1 / 30

AFEPA Master thesis Presentation on

Determinants of the Food Consumption Vulnerability of the Extreme Poor -Empirical Evidence from Southern Bangladesh . AFEPA Master thesis Presentation on . By Mohammad Monirul Hasan Thesis supervisor: Prof . Dr. Imre Fertő CUB, Budapest August 6, 2013. Background of Study:.

Download Presentation

AFEPA Master thesis Presentation on

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.


Presentation Transcript

  1. Determinants of the Food Consumption Vulnerability of the Extreme Poor-Empirical Evidence from Southern Bangladesh AFEPA Master thesis Presentation on By Mohammad MonirulHasan Thesis supervisor: Prof.Dr.ImreFertő CUB, Budapest August 6, 2013

  2. Background of Study: • The southern region of Bangladesh suffers from natural disaster and salinity in the cropping land in every year. • Seasonal unemployment – lean period for labor demand. • Seasonal food deprivation => vulnerability. • Sen(1981) identified lean period as a period when the ability of a large segment of the population is limited in acquiring food, employment and other basic necessities. • Moreover, big cyclones like Sidr (2007) and Aila (2009) caused huge death toll and loss of permanent assets. • Inadequate assistance from govt. and NGOs made them to migrate, cope or simply starve for extended period.

  3. Background of Study:Figure: Tracks of cyclone over last 50 year. Source : CEGIS, DHAKA The household level analysis covers the sample households of following 3 districts of Southern Bangladesh. -- Khulna -- Patuakhali-- Satkhira

  4. Source: Maps are generated by CEGIS. Maps are assembled by Author. Map of Bangladesh (upper captioned) is from PKSF.

  5. Research Questions: • 1. What are the determinants of the food consumption vulnerability of these extreme poor households in the southern Bangladesh? • 2. Does microfinance play any role to reduce vulnerability of the poor households? • 3. What is the impact of two major natural disasters (Sidr and Aila) on food consumption vulnerability?

  6. Research hypothesis: • Household head characteristics (Female, Age, Years of schooling, marriage) • Occupation (Wage worker, Self-employed in agri, Self-employed in non-agri, migration) • Household characteristics (Max Years of schooling, household size, Sex Ratio, migration of any member) • Infrastructure facilities (Access to electricity, tube-well or tap water, sanitary latrine) • Households' distance Information (Distance from main road, small market, big market, nearest microfinance branch) • Village characteristics (Household in char areas) • Physical and financial asset (owned land, agri land, using land, non-using land, free land, cows, goats, poultry, total assets, savings etc.) • Household income and expenditure (total income, food expenditure, non-food expenditure) • Households' loan information (formal loan, informal loan, deferred payment) • Households' coping in disaster (loss and unmet loss in Sidr, Aila and last year crisis, social safety net)

  7. Food consumption vulnerability:

  8. Methods and data • InM-PKSF data. PKSF did census in 2010 of 60,053 HHs. • The criteria for selecting ultra-poor households - (1) Monthly income less than or equal to 3,000 Taka (equivalent to EUR 30) per household during lean period; or (2) Primary occupation of the household head is daily wage earning (in farming, fishing, logging, honey collection or other activities); or (3) Having less than or equal to 50 decimal cultivable land.

  9. Baseline survey 2011 • InM conducted baseline survey in 2011. • 4000 HH were targeted and 3977 HH remains for working dataset.

  10. Theory and Methodology • Theory supported by consumption smoothening- • Permanent Income Hypothesis (PIH) • Credit Constraint Hypothesis • PIH by Milton Friedman in 1957 – Any change in consumption caused by shocks to income (transitory income) could be smoothed sufficiently by perfect capital market borrowing as the household would try to maximize utility. Household will borrow from market when it has transitory low income and by saving when having transitory high income. So consumption patterns of households are largely determined by the change in permanent income.

  11. PIH • households consumption is not completely smoothened if imperfect financial market prevails (Derconet al.2000; Duflo & Udry, 2004; Goldstein, 2004). • According to Morduch, (1995) households are credit constrained if they are not able to fill the income gap by borrowing sufficiently while experiencing income shock. • Not only the credit constraint but also the household precautionary behaviour could result violation of permanent income hypothesis (Deaton, 1991; Morduch, 1990; Paxson, 1992). • According to Deaton (1991) and Kurosaki (2006), savings, other accumulated assets, external assistance and remittances or cash transfer could play the role of absorbers of shocks.

  12. Risk Rationing Theory • According to Boucher & Carter (2001), risk rationing shifts the significant contractual risk to the borrowers when lenders are constrained byasymmetric information resulting voluntary withdrawal of the borrowers from the credit market even though they have the necessary collateral to qualify the loan contract. • Risk is from the both sides- • Supply side: Asymmetric information, distance, group formation modalities. • Demand side: quantity rationing, transaction cost rationing and risk rationing(timely repayment the instalment, peer pressure from the group members, losing the entrance fee, losing option for future credit from the same institution, losing the deposited savings and any collateral they had in the MFIs)

  13. Method selection • Ordered Probit model => as -2 to +2 • Heckman Probit model for selection bias => conditional of MFI. Missing value problem. Endogeneity problem. • Propensity Score Matching => matching on observable characteristics between Treatment and control. (Khandker, Koolwal and Samad, 2010), Godtland et al (2004), Khandker et al (2010), Heckman, Ichimura, and Todd (1998) and Caliendo and Kopeinig (2008). • Mean difference => t - test for mean.

  14. Literature review on similar studies • Chambers (1989) describes vulnerability as “defencelessness, insecurity and exposure to risk, shocks and stress”. • Chaudhuriet al. (2002) argues that “the main difference between poverty and vulnerability is risk”. • Pitt and Khandker (2002) showed that micro-credit helps to smooth consumption by effectively diversifying agricultural income and employment. • Chaudhuri and Christiaensen (2002) showed that the characteristics of the poor were remarkably consistent with the characteristics of the vulnerable: large family sizes, high dependency ratios, illiteracy, location in counties with low provision of public services and poorer regions of the country. • Günther and Harttgen (2005) found that rural vulnerability were caused by both covariate and idiosyncratic shocks.

  15. Descriptive analysis

  16. Seasonality and income

  17. Normal and lean period

  18. Consumption ordering

  19. Degree of vulnerability

  20. Regression Model • => ordered probit model. Where, • = Food consumption vulnerability for individual i (ordered from -2 to +2) • Xi = vector of observed continuous variables like household and family characteristics, infrastructure facilities, occupation, physical and financial assets, household income, expenditure and consumption of individual i. • Di = Dummy of some idiosyncratic and covariate shocks, effect of cyclones, occupations, infrastructure facilities, member of microfinance etc.

  21. Heckman selection bias test ρ is non-significant, Mills lambda is not significant. There is no selection biasness for microfinance participation explaining the vulnerability. So we can directly use ordered Probit model to find the determinants of vulnerability.

  22. Regression results (part 1) (Q1)

  23. Regression results (part 2) (Q1)

  24. Propensity Score Matching (PSM) (Q2)

  25. Propensity Score Matching (PSM) (Q3)

  26. Synthesis of results • Age of household head, years of schooling of household head and household size are the important statistically significant determinants. • wage earning households are more vulnerable than the other households. • self-employed household head can reduce the likelihood of being vulnerable. • safe-drinking water and access to sanitary latrine, the degree of vulnerability will be reduced by statistically significant amount. • More distance from main road exhibits more vulnerability in our study. • Surprisingly distant from small market and MFI exhibit negative relationship with vulnerability. • All kinds of physical and financial assets show negative relationship with vulnerability.

  27. Synthesis of results • HH income and food expenditure show negative relationship. • Informal loan and purchasing on deferred payment increases the vulnerability. • Interestingly HH unmet loss in Sidr(2007) shows negative trend with vulnerability. • Social safety net program shows positive relationship. • Microfinance membership shows negative relationship with degree of vulnerability. • In PSM, microfinance reduces vulnerability by 6% to 8% in different matching techniques. • Last two cyclones contributed increase of vulnerability by 9% to 14% in different matching techniques.

  28. Policy recommendation • Short term: • Increase safe water and sanitary latrine. • Seasonal employment (constructing roads , dams, cannels) • Apply more saline resistant cropping technique. • Offer credit (bank and microfinance institutions) • Information dissemination on cyclone, shelters, emergencies measures. • Build more cyclone shelters. • Long term: • Improve years of schooling • Community health clinic • Roads and highways • Strengthen local government. • Generate employment in that region.

  29. Thank you

More Related