Women-headed households: The need for good data

Tuesday, 19 January 2016 00:44 -     - {{hitsCtrl.values.hits}}

By Yajna Sanguhan of the Centre for Poverty Analysis (CEPA) 

With the Sri Lankan Government’s new task force on Women-Headed Households (WHHs) and the persistence of development actors to aim their programmes at WHHs, it is vital that we have an improved and more nuanced interpretation of who exactly these households are. 

There is no official one-size-fits-all definition and arguably there shouldn’t be, but many women and their families commonly identified within this category do face certain vulnerabilities and do deserve certain entitlements. Yet the targeting of WHHs has been taken up by various stakeholders without a full awareness of who is included in this group and debate on the definition of a WHH persists. 

As previously debated in the series, the concept of the “head of the household” needs to be problematised for numerous reasons (such as identification of a “head” which inevitably reinforces a hierarchy). However, while it continues to be universally used as the unit of analysis, this article makes a case for good quality data that can inform policy and tackle the issues relating to categorising WHHs.

13While we continue to use the concept of “head of the household”, it is imperative that as a first step to deconstructing WHHs and informing policy makers, new methods are employed to collect data and a rigorous, holistic approach is taken to capture the diverse and nuanced experiences amongst Women-Headed Households in Sri Lanka


Current statistics on WHHs and  their limitations

As discussed in the last article in this series, one of the largest and most comprehensive databases that include data on the sex of heads of household is the Household Income and Expenditure Survey (HIES), which defines the household head as “a person who usually resides in the household and is acknowledged by the other members of the household as the head of the household”.

With 1.2 million households (23.5% of all households) reported as WHHs island-wide according to the latest report based on data collected from 2012 to 2013, it is not just a phenomenon in the war-affected regions: the percentage of WHHs is over 20% in every province. Though the high numbers in the north and east are likely a result of the war, the numbers in the other provinces demonstrate that this has been a steady, gradual process. 


That the HIES is one of the best examples of data available for WHHs is troubling for many reasons. Firstly, the latest round of the HIES is the first time that data has been collected for all 25 districts due to previous restrictions on data collection in the Northern and Eastern Provinces (other surveys such as the Census of Population and Housing and the Labour Force Survey also had similar restrictions until recently). Thus, this limited data cannot reliably inform us of any changes or trends in WHHs (particularly those in war-affected areas) over time, despite claims to do so.

Secondly, the HIES only collates data on WHHs based on its narrow definition and its responses then depend on the balance of power within a household. The control over assets, spending and food is often unequal within a household so the responses to the survey questions can vary depending on who answers the survey. 


The definition also assumes that WHHs are a homogeneous group, which means that it cannot capture the differences in income and poverty within WHHs. Without understanding intra-household dynamics and allocation of resources, such as income, we cannot fully understand the true heterogeneity. However, this could be at least partially overcome if we extract the right data in order to identify the diversity within the group. 

As experiences between households in an urban environment and a rural environment are likely to differ, a household will inevitably encounter different issues if the head is a widow relative to a household where the head is a woman whose husband has migrated for work. 

The HIES definition of a WHH fails to capture the reasons for becoming a WHH and as such, the related differences in income and other resources. Therefore, understanding the gendered dimensions of poverty not only requires superior data but a better use of the currently available data.


The need for better data

A key setback in the collection of quality data is a result of the inadequate methods currently employed by most surveys. To take the HIES as an example again, the survey design relies on the enumerator identifying the head of the household according to the HIES definition mentioned earlier. This method results in an underestimation of the number of WHHs, as the presence of an adult male in the house would often prevent a household from identifying a woman as the head and would also produce bias on the part of the enumerator – a common flaw across many surveys. 

The problem with underestimation is that it can mask the scale of any potential issue, thus leading to resources, policies and incentives that are insufficient to manage the needs of WHHs. New methods to collect data are therefore vital to analyse WHHs and deconstruct our notion of who that should be. 

A study conducted in Panama allowed for the use of multiple definitions of the head of the household, which provided multiple poverty rates. Another option would be to conduct a survey without directly asking about the household head and instead obtaining data on dimensions such as decision making, asset ownership and income in order to collate information about potential headship characteristics. 

These types of alternative methodologies grant more flexibility both in terms of definitions for policy makers and in developing a deeper, more realistic depiction of the variety within WHHs as a group. While we continue to use the concept of “head of the household”, it is imperative that as a first step to deconstructing WHHs and informing policy makers, new methods are employed to collect data and a rigorous, holistic approach is taken to capture the diverse and nuanced experiences amongst WHHs in Sri Lanka.