Developing a livelihood-sensitivity matrix

Submitted by Sukaina Bharwani | published 25th Mar 2011 | last updated 30th Mar 2011

This page shows how to develop a livelihood sensitivity matrix that can be used in several ways:

- To synthesize existing knowledge on climate vulnerability in a fairly rapid participatory exercise with stakeholders
- To provide a first-order vulnerability assessment based on expert judgment
- To integrate results from a variety of quantitative and qualitative methods

The matrix also illustrates some of the technical issues in the use of indicators of vulnerability to a range of climatic risks.The analysis works best if focused on a particular region, ecosystem or resource. For instance, it might look at highland land use vulnerable to drought and floods or coastal zones susceptible to sea level rise and cyclones.

Assessing livelihoods

The first step is to list the livelihoods in the case study region.

Then work backward to list the productive activities of these livelihoods and the ecosystem services that support those elements. Thus, the rows of the table are organised according in a hierarchy of the ecosystem services that are essential in productive activities, which are elements of common livelihoods. For example, a general relationship between climate and the soil water balance will affect a variety of crop and livestock production activities, which are the major components of some livelihoods.

More generally, the rows are the 'units of exposure' - those elements in ecosystems, populations and economies that are subject to climatic hazards and trends. Analysts may wish to organise the rows differently - for instance concern for cross-cutting sectors such as infrastructure (roads, electricity, ports, and market facilities) may warrant adding a block on economic services to correspond to ecosystem services. The nature of the exposure elements should correspond to the broad framing of the vulnerable conditions.

Assessing Climate Hazards

The next step is to list the present climatic threats (or opportunities) and trends that are significant for the list of livelihoods (or exposure units). These climatic risks are the columns of the matrix. Some judgment is required to separate the continuum of weather and climate into distinct threats. For instance, drought is almost always a threat in some form for rural livelihoods. In the example below, episodes of drought over a year or more are separated from shorter dry spells during the year.

It is likely that some iteration and refinement will be warranted in rows and columns of the matrix. There are no hard and fast rules for separating ecosystems into services, people into livelihoods, or weather into climatic risks. Indeed, one of the purposes of the matrix is to show how thresholds of vulnerabilities differ between exposure units (and over time). The definition of drought risk is quite different for subsistence farmers and pastoralists.

How sensitive is each element of exposure to each climatic risk? Fill in the matrix by ranking each cell. A rapid, scoping exercise might use high, medium or low score. A five-point scale is probably sufficient for most analyses.

This is a first-cut of a rapid vulnerability assessment. What does it reveal about who is vulnerable? What are the gaps in knowledge? What indicators of vulnerability or adaptive capacity are generic to the matrix or specific for livelihoods and threats? For instance, crop-drought indicators (such as yield) are of different importance for semi-arid subsistence agriculture than for highland commercial farms. What is the range of adaptation options? Are these specific to livelihoods and threats or more generic? What institutions are relevant for implementing adaptation options for each livelihood?

Using a livelihood-sensitivity matrix

Further use of the matrix might involve exploring ratings according to outcomes, or comparing different scenarios of future vulnerability and aggregating the matrix ratings into overall scores.

The rating of sensitivities depends on the outcome of exposure and hazard. For instance, sensitivity to mortality usually has a different rating for each hazard than exposure to loss of livelihood (e.g., economic impacts), or well-being (as a broader category including social and psychological stresses). In most cases, the initial ratings are related to a broad interpretation of economic assets. However, if the matrix is to be used analytically, it is necessary specify what the consequences or outcomes of the identified vulnerabilities. Most commonly, these include loss of life and loss of property (assets), but some stakeholders may be concerned with the full range of livelihood capitals - including social networks and psychological stress.

The matrix is relatively easy to fill in for present conditions. In order to compile a matrix for future vulnerabilities, the analysts need to describe storylines that indicate how livelihoods might change (e.g., their reliance on different ecosystem services and activities, as well as their prevalence), how climate might change (there might be new hazards or trends become significant in the future) and how the sensitivities might change (for example with new technology). These are the typical concerns of building scenarios. The matrix provides an easy means to compare the results.

With some caution, it may be interesting to sum the rows, columns and overall matrix into aggregated indicators. A simple sum (divided by the maximum possible score) of the rows and columns yields relative scores for exposure and impacts. An aggregate score for the total exposure and impacts might be weighted by the probability of the different hazards occurring (for the exposure scores) or the prevalence of the livelihoods (for impacts scores). Analysts may wish to look at different scoring methods - for instance counting the number of high scores (for example those with a 4 or 5).

Note that aggregating matrices for different outcomes is not recommended - the justification for adding exposure to mortality to exposure to property loss raises concerns for equity that are beyond this simple technique (see guidance material on multi-criteria and benefit cost methods).

Onto:

Further resources