Poverty Traps and Resource Dynamics In Smallholder Agrarian

Poverty Traps and Resource Dynamics In Smallholder Agrarian Systems Chris Barrett Cornell University Guest Lecture January 27, 2009

Introduction Oikos (“estate” or “household”) – common etymological root to ecology and economics suggests deep connections. Yet strong latent connections commonly overlooked. Example: burgeoning literatures on thresholds and multiple dynamic equilibria: resilience, catastrophic collapse, poverty traps, etc. Need to integrate better. Conservationists need to consider predictable consequences of human agency Development scholars need to respect the feedback between human and natural processes.

Some African Examples A. The East African ASAL Recent evidence suggests multiple herd size equilibria: - low level associated with sedentarization and localized range degradation … a resource degradation poverty trap - high level associated with mobile pastoralism and resilient range ecology - multiple equilibria arise due to uninsured climate risk, and are faced by those with moderate-high herding ability (low ability herders face unique, low-level equilibrium) Effects are compounded by contested, unclear property rights in land (and insecure rights in animals) which leads to conflict and coordination problems … but different from the classic Hardin tragedy of commons concern.

Those who maintain a herd remain mobile on a resilient landscape, while those who lose their herd collapse into destitution on a degrading local landscape.

Some African Examples B. Western Kenyan Maize Systems - - - Shepherd and Soule (1998), Barrett et al. (2006) find homeostatic systems alongside poverty and severe soil degradation. Lumpy and delayed payoff investments (tea, dairy) and INRM form a reinforcing feedback loop. Collapses into persistent poverty trace back to (health) shocks. Result is apparent multiple equilibria in both assets/incomes and in soil conditions Market failures problems compounded by (i) informational lags and biases in soil perceptions, and (ii) serious coordination problems in striga mgmt

Some (but unfortunately not all) children in this system face a lush future.

Some African Examples C. Rice Systems in Madagascar A national scale resource degradation poverty trap? Heavy dependence on rice, but low uptake of improved inputs or production methods – due to many market imperfections (insurance, credit, land) – leads to low yields, soil nutrient mining, erosion, and slash-and-burn extensification. Imperfect learning about new rice technologies and bounded rationality associated with social customs (famadihana and ritual cattle sacrifice) Coordination problems in forest access and water use to facilitate SRI uptake and reduce deforestation and resulting siltation of irrigation canals.

Some African Examples The result is pockets of productive, seemingly sustainable agro-ecosystems amid broad-scale economic and ecological problems

The Economics of Poverty Traps Poverty trap = “any self-reinforcing mechanism which causes poverty to persist” (Azariadis & Stachurski). This can include: (i) unique dynamic equilibrium systems (convergence on misery) that are empirically uninteresting (ii) conditional convergence systems (unique equilibria for distinct groups, only some below a poverty line) (iii) multiple equilibrium systems (initial condition guides resulting path dynamics)

The Economics of Poverty Traps Welfare. Dynamics. With. Unconditional Convergence Welfare Dynamics With Multiple Dynamic Equilibria Well-beingt+1 Chronic poverty region High group Transitory poverty region W 2 Pov. line Low group ` W 2 Well-beingt Key: unique, common path dynamics with a single stable dynamic equilibrium Key: unique nonlinear path dynamics with awith single multiple stable dynamic equilibrium equilibria that differs and at among least distinct one unstable groupsdynamic or individuals equilibrium (threshold effect)

The Economics of Poverty Traps How might multiple equilibria emerge? Three broad classes of explanations, each with quite different policy implications: A) Market Imperfections B) Imperfect learning and bounded rationality C) Spillovers, coordination failures and economically dysfunctional institutions

The Economics of Poverty Traps: Market Imperfections Multiple variants of market imperfections story: Nonlinear pricing Internal economies of scale (incl. sunk costs) Credit (financial liquidity) constraint Uninsured risk Unobservable labor effort and resulting moral hazard and efficiency wages Implication: poverty traps can be overcome with adequate resources to overcome the market imperfections that presently obstruct capital accumulation and technology adoption in the poorest areas of the tropics.

The Economics of Poverty Traps: Imperfect Learning/Bounded Rationality Imperfect learning/bounded rationality alternative: Agents may have a difficult time observing changes in the environment around them, especially changes occurring at some distance from their current position. Differences in beliefs or subjective expectations can generate “inertial self-reinforcement” (Mookherjee and Ray 2000) 3 variants of problem: informational lags (e. g. , mosquito control, technology treadmill), differentiated social networks, norms/conventions under bounded rationality. Implication: Additional resources need not generate the most productivity-enhancing investments. Rather, the highest return interventions would provide more timely, accurate and universally available information so as to surmount barriers to learning and innovation.

The Economics of Poverty Traps: Spillovers, coordination failures and economically dysfunctional institutions Several variants of this explanation: Technological and pecuniary externalities Coordination failures more generally (Formal and informal) rules of behavior that breed economically dysfunctional institutions: corruption, weak property rights, failure to contribute to public goods and services, etc. Implication: Need to craft rules of interaction – and rules for transitioning to new rules – to facilitate coordination, create focal points at Pareto dominant equilibria, and discourage venal behaviors.

Poverty Trap – Resource Dynamics Link A. The Poor’s Assets Poor smallholders’ heavy dependence on natural capital creates intrinsic linkages. When the key state variables of two systems are shared in common, strong interdependence follows automatically. Resource dependence need not lead to a poverty trap; indeed, resource exploitation has often been the pathway out of poverty. Key biophysical assets: human health is primary for the poor; complementary inputs from nature (forests, soils, water, wildlife), especially land (>70% of natural capital in low-income countries). Most biological assets follow highly nonlinear dynamics: generates coupled collapse or abundance in human wellbeing and biological resources … no automatic vicious cycle, rather multiple equilibria.

Poverty Trap – Resource Dynamics Link B. Poverty Trap Mechanisms Apply to NRM, too Market imperfections – e. g. , credit constraints, uninsured risk, unobservable labor effort – cause underinvestment in natural resources conservation by smallholders. Information lags and flow barriers, and norms/conventions associated with bounded rationality all inhibit adaptation and lead to inertial self-reinforcement. Externalities, coordination failures and weak institutions pervasive and long recognized as central to NRM problems in the rural tropics: insecure property rights, lack of rules (or enforcement), corruption and powerlessness/voicelessness.

Policy Implications “Divergence, big time”: most low-income countries face declining per capita wealth while most high-income countries enjoy increasing per capita wealth. Poverty traps imply a clear compulsion to intervene … only reinforced by close coupling to environmental state But how to intervene is much less clear because alternative mechanisms imply different responses: De. Soto vs. Sachs. Most likely, face “fractal poverty traps” (Barrett and Swallow WD 2006) – interlinked processes across different scales, with micro-level market imperfections reinforcing (and reinforced by) meso-level information problems and macro-level institutional failures. Implication: high returns to detailed empirical study to identify proximate causes in a given setting.

Thank you for your time and interest!