In 2016 we prepared aCommon Country Analysis (CCA) for Palestine. A CCA is UN speak for a detailed analysis of a country in preparation for a multi-year action plan of the UN. It identifies key development challenges and where the UN needs to focus its development investments.
For our analysis this time, we decided to look at people. In hindsight it appears to be the obvious thing to do, but we were not the first to think of this. The Nepal UN Country Teamdid it before us.
For our CCA we asked ourselves two questions:
Who are the most vulnerable groups in Palestine?
What are the structural drivers of their vulnerability?
We thought if we could identify the most vulnerable groups and analyze the structural drivers of their chronic vulnerability, we will have a good sense of what it will take to ensure that our sustainable development investments leave no one behind.
The first call for ideas brought out 61 proposed groups, each backed by passionate arguments as to why they are the most vulnerable.
We merged some groups, reduced duplications, clarified categories, tinkered with definitions, and after extensive discussions, honed our focus to 20 vulnerable groups. This gave us a window to the factors that keep some groups in Palestine systematically at a disadvantage.
Next, we did a deep-dive to understand why development was leaving some groups behind. For some groups, including out-of-school children and children in the labour market, the lack of adequate data makes it difficult for government to formulate specific policies and programmes for these groups.
Alternative data collection methods for groups that are small compared to the population
After a comprehensive exercise to account for the data, especially looking at Sustainable Development Goals indicators, we noted that relevant data on smaller groups couldn’t be collected only through existing surveys.
The Palestinian Central Bureau of Statistics (PCBS) uses representative samples for each geographical area of the occupied Palestinian territory (oPt), and even though it produces high quality data consistent with international standards, there is a lack of up-to-date and periodic disaggregated data on several smaller groups.
Take for example, the fishermen of Gaza.
There are some 4,000 registered fishermen in Gaza, accounting for 0.2 percent of Gaza’s population of two million.
If PCBS samples 1,000 people from Gaza for one of its quarterly labour force surveys, it will have at most two fishermen in its sample. We cannot draw any reliable conclusions about the socio-economic conditions of fishermen in Gaza from a sample of two people. And if PCBS included more fishermen in their sample, the percentage of fishermen in the sample will be larger than the percentage of fishermen in Gaza’s population.
To create a large enough sub-sample for fisherfolk, PCBS will need to do a new level of sub-sampling by profession or sector on top of the two layers it is already subsampling. This would significantly increase its cost of surveys. Are you still tracking with us? Keep reading.
Flash surveys to the rescue
So, for the smaller groups, we at the UN looked for an approach to gather data that would not cost too much, would not create too much additional work and most importantly, that is able to produce good quality data.
The first thing we tried is a series of flash surveys – with small samples, and short questionnaires.
These flash surveys had several benefits over the more traditional surveys with bigger samples and longer questionnaires:
They allowed us to test our systems for collecting primary data and iterate quickly and cheaply if necessary to work out the flaws in the system.
They enabled our enumerators to get hands-on training at a relatively low cost to us.
They are also particularly suitable for understanding the smaller groups that don’t get adequately represented in the bigger surveys.
We chose four vulnerable groups: adolescent girls, children in labour, the elderly and persons with disabilities as pilot cases.
UNFPA took the lead in this. They engaged the Sharek Youth Forum, a non-profit, and one of UNFPA’s implementing partners to conduct the surveys.
OHCHR, FAO, UNRWA, helped with the quality control.
37 university students (28 from the West Bank and 9 from Gaza) were recruited from Sharek’s network and trained as enumerators by an expert.
The survey questionnaires in Arabic were uploaded on KoBoToolbox, a free and open source suite of tools for collecting data. Many of the young enumerators owned smartphones so they downloaded the app on their phones and entered the data for each person they surveyed into their smartphones. Sharek provided the others with tablets.
A village, a town and a refugee camp were selected in each governorate. Sharek’s enumerators visited schools to survey adolescent girls, reached out to the elderly in their local communities, and found persons with disabilities through support groups.
ILO provided information on the areas with high concentration of child labour. The enumerators collected the data over a period of two weeks, and, in some cases, they used paper forms to collect the data and documented problems as they arose.
The enumerators collected data on a small number of key demographic variables for each group.
Before even looking at the data, we noted a few things.
First, we now have 37 trained enumerators who can be deployed again at short notice to conduct other flash surveys. The investment in training and the hands-on experience they got has started the process of creating systems to collect data on vulnerable groups.
Second, we need to finesse our sample selection if we want to use the surveys to provide baseline indicators and monitor progress.
Third, we need to think through how to combine the data from smartphones and paper surveys.
Fourth, we need to figure out how to identify our target groups based on more rigorous definitions. For instance, not all work done by children should be classified as child labour. According to ILO, child labour refers to work that “deprives children of their childhood, their potential and their dignity, and that is harmful to physical and mental development”.
Fifth, flash surveys need more quality control if they are to serve the same purpose as traditional surveys. This is because with smaller samples of flash surveys, the choice of location will need extra attention to ensure that the sample is indeed representative. This year, we will work through these wrinkles.
Engaging people in their own data analysis
In data circles, we often hear the idea of engaging communities to collect and use their own data. But the instances of it being done in a meaningful, low cost, sustainable way to generate usable data are few and far between. Could we pull it off?
We decided to experiment with combining data collection and empowerment for one of the most vulnerable groups in the oPt, namely, Area C communities. Area C accounts for 60 percent of the West Bank. It has some of the most fertile agricultural land and almost the entirety of Palestine’s natural resources. An estimated 300,000 Palestinians live in Area C and a greater number depend on its resources for their livelihoods. Area C is controlled by the Israeli military, which has exclusive control over land, planning and construction.
Significant portions of Area C land are allocated for Israeli settlements and declared as Israeli state land.
Only about 30 percent of Area C is available for Palestinian construction, but so far Palestinians have been issued permits to build on less than one percent of the land.
Since construction permits in Area C are closely tied to Israeli spatial plans, spatial plans driven by Palestinian communities have been used in recent times to empower communities, and to rally the Israeli Civil Administration to issue permits to Palestinians for construction.
In addition to Israeli military orders, land ownership in Area C is governed by a complex legal framework resulting in insecurity of land tenure and confusion about ownership and user rights of private land. Consequently, land registration has been a long-time priority of local and international development actors in the oPt.
As the next activity of our project, we integrated a community-driven process to map land ownership and user rights.
UN-Habitat took the lead in developing a system called the Social Tenure Domain Model. This participatory tool is a pro-poor, gender responsive system based on free and open source software, which means that all the data collected and stored is available to the communities and owned by the users.
The system is based on information and evidence shared by local communities making them a part of the decision-making process. The system records and analyzes the social tenure relationship of people and land, and the social services/amenities that available to the inhabitants of a location. It fits the oPt’s highly complex tenure system, because it supports a continuum of land rights ranging from formal to informal.
An Arabic interface was created for the system so it can easily be deployed in other Arabic-speaking countries. UN-Habitat also provided training for the Palestinian Land and Water Settlement Commission staff.
This system for community mapping of land rights with a special focus on women and youth will help us empower the community, build social cohesion, and generate data on land rights.
The resulting database will serve as a shadow land register, support land valuation, raise awareness about land governance in Area C, and inform advocacy efforts to defend land rights of Palestinian communities. These efforts are supported by the ‘Road Map for Reforming Palestinian Land Sector’ of 2017. Right now, the background work is still ongoing. The model will be piloted in 2019.
Will this actually work?
We don’t know. For now, we know that we now have the systems in place to replicate or update the data collection of smaller groups through flash surveys, we can engage communities participate in collecting and analysing their own data and integrate a community-driven process to identify land ownership and user rights, at a lower cost than in the first run. And we will use whatever we learn from these initiatives to finesse our methods in our next set of data collection initiatives in 2018.