilisation of even the most basic preventative and curative health services (WHO 2010). With rapid urbanisation, the urban poor population is also increasing rapidly. This group, in India, suffers from poor health: for example, the mortality rate (per 1,000 live births) for under-fives at 59.3 is significantly higher than the urban average of 34.4 (IIPS and ICF 2017). Consumer expenditure data shows that urban households spend 5% to 6% of their total expenditure on health, with half of this spending on primary healthcare alone (Ministry of Health and Family Welfare 2013). Whereas in rural areas it is a lack of proper medical practitioners that limits access to healthcare, in urban areas it is cost. These realities led to the neighbourhood clinics) in 2015, which aim to provide primary healthcare for free at a neighbourhood level, thereby making it more accessible. GoNCT-D also raised healthcare higher up the political agenda, including by shifting the political discourse away from physical infrastructure and towards social infrastructure (Lahariya 2017). This paper aims to understand the promises vs. the practices of Mohalla Clinics in providing the people of Delhi with primary urban healthcare, with particular emphasis on their governance and the politics surrounding it.
South Africa at 8.20% and Brazil at 8.91% in the same year (WHO n.d.). If we compare budget trends, there has been only a moderate increase in the funds allocated in the union budget, 3 such that when adjusted for inflation the situation can be viewed as virtual stagnation (Sundararaman et al. 2016).

Community participation in health
Community participation is one of the key principles of Mohalla Clinicsalong with accessibility, equity, an assured package of essential health services and defined quality standards for services (Lahariya 2017). The concept of community participation in primary health was formally introduced at the Alma Ata conference in 1978 and has been explored in detail in various studies. Rifkin (1986) reviewed 200 global case studies over a decade (1976)(1977)(1978)(1979)(1980)(1981)(1982)(1983)(1984)(1985)(1986) and drew the following lessons: 1. there is no universally accepted definition of community participation; 2. health services alone will not lead to community participation; rather a community participation framework should already exist, and health should be one of its components; 3. a 'one model fits all' approach should not be used; and 4. it is not possible to consider community participation in isolation from its political context. The present study discusses how community participation operates in practice in Mohalla Clinics, using a framework provided by Rifkin (1986) which is depicted in Figure 1. Since healthcare, particularly in its diagnostic and curative aspects, is a technical subject, and given the context in which Mohalla Clinics operate (i.e. mainly for the urban poor), the study focuses on whether community participationfor example in deciding the location and spatial distribution of clinics and their operating hourscan help improve the efficiency of their implementation.
CJLG 2020 4 Figure 1: Community participation in healthcare programmes Source: Rifkin (1986, p. 247) Literature review: history of participation and participatory governance Participation as an approach to policy-making, governance and project implementation is popular in both developing and developed countries (Hickey and Mohan 2004). The first attempt to classify levels of citizen participation and 'nonparticipation' was by Arnstein (1969), through an eight-rung ladder ranging from citizen control on the highest rung to 'manipulation' on the lowest rung ( Figure 2).

Figure 2: Arnstein's ladder of citizen participation
Source: Arnstein (1969, p. 217) Citizen participation evolved into community participation as its context changedfrom citizen participation in 1960s government schemes in developed nations to community participation in the neoliberal era in underdeveloped or developing nations (Arnstein 1969;Choguill 1996). Whereas 'manipulation' was the lowest rung in Arnstein's ladder, Choguill's ladder (1996), shown in Figure 3, breaks down community participation somewhat differentlyfor example including 'conspiracy' and 'self-management' as two other rungs, as these are prevalent in developing nations. Manipulation is

Patel & Pant
Delhi's Mohalla Clinics CJLG 2020 5 also explored further by drawing out its types: dissimulation, diplomacy and informing. Community empowerment, rather than citizen control, is at the top. It is noteworthy that Choguill's ladder focuses more on the process of development than does Arnstein's ladder.

Figure 3: Choguill's ladder of community participation
Source: Choguill (1996, p. 442) In the globalisation era, decentralisation and neoliberalisation are two sides of the same global restructuring process and are usually simultaneously launched in urban governance. While neoliberalisation heralds policies towards market-oriented economic growth, decentralisation or participatory governance purports to offer opportunities to those without exchange entitlements to participate in decision-making processes in urban governance and thereby seek rights to improve their well-being (Purcell 2003;Goodwin and Painter 1996;Patel et al. 2016a). Hence new spaces and opportunities have emerged for citizen engagement in policy formation. There are policy instruments, legal frameworks and programmes that promote citizen participation. However, it is also becoming apparent that merely creating new institutional arrangements does not necessarily result in greater inclusion or pro-poor policy change, despite the prevalent assumptions. The critical question is where power truly lies, and this can be assessed through mapping onto the 'power cube', which is a framework for investigating the spaces, places and forms of power, and their inter-relationship (Gaventa 2006).
The three levels of governments operating in Delhi with no clear demarcation of their roles and responsibilities, explained in the next section, can be best assessed through Gaventa's power cube.
Visually presented as a cube (Figure 4), in the model, each side represents a dimension or set of relationships, not a fixed or static set of categories.

Figure 4: Gaventa's power cube visualised
Source: Gaventa (2006, p. 25) Firstly, power may be exercised at local, national or global levels. Secondly, the forms that power take have been defined as visible, hidden and invisible. Visible forms of power are contests over interests that are visible in public spaces like legislatures, local assemblies which are bound by formal rules, structures, authorities and procedures for decision-making (VeneKlasen et al. 2002;Gaventa 2006).
Hidden forms of power are used by certain power players to maintain their power and privilege by creating obstructions to people's participation or by gatekeeping over the facts to be brought forward (VeneKlasen et al. 2002;Gaventa 2006). Invisible power goes one step further: it involves dominant groups continuing to hold power by keeping vulnerable groups unaware of their rights, for example by manipulating the narrative of how the powerless groups perceive themselves, so they feel unable to change and therefore accept the status quo (VeneKlasen et al. 2002;Gaventa 2006).
Thirdly, spaces to exercise power may be closed, invited or claimed. Closed spaces are where powerholders or authorities such as politicians, bureaucrats, professionals, executives make decisions with little broad consultation or involvement of the ordinary people (Gaventa 2006). Invited spaces are those where the people as users, beneficiaries or citizens are invited to participate by various kinds of power players; not necessarily government, but also international agencies or non-governmental organisations. Claimed spaces are carved out by less powerful actors from within the spaces available to the powerholders through social movements or community mobilisation for an issue (Cornwall 2002;Gaventa 2006).

Governance framework in Delhi
Delhi is a city-state in India, with a population of 16.8 million in 2011 and an area of 1,483 square kilometres (km 2 ) and nearly 1.   Figure 6 as conceptualised by Rao (2016). Even this structure, however, does not capture the full complexity of Delhi's governance arrangements, and Rao herself notes that it is "misleading because activities within functional heads always involve other agencies and also because some agencies wield more de facto power [than others]" (Rao 2016, p. 11). This complexity can also be seen in public health sector governance.

Patel & Pant
Delhi's Mohalla Clinics  Table 1. At the primary level healthcare is provided by GoNCT-D and the local bodies, as shown in Table 2 (see also Figure 7). Sources: GoNCT-D, NDMC, MCD; compiled by the authors

Mohalla Clinics
This study focuses mainly on Mohalla Clinics. The concept of the Mohalla Clinic is to provide quality primary healthcare services on the doorstep of local communities in Delhi. These clinics were commissioned as a direct result of AAP's promise to improve primary healthcare in its party manifesto.
There is widespread acknowledgement that Mohalla Clinics have improved access to health services by qualified providers, for the poorest of the poor, in Delhi. They have also brought health, education and social services back to the forefront of politics and governance (Lahariya 2017). These clinics featured prominently in the AAP's campaign for GoNCT-D elections in February 2020, along with the other works carried out by AAP while in power at NCT. In the event, the AAP won the elections again with a comfortable majority (Bhaduri 2020).
According to the GoNCT-D website (2017a), Mohalla Clinics are to provide the following services: 1. Primary medical care based on standard treatment protocols, including curative care for common illnesses (fever, diarrhoea, skin disease, respiratory trouble etc.), first aid and referral services.
2. A referral service for lab investigations, which are to be carried out by an accredited partner laboratory.
3. Medicines, free of cost to patients, based on the essential drug list published by the Delhi government.
4. Preventive services such as antenatal and postnatal care of pregnant women and new mothers, assessment of nutritional status and counselling; and the preventive and promotive components of ongoing national/state health programmes.

Objectives
The present research aims to understand the promises and practices of Mohalla Clinics, with particular emphasis on their governance and the politics surrounding it. The specific research objectives are: 1. Evaluation of community participation in Mohalla Clinics and their effects on the performance of the clinic.
2. Comparing the promises for participation in running Mohalla Clinics with actual practices.
3. Propose measures to improve efficiency and sustainability of Mohalla Clinics.
CJLG 2020 12 Through this research, we attempt to assess whether the Mohalla Clinics have delivered on the government's promise of making primary healthcare more accessible to the people and the role of community participation in the process.

Methods
The study is based on a case-oriented mixed-methods fieldwork in five different areas of Delhi (  Dev Nagar, with the highest Scheduled Caste (SC) 6 population (69%) in Delhi, is an example of a lowincome neighbourhood. Here, the SC population is taken as a proxy for income data. Similarly, Azadpur is also an example of a low-income neighbourhood, and the Mohalla Clinic is located in Azadpur fruit mandi (i.e. wholesale market) where casual workers make their living. Munirka Village and Paschim Vihar are examples of affluent neighbourhoods. New Kondli is just 1.5 km away from the NCT boundary. These areas were selected to understand how the Mohalla Clinics operate in neighbourhoods with different income groups, and to establish whether people from outside NCT also use the clinics.
In the areas mentioned above, primary surveys were conducted among 30 randomly selected users of the clinics to understand the accessibility, waiting time, cost incurred and to compare these parameters of Mohalla Clinics to the healthcare facilities they were using. Semi-structured interviews were also conducted in this area with both users and non-users randomly selected in an area within 500 metres 6 Scheduled Caste are a group of people recognised officially by the Indian constitution as socially and economically weak because they were historically discriminated against and require affirmative action. The study uses them as proxy for an economically weak section.

Limitations
The authors would like to highlight two principal limitations of the methodology described above. First, there may be recall bias in the study, as users were asked to compare their Mohalla Clinic's services to the healthcare facilities they were using prior to Mohalla Clinics. Second, seasonality of morbidity may also be a limitation in this comparison, as the survey was conducted in February and March 2019, while the recalled data may relate to other months.

Results
This section discusses the study's findings relating to each of the research objectives. These objectives are: 1. Evaluation of community participation in Mohalla Clinics and its effects on the performance of the clinic.
2. Comparing the promises for participation in running Mohalla Clinics with actual practices.
3. Proposing measures to improve efficiency and sustainability of Mohalla Clinics.

A curative initiative with people playing the role of beneficiaries?
The evaluation of community participation in Mohalla Clinics was carried out by applying the framework provided by Rifkin (1986) in Figure 1. The top-level findings were as follows: 1 4. Implementation: The implementation of the scheme, including site selection, is also top-down (see Figure 9). The first 100 clinics were set up in properties rented from private owners. PWD CJLG 2020 15 would set the rent for these, however, the property owners were not satisfied with these rates. 7 Opposition political parties have also alleged misconduct regarding the rents' disbursement (Bhatnagar 2016). There is also no standard size for the rental clinics. These reasons have resulted in GoNCT-D discontinuing the practice of opening Mohalla Clinics in privately rented premises. Since 2017, GoNCT-D mainly constructs the clinics in temporary structures (called Portacabins; short for 'portable cabins') on sites where it owns the land. 8

Performance of Mohalla Clinic as provider of primary urban health services
This section analyses the performance of Mohalla Clinics as a provider of health services based on the interviews and surveys conducted in the study areas. Residents living within 500 m of a clinic who did not use the clinic cited a range of reasons. These included having no need to go to the doctor for primary healthcare services; needing healthcare of secondary level and above; having an alternative such as However, for those residents who do use them, the clinics have provided several substantial benefits, discussed below.
CJLG 2020 20 Accessibility First, they are accessible to the people. The average distance travelled by the users is 1.3 km and about 51% of users live within 500 m of a clinic ( Figure 15). Average travel time is 18 minutes and the average travel cost negligible. This goes to show that Mohalla Clinics bridge some of the gaps in access to healthcare in the existing system.

Figure 15: Localities from which users come to Mohalla Clinics
Source: Compiled by authors from primary survey data Before the Mohalla Clinics, these patients either used alternative government or private healthcare facilities or self-diagnosed and -medicated (Table 3).

Reduced waiting times
There is a considerable pre-consultation delay at Mohalla Clinics: 30 minutes on average but ranging from five minutes to two hours. However, let us compare this with average waiting times in other places providing public health services in the city, such as outpatient departments of hospitals and dispensaries (which 56% of the respondents used before their local Mohalla Clinic opened); we find that people are saving two hours and 19 minutes on average. Nor did users complain about waiting time during the survey; most users responded that they were saving time. There is, however, an inverse correlation between waiting time and income, as shown in Figure 16. It could be that in the more impoverished neighbourhoods, people have specific health needs due to poverty or a lack of healthcare earlier in life.
It could also be because they have few options other than the Mohalla Clinic for primary healthcare, and the long waiting time indicates their dependency on the clinic. It may also be because people who can afford private healthcare or have other alternatives (such as a CGHS dispensary) do not use Mohalla Clinics (although the study did find a preference for GoNCT-D-run Mohalla Clinics and dispensaries over the PHCs run by the MCDs). 12 However, the hypothesis that alternative healthcare providers are preferred is supported by the views of residents of the more affluent neighbourhoods, who reported that they prefer their own (private sector) doctor who knows their medical history, whose availability in the evening suits them and with whom they can make an appointment.

Lower cost
There is no direct cost for accessing the health services of Mohalla Clinics. CJLG 2020 23 encouraged the 10% of respondents who previously practised self-medication to get proper healthcare in the Mohalla Clinics. There is, however, the opportunity cost of missing work due to the opening hours. In our survey, 57% of respondents said they had missed work due to attending the clinic, and only 18% said they had not. The remaining 25% did not work.
A number of respondents expressed dissatisfaction with regard to the clinics' opening hours, management of the diagnostic centre, and time and effort spent by the doctor on each patient. They reported that there is no way to report their dissatisfaction to GoNCT-D. Therefore, we can say that while Mohalla Clinics are improving the lives of the people, some form of community participation will improve its functioning. The next section discusses the promised participatory governance for Mohalla Clinics vis-à-vis the practice.

Tangled governance and a lack of accountability
In order to compare the promises and practices for participatory governance concerning Mohalla Clinics, it is useful to look at the history of participatory governance in Delhi. This is summarised in Figure 17. In Delhi, through the Municipal Corporation (Amendment) Act 1993 provision was made to establish ward committees at the zonal level. This was a direct consequence of the 74 th CAA. These committees consist of all the elected ward councillors within their respective zones, the nominated members of the council living in the zone, and the zonal municipal corporation officials. However, ward committees are not spaces for direct participation. Nevertheless, in India community participation in urban governance was directed in the 74 th CAA, which made it mandatory for ULBs with a population over 300,000 to form one or more tiers of governance, as a participatory space where elected councillors, administrative officers and citizen groups would work together (Patel et al. 2016b).
However, not all cities created the participatory spaces as mandated and, where they were created, there were vast differences across cities in terms of composition, proximity to citizens, functions, powers and resources (Patel et al. 2016b).

Congress), and resident ward committees (RWCs) launched in 2008 by the BJPwere attempts by
GoNCT-D and the MCD (before it was split into three) respectively at direct participation. These, however, resulted in elites capturing the invited space for participation, as only landowners were allowed to participate and, in the case of bhagidari, the local government was subverted 14 (Kundu 2011;Mohanty 2014;Rao 2016).
14 Bagidari had no constructional basis. The local government, i.e. the MCD comprising of elected ward councillors, were side-lined as Bagidari connected the GoNCT-D's agencies like PWD directly with the RWAs which comprised of rich landowning population, a space closed for the tenants. Also, projects carried under Bhagidar like solid waste segregation, were the responsibility of MCD.

Figure 17: Incremental but fragmented initiatives of participatory governance in Delhi
Source: Compiled by authors The Mohalla Sabha is the latest attempt of GoNCT at participatory governance. The AAP, the ruling party in GoNCT-D, promised in its manifesto (2015) that it would be a local, visible and invited space.
According to the manifesto, each ward in Delhi would be divided into 10-15 mohallas to be governed by an assembly of voters called the Mohalla Sabha. This sabha would be empowered to make decisions concerning the locality and would monitor the functioning of public institutions like the schools and PHCs, which include Mohalla Clinics (AAP 2015). However, in practice, the system of sabhas has been implemented under the GoNCT-D instead of by local bodies; the district magistrate is responsible for organising these meetings and for providing two mohalla coordinators to conduct them. Furthermore, each Mohalla Sabha's powers have been limited to urgent repairs in its neighbourhood within a budget of INR 1 Million (USD 13,595), 15 according to the Mohalla Sabha website (GoNCT-D n.d.). Delhi's 'tangled governance system' is to blame for its non-performance (Rao 2016, p. 11) and ultimately, the subversion of its local governments.
GoNCT-D is also looking at participatory governance for Mohalla Clinics through the system of Jan There are, however, some significant concerns with the JSS system. According to one respondent, JSS members are neither interested in the meetings nor accountable to anyoneindeed, they sometimes delegate their meetings to others. Also, because the process of setting up JSSs is in its initial stages, on the ground, neither Mohalla Clinic doctors nor the patients surveyed are aware of the JSSs. The JSS, therefore, is a local, closed and hidden space.
When respondents were asked their view of community participation in Mohalla Clinics, they gave divided responses. Some people thought the community should decide the operating hours, and oversee the management of diagnostic sample collection in the clinics; that is, they wanted the community to be involved in some activities and some aspects of implementation at their Mohalla Clinic. Others thought that the community should take part in the monitoring and evaluation of services in clinics. On the other hand, some respondents thought the community did not have the capacity for these tasks, which were, therefore best, left to professionals (i.e. doctors). As for respondents who were doctors and officials, they seemed reluctant to see community members in any position with decision-making power, claiming that this would lead to delays.

Mohalla Clinics during COVID-19
On a positive note, Mohalla Clinics have improved access to healthcare for ordinary people during the COVID-19 pandemic of 2020, as the major hospitals of the city have been treating patients with COVID-19 and the private clinics have been shut (Dutt and Chitlangia 2020). During the lockdown, 16 however, reports of some Mohalla Clinics being shut, despite the instructions of the chief minister, came to light. It turned out this was due to the inability of those staff who were contract workers to commute to their place of work, as they had no identity proof or travel facility (Babu 2020 (Bedi and Sirur 2020), and GoNCT-D also plans to deliver vaccines through Mohalla Clinics (Press Trust of India 2020).
The following section discusses recommendations to improve the efficiency and sustainability of Mohalla Clinics.

Policy recommendations
GoNCT-D currently shares responsibility for healthcare with local governments. It is the local bodies who are responsible for the preventive aspects of healthcare such as antenatal and postnatal care, immunisation drives, health education, solid waste management etc. Mohalla Clinics however are run by the GoNCT-Dnot local bodies. We have recommended that because GoNCT-D has better resources the power and responsibility for healthcare should be with them and sanitation can continue to be with the local governments. MCD dispensaries and PHCs can also be converted to Mohalla Clinics and provide a better quality of service. This will also avoid redundancies in the health infrastructure of the city and the two governments will not be competing with each other while providing a public service.
There is no community participation in planning the healthcare programmes, or in their implementation: these are top-down activities in Mohalla Clinics. The government did initially attempt to get the community involved in monitoring clinics and participating in their activities through setting up a management committee, which was planned to have four to five volunteers (GoNCT-D 2017b)but it is not functional on the ground. Instead, these functions also are top-down. In their current form, therefore, Mohalla Clinics are limited to providing primary curative healthcare, and in this form, they have achieved some success.
The top-down process for planning and management, however, is not proving successful, and the authors believe this will adversely impact the Mohalla Clinics in the long term. It is very tempting at this point to claim that community participation can solve all these problems; however, studies have shown that this line of thought by planners in the past has led to the failure of participation programmes (Rifkin 1996). Therefore, it is essential to consider which processes are best managed through participation and which through top-down decision-making.
Mohalla Clinics have shifted the approach to primary healthcare away from vertical health programmes (such as the nationwide Pulse Polio vaccination programme) and towards horizontal interventions, where a single clinic is looking at all diseases, communicable or non-communicable and even other According to Verbeek (2014, p. 14), many aspects of primary health relating to mental well-being, injury due to violence, or chronic diseases due to environmental pollution can be controlled through integration of public health and urban planning. However, he cautions that both disciplines are "institutionally scattered across a tangle of administrations and policy levels, and the major connection between them is made through a generic regulatory framework with fixed environmental norms and a strictly organized environmental impact assessment"and that this remains the case today even though urban planning started over 150 years ago in response to 19th-century unhealthy and overcrowded industrial cities. The statement certainly holds for India, and especially for Delhi where the DDA (one of whose roles is as the urban planning body) sets unrealistic standards for construction byelaws and makes impractical plans which it cannot achieve. For example, MPD 2021 suggests having one dispensary per 10,000 population, with a floor area of 800-1,000 m 2 , and two polyclinics per 100,000 population, with a floor area of 1,000-2,000 m 2 . This would mean, for a projected population of 23 million in 2021, about 460 polyclinics and 2,300 dispensariesbut these are not being implemented by the DDA. This brings us back to the importance of participatory governance.
To understand the process of participatory governance better, the authors also reviewed another model  According to Olum (2014, p. 37) "the successful implementation [of decentralisation] has to take into account six preconditions, namely: the establishment of institutional mechanisms, the creation of spaces for citizens' participation, political will and civil will, capacity development at the local level, careful implementation, and democratic governance." Delhi fulfils the first, second and the last conditions easily, but there is ambiguity in both the political will and civil will. For example, the GoI's reluctance to devolve power over land and local governance to GoNCT-D shows a clear lack of political will to decentralise. Similarly, GoNCT-D tries to create spaces for citizens' participation whenever there are opposition parties in power at a particular governance level: e.g. bhagidari were created by the INC to subvert the MCD (under the BJP), and Mohalla Sabhas are being used as a similar instrument by the AAP. People's mixed responses on participation also show ambiguity in the civil will. The people's preference for Mohalla Clinics run by GoNCT-D, ahead of PHCs run by the MCD, further indicates a lack of capacity for participation at the local level.
Given the current state of affairs and based on the findings of this study, the authors recommend that a useful first step towards meaningful community participation would be decentralisation of the process of selecting the location for clinics and including the three MDCs in it. Community voices should ideally also be heard in deciding where to site the clinics, but this may lead to delays as with the absence of a proper Mohalla Sabha, there is no proper community participation framework in place. The community can be involved in monitoring and evaluating the services of the clinic. They can also help decide opening hours for the clinics, which could end up being different in different neighbourhoods.

Measures to improve efficiency and sustainability of Mohalla Clinics
Based on above observation and findings, we therefore suggest the following measures to improve the efficiency of Mohalla Clinics and make this a sustainable programme in the long run.
1. People in their interviews and during the survey indicated a preference for Mohalla Clinics over clinics run by other agencies like MCDs and a lot of them have stopped using the services of MCD dispensaries. This vote by feet indicates public confidence in Mohalla Clinics and GoNCT-D. Therefore, we recommend that responsibilities related to health be shifted away from MCD to GoNCT-D, a higher level of government with more money and resources to spend on health.
2. The monitoring and evaluation of the services should be community driven. It would be more economical than CCTV cameras or other IT-based surveillance and would also develop, in the community, a sense of ownership and responsibility for the Mohalla Clinic.
Community voices should also be heard in deciding the operational hours of the clinic, which could end up being different in different neighbourhoods.
3. The process of site selection can be decentralised to involve the three MCDs as they are a principal landowning and land regulating agency in Delhi. This would ensure that the sites selected for Mohalla Clinics are located optimally, and it may also speed up the site selection process.

Conclusion
Mohalla Clinics, in their current form, are limited to providing primary curative healthcare. They have shown some success in this form. For example, users of Mohalla Clinics who previously used government facilities, save on an average two hours and 19 minutes of their time in using primary health care services. About 10% of the users were citizens who admitted to practising self-medication previously but are now getting proper healthcare at the Mohalla Clinics. Thus the Mohalla Clinic is a positive step towards universal primary health coverage and for many ordinary citizens, it has successfully bridged the gap in accessing primary healthcare in the current system. However, there is no community participation, whether in the planning of healthcare programmes, the monitoring and evaluation of clinics, the implementation of programmes or the execution of any activities. These are all organised top-down. Top-down interventions have failed to be effective in finding sites or locations for the clinics and organising community-friendly opening hours. It is also regrettable that Mohalla Clinics are not managed by Mohalla Sabhas, but by the JSS, and that power struggles between political parties are played out within local government bodies, subverting their important local governance roles.
This paper therefore proposes greater community participation in the evaluation of Mohalla Clinics, the shifting of resources away from MCD run (non-Mohalla) clinics to GoNCT-D administered Mohalla