Pakistan’s unsafe water

By Hina Shaikh and Ijaz Nabi

Safe drinking water is a central plank of a country’s health strategy as it affects nutrition intake and therefore infant mortality, child growth, and the ability of adults to be productive. Exposure to unsafe water also leads to skin-related disease. For these reasons, access to safe drinking water is a right enshrined in the constitution and is a critical Sustainable Development Goal.

The Consortium for Development Policy Research (CDPR) recently brought together a panel of experts to discuss the current status of drinking water in Pakistan and what is being done to ensure that citizens enjoy this constitutional right.

Pakistan’s “water stress”

The panel distinguished between “water scarcity” and “water stress”. With the world’s fourth highest rate of water use, Pakistan’s economy is one of the most water-intensive in terms of cubic meters consumed per unit of GDP. Subsequently, water availability per capita has shrunk to under one thousand cubic meters by 2017 from over five thousand in 1951. Pakistan crossed the “water scarcity line’” in 2005, indicating a shortage of overall supply. With higher than expected population growth, this is likely to get worse.

The focus of the discussion was primarily on “water stress”, or the prevalence of polluted water that is unsafe to drink. Panelist Syed Hasan, Assistant Professor of Economics at the Lahore University of Management Sciences, noted that Pakistan became a water-stressed country in 1990 and is expected to be among the most water-stressed countries in the world by 2040.

Panelist Hammad Khan, Director General of World Wild Fund for Nature (WWF) Pakistan, pointed out that the level of arsenic in the water supply far exceeds the government’s own thresholds for contamination, which are in fact less conservative the World Health Organization’s (WHO) standards. A 2015-16 nation-wide survey by the Pakistan Council of Research in Water Resources (PCRWR) found that only a third of the 369 samples tested for water quality were safe for consumption. A separate PCRWR study conducted in 2011 found 100 percent of water samples in Lahore were polluted with arsenic. A study led by Joel Podgorski, a scientist at the Swiss Federal Institute of Aquatic Science and Technology, found that water in two-thirds of the 1200 wells sampled across Pakistan exceeded the WHO-recommended threshold of arsenic. Based on this data, nearly 60 million citizens are estimated to be consuming toxic ground water.

Microbial pollution is also common. In cities, water becomes contaminated due to improper disposal of solid waste and continued usage of outdated water and sewage networks. Chemical pollutants from industrial waste also infect water. In rural areas, open defecation and animal waste are the leading sources of contamination.

Poor quality of water burdens the health sector, increases missed days of work, and reduces labor productivity. High levels of arsenic in water contributes to underweight birth, skin defects and miscarriages. Syed Hasan mentioned the burden of poor health outcomes in the form of waterborne diseases costs Pakistan 1.6 million disability-adjusted life years – and almost four percent of GDP.

Poor governance

Pakistan has been unable to incentivize conservation and efficient usage of water. Syed Hasan commented on the pricing mechanism and its failure to reflect the true market value of this critical resource. The tariff for water used for household consumption in urban Pakistan was last revised in 2004. Operation and maintenance costs incurred by water authorities continue to exceed the revenue they collect, while water metering covers only eight percent of the households.

Hammad Khan explained that if sources of water remain unprotected, the availability of drinking water will keep dwindling. Ground water once contaminated cannot be treated. The installation of filtration plants by Punjab’s Saaf Pani Company (see below), meant to cover all union councils in Punjab, are remedial measures – not sustainable solutions. The unrelenting adulteration of water sources despite decades of several dedicated water authorities in operation reflects a serious governance failure.

The government response in Punjab

Following devolution, water became a purely provincial subject. Punjab set up the Saaf Pani Company three years ago to ensure provision of drinking water in rural Punjab. The company’s progress is personally overseen by the chief minister.

Panelist Tahir Majid, Chief Technical Officer, Punjab Saaf Pani Company, agreed that despite being the Punjab government’s flagship initiative and substantial public expenditure, progress has been slow. A third of the water schemes in the province remain non-functional while 79 percent provide water that is unsafe for consumption.

The problem with Saaf Pani Company reflects a deeper problem of governance pervasive across several other sectors. When parallel governance structures are set up in the presence of existing departments, such as the Punjab Health and Engineering Department, inefficiencies slide in. The company saw several quick changes in senior management (some resulting in criminal inquires) and frequent changes in the operational design. This has resulted in delays and has not encouraged strong private sector engagement in the delivery of safe drinking water to the citizens.

After being heavily scrutinized for its performance, the company is now restructuring itself to improve delivery and remains committed to providing clean drinking water to Punjab’s entire unserved population of 60 million by 2025.

What can be done?

It is encouraging that “water-stress” can be overcome. Singapore’s example, cited by Syed Hasan, shows how the risk of extreme water stress can be countered by efficient regulation and management. Inaction, however, may result in a crisis similar to the one in Cape Town, a city that is now on the verge of rationing clean drinking water.

The panel suggested immediate steps Pakistan can take to tackle the water crisis and avoid the Cape Town outcome:

Set the right priorities: An over-arching water policy framework is critical. The National Water Policy, in circulation since 2004, should be updated in light of changes and approved.

Set water classification standards: Every country (including India, Bangladesh, and China) has water classification standards where all the water bodies are categorized according to their usage. This should also be done in Pakistan. WWF has offered to use remote sensing and GIS mapping to help government conduct this exercise. Authorities will then be able to ensure more effectively that water bodies classified for drinking purposes are kept clean and used for that purpose alone. Water classification will also help avoid disputes between different stakeholders (agriculture vs. industrial vs. household consumption).

Let prices work: Even though water is a basic right, it is a limited resource and hence water pricing is important. The fact that people pay for bottled water indicates that there is willingness to pay.

Mobilize community ownership: Community ownership is key to ensuring that water schemes remain functional and well-maintained. The WWF for example signs a legal contract with the community for joint ownership of the filtration plants it has provided.

Hina Shaikh is a Pakistan country economist at the International Growth Centre.

Understanding Punjab’s rural non-farm economy


Usman Naeem

Where is Pakistan’s growing labor force going to find jobs? Many argue that, given the limits of arable land, agriculture cannot create much new employment, pushing job-seekers to cities. However, just because future job growth may not be agricultural, that doesn’t mean it won’t be rural. This is because of the rural non-farm (RNF) economy, which has the potential to absorb a large part of the labor force, slow migration to already congested cities, add to GDP, and alleviate poverty and inequality.

The RNF economy has been neglected by policymakers because few understand the sector and its role in development. That is not surprising given the dearth of research on RNF activity in Pakistan.

To fill this knowledge gap, the most rigorous evidence on Punjab’s RNF economy has been digitized by the International Growth Centre (IGC).[1] The IGC project is based on a census of small and cottage industry in Punjab that was conducted by the Punjab Small Industries Corporation[2] (PSIC) from 2011 to 2013. Following strict protocols[3], data for 352 unique activities covering 24,210 rural clusters (villages)[4] in 36 districts of Punjab was recorded.

The PSIC survey reveals important data to understand RNF activity in Punjab, where 63 percent of the population now live in rural areas. Following is a brief overview of the most important data.

Dividing Punjab into North, Center, South, and West regions based on the classification adopted by Cheema, Khalid and Patnam (2008), the project revealed significant regional differences in the kinds of economic activity found in villages. These differences can be explained by historical variation in the economy, agrarian structure, road and irrigation infrastructure, human development, and other factors that distinguish these regions.

Table 1 reports the average per cluster of different types of RNF activity at the level of Punjab and regions.

Table 1: Snapshot of the non-farm economic activities at Punjab and regional levels

Total clusters Average per cluster
NDUs NDUs*: Non-farm economic activity NDUs
Trade Personal services Others Rest**
Punjab 24049 430.84 426.43 61.89 3.83 19.95 7.54 4.26 26.31
North 2309 394.36 377.34 35.96 2.28 15.06 5.9 3.54 9.18
Center 11961 435.75 428.13 70.4 3.99 19.64 6.98 4.43 35.35
West 4094 402.57 461.58 53.03 5.03 20.62 7.55 4.26 15.56
South 5685 455.69 417.46 60.92 3.26 22.11 9.39 4.19 21.97

* Non-dwelling units (NDU’s) were units that were being used for non-residential purposes, i.e., for economic and non-economic activities.
** This includes farm and non-economic activities and those that were unidentified because of incomplete or illegible information.

The PSIC data not only shows the kind of RNF activity, but also the intensity of it. The intensity of economic activity (IEA) in villages, defined as the number of non-farm economic activities in a village per 100 households, is mapped at the district level in Figure 1:

Figure 1: Average intensity of non-farm economic activity in a village by district


IEA is low in the North region, high in four out of the seven districts in the West region, high in all but one district in the South region (with the exception of D. G. Khan), and it varies from high to medium in the Centre region.

We then looked at how the seven most important categories of RNF activity were distributed regionally, as shown in Table 2:

Table 2: Main non-farm economic activities in percentage by region

Non-farm economic activities’ categories North Centre West South
Retail Food/ Beverages 35.93 41.66 38.6 39.89
Other Retail 15.05 11.97 7.95 7.6
Household Goods and Home Appliances 6.65 7.27 7.83 8.29
Miscellaneous Services 13.27 10.15 10.89 12.58
Social Services 9.59 11.21 9.44 8.76
Wholesale 4.57 4.23 7.49 5.15
Motor Vehicles Related Activities 5.87 6.37 8.82 8.54
Activities not included above 9.07 7.13 8.97 9.2

Going forward, researchers should use the data set to answer a broader set of questions for this sector. Policy research stemming from this will act as a tool for policymakers to understand and subsequently tap into the true potential of Punjab’s RNF economy.

The data set is available upon request at Anyone wanting access to the data set should send in a request with his or her name, title, organization and a brief write-up on the proposed research, to the above address.

Usman Naeem is a Pakistan Country Economist at the International Growth Centre.

[1] This post is based on the IGC-funded project, ‘Development of an electronic database of Industrial and Commercial Activity and a Spatial Analysis of Small and Cottage Industries in Punjab, Pakistan’, the research team on which included Syed M. Hasan, Attique ur Rehman, Ijaz Nabi, Naved Hamid, and Usman Naeem.

[2] PSIC is an affiliated organization of the Industries, Commerce and Investment Department (IC&ID) with a mandate to “promote sustained industrial development through provision of market driven credit, infrastructure and technological support to contribute towards poverty alleviation through job creation and socioeconomic uplift of the province”.

[3] Data was only entered for rural clusters in Punjab, and in order to ensure quality, the data was entered in-house and under the direct supervision of the research team. In order to minimize subjectivity, a macro-enabled excel-based data entry system was designed that automated much of the data entry operations. In addition, three Stata programs were also written for improved data entry management and quality assurance.

[4] This consists of Mouzas (villages), which are the smallest revenue units recognized by the unique Hadbast number, within a Tehsil (an administrative sub-division of a district). There are a total of 25,914 Mouzas or rural clusters in Punjab.

Why a grant to help people migrate during seasonal hardship could be a game changer


By Agha Ali Akram

Estimated conservatively, half of the world’s 600 million food insecure people who live in rural areas suffer from seasonal hunger during the “lean season”, the period of the agricultural cycle between planting and harvest when there is less work. As much as 50 percent of Pakistan’s population is food insecure, and while there are no specific estimates, it is likely that Pakistan’s large agriculture sector – which employs 45 percent of its workforce – is vulnerable to seasonal deprivation.

Could offering residents of rural areas a small travel grant to temporarily migrate alleviate their suffering? In a recent study¹ Mushfiq Mobarak (Yale University), Shyamal Chowdhury (University of Sydney), and I conducted in Bangladesh, we show that travel grants targeting the landless improve their own welfare and benefit rural village economies in the midst of seasonal deprivation.

Seasonal food insecurity in Rangpur and a potential solution

Every year, September ushers in a three-month lean season for landless farmers living in Rangpur, a region in the northwest of Bangladesh. Agricultural work dries up and the landless lose their source of income. And since they do not own property, they are unable to grow enough of their own food. This sets off a period of considerable hardship.

Small travel grants to temporarily leave the region and find work in nearby urban areas that are not affected by the lean season show promise as a solution. Gharad Bryan and my co-authors previously tested this idea and found that grantees increased both consumption and caloric intake. They also migrated during the subsequent lean season without being offered a travel grant. The small travel grant reduced the risk households associated with temporary migration and opened a highly profitable opportunity for them.

Studying spillovers: What we did differently

Given these results, we wondered if the consequences of temporary migration were “spilling over” to villages of origin. For instance, migration might create a relative scarcity of labor for village employers, driving up local wages and benefiting landless workers who stay back. Conversely, if non-migrants and migrants have a complimentary relationship – that is, non-migrants depend on migrants to find work – then migration would hurt the rural economy as non-migrant productivity would drop.

To measure these potential spillovers, we built on Bryan et. al in our own study by varying the amount of travel grants offered to landless workers. In 47 “high-intensity” villages, we offered 70 percent of landless residents a grant, while in 48 “low-intensity” villages we offered 14 percent of landless residents a grant (for a total of 5,792 offers made across the two treatment groups). 38 other villages were assigned to a control group, in which no one was offered a grant.

This variation in the intensity of travel grant offers allowed us to understand the spillovers of migration. The higher the migration rate, the greater the potential impacts on the local village economy.

The travel grant scheme was coupled with extensive surveys. We surveyed both households that were offered the grant and those that were not offered. Surveying households that were not offered allowed us to detect whether a migration opportunity they could not access benefitted them nonetheless. We also surveyed employers in all our villages to understand wage and other demand-side impacts. Finally, we repeated the household and employer surveys in the subsequent lean season. Our surveys sampled 3,600 households and 1,099 employers across control and treatment villages.

The results

First, we find that grantees were much more likely to migrate and find work in nearby cities. Due to this increase, migrants in the high-intensity group increased their total income by 20 percent from working in destination cities.

Next, more strikingly, we find that those offered the travel grant in high-intensity villages were twice as likely to migrate than those in low-intensity villages, suggesting that people are more likely to migrate if more people around them are also likely to migrate. In fact, we also find that landless workers not offered the grant were more likely to migrate in high-intensity villages, which suggests that just knowing people who are likely to migrate can increase one’s own likelihood of migrating without needing any subsidization. This is important. A program built around a travel grant like ours could save costs because not everyone needs to be subsidized.

Moreover, under the assumption that rural labor markets are not well integrated (i.e. the village is the relevant labor market), we also find that the local agricultural wage rate increases. In our high-intensity villages, more than half of the local landless population migrate away because of the travel grant we offer, “emptying out” the local labor market. This bids the origin village wage rate up, as reported by employers in our survey work, since there is relative labor scarcity.

Migrant workers take advantage of this increased wage rate by supplying more labor in the origin village labor market too. They do this by migrating back and forth between their origin village and destination over the course of the lean season. Because so many people do this, there is slack in the village labor market that landless workers can take advantage of. Where previously village labor markets had few opportunities for paid work, there were now relatively more opportunities available. This is good news for landless workers as they can take advantage of opportunities both in the city to which they temporarily migrate and their own village.

Policy implications and the bigger picture

Our results point to a new and cost-effective policy option to improve rural welfare and address the terrible suffering caused by seasonal deprivation. With a small travel grant, many landless rural workers can temporarily migrate and benefit from work opportunities in cities. A key insight from our study is that a seasonal migration grant policy at scale would potentially benefit the rural economy since it bids up the local agricultural wage and allows previously unemployed workers to find opportunities. And any program that uses travel grants would not need to be offered to all landless workers because those not offered a grant also migrate, suggesting that there is an optimal level of coverage below 100 percent.

Additionally, our findings point the way to a policy that can complement existing rural development programs. Typically, rural development policy efforts focus on developing human and physical capital by providing very direct financial and material support to the rural sector through rural support programs (Pakistan has at least 11 major rural support programs). The seasonal migration travel grant provides an innovative and scalable solution to add to the existing set of rural support programs. In fact, rural support programs provide a natural vehicle to scale the travel grant, adding it to their programming.

Finally, the travel grant has potential for impact if applied to other kinds deprivation. For example, the deaths, water shortages, and loss of livelihood brought on by seasonal drought in Pakistan’s Thar region could be ameliorated if residents had access to a travel grant. Given the inherent unpredictability of disasters, mobilizing resources could be challenging but is well within the realm of possibility.

Agha Ali Akram is an environmental economist who recently completed a fellowship at Yale University’s School of Forestry and Environmental Studies. He holds a PhD in Environmental Economics from Yale University.

[1] This links to a summary of the paper, which is being readied for publication at the time of writing.

Six insights for engaging community activists in rural development


Hamna Ahmed, Asha Gul, Saheem Khizar, Simon Quinn, and Kate Vyborny

You can’t overstate the importance of local organizations in providing services for their communities.

Pakistan’s flash flood of 2012 washed away the water supply scheme and 1.5 kilometers of a road in the remote village of Kiyar in Khyber Pakhtunkhwa’s Karimabad valley. After a week, the local community restored these services itself.

In the hilly areas of northern Punjab, open defecation polluted the soil and water supply, making water-borne diseases widespread in the region. Activists in Bhattian union council (UC) in Kotli Sattian tehsil convinced over a thousand households to construct latrines, eventually leading the UC to be declared open defecation-free.

The local community in Rostam UC in Sindh – with support from a local NGO – linked women artisans to the Sukkur market, increasing their profit threefold.

These are just a few examples of the vast diversity of activities local communities in rural Pakistan are engaged with. Rural communities have been equipped to take responsibility over local problems through decades of local and international investment in social mobilization.  Community organizations – also known as Third Tier Organizations (TTOs) – have been set up at the neighborhood, village, and UC levels. They are managed by local volunteers, who form elected governance bodies within them.

These organizations are integral to local development and ought to be included in community-level development initiatives. However, UC organizations vary widely in their activities and performance, signaling the need to better tailor social mobilization efforts and policies according to the relevant regional characteristics to maximize success in rural development outcomes.

A 2014 survey[1] of 850 UC organizations supported by Pakistan Poverty Alleviation Fund (PPAF) and its partner organizations captures this variety. Here are its six most important insights:

1. Rural organizations focus on a variety of issues

As Figure 1 shows, UC organizations work in a wide range of sectors. More than 50 percent of UC organizations are active in health, education, human rights, and assistance with legal documents and procedures.[2] Some frequently reported activities include holding free medical camps, providing first aid training, raising awareness on various health issues (such as polio, dengue, tuberculosis, cleanliness, and hygiene), providing financial and non-financial resources such as school uniforms, books, and stationery to encourage school enrollment, organizing education awareness campaigns, and helping community members acquire legal documents such as national identity cards, marriage certificates, and birth and death certificates.

Figure 1


2. There is significant regional variation in the level of organization activity

The amount of services offered by UC organizations differs by area. Organizations in KP and Gilgit are most active, working on an average of six activities from August 2013 to August 2014. On the other hand, their counterparts in Balochistan worked on an average of two activities over the same period. These discrepancies may be explained by differences in the timing of social mobilization efforts across different regions, geographic factors (difficult terrain, sparse population, and limited infrastructure), security issues, and limited financial resources which make it harder for local organizations to operate.

3. The level of donor involvement also varies widely

Partner organizations, which consisted mostly of the Rural Support Programmes[3] for this survey, engage with UC organizations in different ways. Figure 2 shows that around 35 percent of UC organizations report active involvement of their respective partners in more than 75 percent of their activities while almost 40 percent of them report no involvement of their partners. Some UC organizations may need more assistance – for example, in the early stages of their development or when they are scaling up a new activity – while others may be more independent. The variation may also be due to partners having more funding to support field staff time and travel in some areas than others, as well as greater distances between organizations in sparsely populated areas like Balochistan.

Figure 2


4. Partners are most involved with the most active organizations

The busier the UC organization, the more likely a partner will be involved with it. Partners were most engaged with organizations that had more than ten activities in a year. 80 percent of these highly active organizations reported that a partner was involved in at least one of its activities, and 40 percent reported that the partner was involved in all of them. In contrast, about half of the least active organizations (those that had three or fewer activities in a year) reported no partner involvement in any activity. This suggests that partner support helps increase activity, or that organizations with the most initiative are rewarded with more time and assistance from partners (or both). This also suggests potential for more support from partners to less active UC organizations.

5. Funding is the most common form of partner involvement

Almost 50 percent of surveyed UC organizations report that they receive some funding from their partners. However, funding is not the only form of partner engagement. Partners also assist these organizations in developing connections, getting permissions from relevant government departments, providing technical advice, and helping staff carry out activities (Figure 3). This kind of support is likely to help organizations work more independently in the future. Partner engagement also comes in the form of training on issues such as financial management and development of UC/village plans. Around 90 percent of the organizations received some form of training from partners in the last three years.

Figure 3


6. Lack of resources does not necessarily mean low activity

Although UC organizations with the higher funding carry out more activities and serve more people, the differences in activity between them and organizations with little funding are not so stark. The lowest-funded organizations worked on an average of three activities while those with the most funding worked on an average of seven activities between August 2013 and August 2014. The average well-funded organization is involved in the construction of infrastructure projects regularly used by 4,500 people. These organizations also served about 15,000 people over the year by providing services, information, or training. But even the organizations with the lowest funding delivered services to an average of 5,000 people.

Further research

Although these numbers do not account for differences in the quality and extent of services delivered by UC organizations, they demonstrate that positive outcomes are possible even with minimal financial resources.

We are running an experiment[4] in collaboration with PPAF to test the impact of both self-reporting and non-financial incentives on the governance and service delivery performance of UC organizations across the country. The organizations are being asked to regularly report their own progress and achievements, rather than having achievements reported primarily by staff of the partner organizations after field visits. Some organizations are also being offered non-financial incentives in the form of a recognition package, as these are volunteer organizations that may be motivated by social praise especially within their own communities.

The findings from this project will help donors, the government, and other development sector organizations devise and revise their engagement strategies with community organizations to increase their effectiveness and sustainability.

Hamna Ahmed is a a faculty member at the Lahore School of Economics and PhD candidate in Economics at the University of Kent.

Asha Gul is a PhD candidate in Economics at the University of New South Wales.

Saheem Khizar is a field experiment coordinator at the Pakistan Poverty Alleviation Fund.

Simon Quinn is an associate professor of Economics at Oxford University.

Kate Vyobrny is a post-doctoral fellow at Duke University.

[1] This survey was conducted in the fall of 2014 by a joint research team from the Lahore School of Economics, Oxford University, and Duke University in collaboration with PPAF which is the apex institution managing this social mobilization process with support from local partner organizations. The survey gathered data through meetings with each UC organization’s executive body (EB) on their governance, activities, future plans and characteristics of the EB members. The survey also gathered data from 109 field offices of the 19 partners that are directly involved with these local community UC organizations.

[2] An organization is classified as active in a sector if it worked on at least one project/activity within that sector between August 2013 and August 2014.

[3] The Rural Support Programmes Network (RSPN) is the largest development network of Pakistan, with an outreach to over 43 million rural Pakistanis. It consists of 11-member Rural Support Programmes (RSPs), which have been operating since 1982. The survey was conducted with both RSP and non-RSP organizations. The largest partner organization for the survey is the National Rural Support Programme (NRSP) that is engaged with 60% of the TTOs, followed by Sindh Rural Support Organization (SRSO) (11%), Sarhad Rural Support Programme (SRSP) (8%) and Agha Khan Rural Support Programme (AKRSP) (7%).

[4] This research is being supported by funding from PPAF, International Growth Center (IGC) and the National Science Foundation (NSF).