Equity workshop: Understanding links between ecosystem services/governance and human well-being

Equity workshop: Understanding links between ecosystem services/governance and human well-being

• 1.
  Understanding links between ecosystemservices/ governance and human well-being: reflections on conceptualization and operationalisation Frank Vollmer School of GeoSciences, University of Edinburgh [email protected]
• 2.
  Fisher et al(2013): Conceptual frameworks for ecosystem services and poverty alleviation research reviewed: o Environmental Entitlements Framework (Leach et al., 1999) o Framework for Ecosystem Services Provision (Rounsevell et al., 2010). o Millennium Ecosystem Assessment (MEA, 2005b). o Political Ecology (Blaikie and Brookfield, 1987). o Resilience (Folke, 2006 and Holling, 1973). o Sustainable Livelihoods (Chambers and Conway, 1992 and Scoones, 1998). o The Social Assessment of Protected Areas (linked to Sustainable Livelihoods) (Schreckenberg et al., 2010). o The Economics of Ecosystems and Biodiversity (TEEB, 2010a). o Vulnerability (Adger, 2006 and Fussel, 2007). Conceptual Frameworks Source: Fisher, J. A., G. Patenaude, P. Meir, A. J. Nightingale, M. D. A. Rounsevell, M. Williams and I. H. Woodhouse (2013) 'Strengthening conceptual foundations: Analysing frameworks for ecosystem services and poverty alleviation research.' Global Environmental Change, 23(5), 1098-1111 2
• 3.
  Conceptual Frameworks Source: Fisher,J. et al (2013: 1108-09) 3
• 4.
  Source: Fisher, J.A., Patenaude, G., Giri, K., Lewis, K., Meir, P., Pinho, P., Rounsevell, M., Williams, M. (2014) Understanding the relationships between ecosystem services and poverty alleviation: a conceptual framework. Ecosystem Services, 7: 34-45 4
• 5.
  • Sense ofcomplexity • Comprehensive frameworks such as [Fisher et al (2014)] make things harder to overlook, [and] they dictate what is on the agenda. This leads to a central limitation: if frameworks are used mechanistically or uncritically, they can hinder a deeper, questioning analysis, that remains open, for instance, to factors that do not feature in the framework [“other means than ES”] (Fisher et al., 2014: 35). Conceptual Frameworks: Pros and Cons of Operationalization Source: Fisher, J. A., G. Patenaude, P. Meir, A. J. Nightingale, M. D. A. Rounsevell, M. Williams and I. H. Woodhouse (2013) 'Strengthening conceptual foundations: Analysing frameworks for ecosystem services and poverty alleviation research.' Global Environmental Change, 23(5), 1098-1111 5
• 6.
  “Forest resources [andtheir derived ES] may contribute to local livelihoods through: (1) a needs-driven forest reliance, whereby local poor people depend on low-value forest resources to some extent for their livelihoods, perhaps in response to shocks (“safety nets”), or (2) because they are unable to make the transition out of this resource dependent mode (“poverty traps”); and (3) an opportunity driven forest reliance, whereby local people use higher-value forest resources as a source of cash products in order to get richer (“pathways out of poverty)”(Clements et al (2014: 125-126) Links between ES and livelihood: what do we know? Source: Clements et al (2014), “Impacts of Protected Areas on Local Livelihoods in Cambodia”, World Development Vol. 64, pp. S125– S134, 2014 6
• 7.
  Impact evaluation methodsutilised to investigate the effect of protected areas (PAs) on poverty and livelihoods in Cambodia (comparing households inside PAs with bordering villages and controls) found that:  There was no evidence that PAs exacerbated local poverty or reduce agricultural harvests in comparison with controls (Households bordering the PAs were significantly better off, not because of the PA but due to greater access to markets and services).  Non-timber forest product collectors inside PAs were significantly better off than controls and had greater rice harvests, because they had more secure access to land and forest resources.  The PAs in Cambodia therefore have some positive impacts on households that use forest and land resources for their livelihoods (Clements et al (2014: 125) Links between EG and livelihood: what do we know? Source: Clements et al (2014), “Impacts of Protected Areas on Local Livelihoods in Cambodia”, World Development Vol. 64, pp. S125– S134, 2014 7
• 8.
  Development as 1. welfare, 2.utility or 3. freedom maximisation (and equalisation) All concepts still in use, dependent on • Academic discipline, • Underlying assumptions, e.g. regarding the welfare models (growth-mediated vs. support-led strategies), • Practical concerns (used in structural equation models, linear correlation analysis, qualitative evaluations, impact assessments, etc). Dependent variable “Well-being”: Equality of what (Sen, 1980)? 8 Source: Sen, A. (1980), “Equality of What?”, in McMurrin, S. (ed.): Tanner Lectures on Human Values, Cambridge, Cambridge University Press
• 9.
  Money-metric poverty assessments: •Identification a) per adult equivalent consumption as the welfare metric, • Identification b): an absolute poverty line, usually based on the Cost of Basic Needs Method • Aggregation: Use of FGT method (1984) - FGT0 (poverty headcount (incidence of poverty), FGT1 (poverty gap – incidence, intensity and depth of poverty) or FGT2 (squared poverty gap - incidence, intensity, depth of poverty and inequality among the poor). Measuring poverty 9
• 10.
  Measuring poverty 10 Gaddis, I.and Klasen, S. (2012), Mapping MPI and Monetary Poverty: The Case of Uganda, at Dynamic Comparison between the Multidimensional Poverty Index (MPI) and Monetary Poverty Workshop, November 21-22, 2012, Oxford University. Available at: http://www.ophi.org.uk/wp- content/uploads/Stephan-Klasen-Mapping-MPI-and-Monetary-Poverty-The-Case-of-Uganda.pdf?0a8fd7 (23/03/205) Poverty line region Line or Rate Household or People Mozambique Poverty line (MZN) US$1.25/day US$2.5/day MPI All Mozambique Line Rate Rate Households People 18.41 (US$0.53) 47.3 54.7 20.05 MZN 53.2 60.6 40.10 MZN 85.6 90.1 69.6% Gaza and Inhambane, rural Line Rate Rate Households People 18.37 55.2 65.2 20.02 60.3 69.9 40.04 89.4 93.9 60.1% Each computation of poverty is imperfect and can be critiqued from different angles. Substance: • Caloric intake not suitable to assess nutritional quality of diet • Each person converts means (income) differently to ends (human development, e.g. a healthy diet) • Non-deprived in income does not mean access to health care is ensured Utility: • “Empirically, MPI poverty much less varied spatially than income poverty” (Gaddis and Klasen, 2012) – do you want to see spatial differences?! Do you want to see abrupt/non-linear changes?
• 11.
  Identification 1. The Unitof Analysis 2. Dimensions of poverty 3. Variables/Indicator(s) for dimensions 4. Poverty Cutoffs for each indicator/cross-dimensional 5. Weights within and across dimensions Aggregation 1. Dashboard approaches 2. Axiomatic measures (Counting approach (e.g. Alkire- Foster method)) 3. Fuzzy set 4. Statistical approaches (e.g. Multivariate Analysis) Poverty Index 11
• 12.
  Identification 1. basic materialneeds for a good life, 2. health, 3. good social relations, 4. security and 5. freedom of choice and action (MEA, 2005) Poverty Index 12 Well-being components No of BBNs in which they were included Total Villages WSs National WS Provincial WS Food Security 9 3 3 3 Good quality farm 6 3 2 1 Cattle 3 3 0 0 Access to drinking water 6 2 2 2 Good quality housing 3 2 1 0 Health care 2 2 0 0 Purchase capacity 3 1 2 0 Education 2 1 0 1 Achieve your dreams 1 1 0 0 Freedom 1 1 0 0 Peace 1 1 0 0 Energy availability 3 0 0 3 Protection against extreme weather events 2 0 1 1 Wild food 1 0 1 0
• 13.
  13 Poverty Index Domain DimensionDeprived if… Human capital 1. Sanitation 2. Water 3. Health (under-five mortality, access to health care) 4. Formal Education (illiteracy, highest qualification achieved) • The household´s sanitation facility is not improved (according to the MDG guidelines), or it is improved but shared with other households • The household does not have all-year long access to clean drinking water (according to the MDG guidelines) or clean water is more than 30 minutes walking from home • Any child has died in the family; illnesses remain undiagnosed by professional health specialists • No household member is able to read and write; no household member achieved EP1 or attended the Portuguese colonial school system. Social capital 1. Food security 2. Access to services, associations and credit • Household did experience a food shortage in the past • The household did not receive advice from an extension agent during the last 12 months, and did not receive a credit in the last 12 months, and is currently not a member in either an agricultural or forestry association. Economic well- being 1. Income (cash + subsistence) 2. Assets owned 3. Housing (floor, roof, walls) • Quintiles • If do not own more than one of: radio, TV, telephone, bike, bed, motorbike or refrigerator and do not own a car or truck • The household has sand or smoothed mud floor; the household has grass or poles roof; the household has sand, mud, grass or poles walls
• 14.
  14 Poverty Index Challenges: 1. Explicitvalue judgments: as cardinal data is mixed with ordinal and categorical data, value judgments to set poverty lines are required 2. What constitutes “adequate housing”, “access to health care”, “food security” is often multidimensional itself and thus hard to capture by a single indicator or a proxy 3. Ideally, variables do not correlate much – challenges to link ES or EG to well-being (e.g. sanitation, clean water access) 4. “Change” analysis: Panel data often not available, necessitates alternatives (space-for-time substitution). Practical challenges occur - controlling for similar soils and woodland vegetation and a similar provision of public services within study sites is, in reality, a much harder task than on paper
• 15.
  15 Use of multipledependent variables in regression Studies increasingly use multiple dependent variables in regression analysis • Hossain et al. (2015) used linear regression, among other statistical techniques, to analyse how ecosystem services are coupled to economic growth and well-being in the in the Bangladesh coastal zone (Well-being defined as poverty (% of population below poverty line), Per capita income, Gross domestic product)). • Santos et al (2013), in “Ecosystem and Human Wellbeing in Spain”, used structural equation modelling to explore "the relationships between biodiversity loss, ecosystem services, human wellbeing, drivers of change (both direct and indirect) and policy responses” (10 well-being indicators) Sources: Hossain, M.S., Dearing, J.A., Rahman, M.M., and Salehin, M. 2015. Recent changes in ecosystem services and human wellbeing in the Bangladesh coastal zone. Regional Environmental Change (Published onlune 21 January 2015) Santos-Martin, F et al (2013), Unraveling the Relationships between Ecosystems and Human Wellbeing in Spain, PLoS ONE 8(9)
• 16.
  16 Enhancements of ourunderstanding, some observations… • Preference for cardinal indicators (less room for different interpretation of results/easier to show trends (linearity)/ use of quintiles rather than a poverty line – while it adds knowledge to the picture, it does not capture the entire picture)  Access to services or markets = distance (physical accessibility to services), but says little about their financial affordability, social acceptability, quality of services • Covariates: “Poor matching designs might identify an effect when in fact none exists or mask effects […]. A simple comparison of households inside the PAs with bordering villages would come to the conclusion that PAs exacerbate local poverty. The results of the impact evaluation show that this would be a misleading comparison, because border villages were closer to market centers, other services, and main roads, all of which had positive impacts on local poverty status” (Clements et al (2014), S129 – S130)  Finding the right control variables might be challenging if the dependent variable is a composite index with various types of variables, links to different dimensions of well-being, and variables that link either to public or private goods (Keyword: Endogeneity)  “Soft variables" (social dynamics, exclusion, etc) are harder to use Use of multiple dependent variables in regression
• 17.
  17 The way forward:Use of Poverty Index in regression 1. Micro regression (determinants of poverty of a person or household) 2. Macro regression (determinants of poverty at the district, state, province or country level, ethnic group, gradient level)  Endogeneity is a great challenge with multidimensional poverty/well- being: high correlation between a variable constituting the dependent variable with an independent variable (the same forces that influence the input also influence the output – ownership of goods (motorcycle) to explore forest resources). Alternatives: • Instrumental variable (exogenous variable thought to have no direct association with the outcome (harder to find with multidimensional poverty composed of indicators that are not highly correlated) • Nonindicator measurement variables, e.g. certain demographic characteristics or additional socioeconomic characteristics of the household (ethnicity, hh size, etc.) (possibly not very satisfying)  Well-being determinants might change across spatial differences Source: Alkire et al (2015), “Multidimensional Poverty Measurement and Analysis: Chapter 10 – Some Regression Models for AF Measures”, in Alkire, S. et al. (eds), Multidimensional Poverty Measurement and Analysis, Oxford University Press (forthcoming)
• 18.
  Understanding links between ecosystemservices/ governance and human well-being: reflections on conceptualization and operationalisation Frank Vollmer School of GeoSciences, University of Edinburgh [email protected]