The channels of poverty reduction in Malawi : a district level analysis
Abstract
The study investigated on the channels of poverty reduction in Malawi, using household data aggregated at district level. Malawi is divided into 31 districts with different demographics and opportunities. Macro level data which was calculated in terms of district percentages were used in the study. The study emanated from the premise of the link between economic growth and poverty reduction. With the trend of growth that was seen in Malawi from 2004 to 2012; there was an interest to further investigate if there had been any significant change in the poverty levels as measured in the country by the National Statistical office. The objectives of the study were two pronged; the theoretical and the imperial. The theoretical objectives were; to provide a background of Malawi, to review the literature on poverty theories, to review the literature on the link between poverty reduction and the channels of potential impact, namely: economic growth, education attainment, access to loans and enterprises, agricultural production, population growth and employment or unemployment. The empirical objectives on the other hand were; to investigate if there has been any poverty reduction in the years 1998 to 2012 in Malawi, to assess how economic growth at a district level proxied by agriculture production and land holding affect poverty at district level in Malawi, to assess how education attainment affect poverty reduction at a district level in Malawi, conduct an analysis on how employment or unemployment affect poverty reduction at a district level. Also investigate the relationship between access to loan and poverty reduction in Malawi and to determine if different poverty measures exhibit statistically significant different responses to channels under investigation namely economic growth, education levels, population growth and access to loans at district level. The study employed descriptive and regression analysis to arrive at the results for the set empirical objectives. Due to the fact that panel data was used for districts, a random effects regression model was used for the estimations. A Breusch-Pagan test was used to decide on random effects as opposed to fixed effects model. The results from the regressions showed that all the channels that were hypothesised to be of importance, came out significant from objective based regressions. These regressions were run separately for each channel, with the district poverty rate as a dependent variable. The study found the considered channels of poverty reduction to be significant at different levels. First, it was established that there has been significant growth in Malawi. This growth however was seen to be erratic where in other years it was higher and in other years lower. A more important conclusion from the first objective was that there had been poverty reduction in the country between 1998 and 2012. A t test was also used for mean difference in the years where Integrated Household surveys were conducted namely, 1998, 2004 and 2012. The t-test showed a statistically significant reduction in poverty between 1998 and 2012 of up to 15.07. The study also found that the relationship between agricultural production and poverty was significant especially looking at local maize production which had a negative significant coefficient. Implying that, an increase in agricultural production has an associated reduction in the district poverty rate. It was also established from the results that input subsidy had a significant impact on poverty at district level. This input programme which helps poor households to access fertilizer at a highly subsidised price had a negative relationship with poverty that was significant. This shows that government’s effort in funding the national wide fertilizer subsidy has some bearing on the poverty level of the country. On the relationship between education and poverty reduction, the study also found a significant relationship. This was clear on the impact of literacy rate on poverty reduction. The regression results showed a significant negative relationship between literacy rate and poverty reduction. The channels of employment in poverty reduction was found to be significant but in a direction unexpected. Labour force participation had a positive influence on poverty rate at district level. A number of things were discovered; first the employment rate as reported in the statistical year book is misleading. What is considered employment in these statistics is basically subsistence farmers who take up more than 80% of the employment rate. Second, most of what is recorded as employment is non-skill labour with people without education recoding a 99% employment rate. This is a misleading record in as far as what employment for poverty reduction is concerned. It is therefore not a surprise that, most of the people reported as employed are also found below the poverty line some even below the ultra-poverty line. A special contribution resulting from the study is the framework on the interconnection between the channels. The study points out the fact that for agricultural production to thrive there is need for education. Also for agricultural production to succeed there is need for the farmers to have access to loans. the study discovered that more than 45 percept of the loans people obtained were for agricultural inputs. There is also a link between education and employment, education and access to loans and access to loans and employment through business start-ups that create employment. The conclusion of the study is that policies that are intended to reduce poverty should be aimed at promoting education participation. There is also need to create an environment that enables the poor to access loans and credits at a reasonable interest rate. The government should continue with the input subsidy programme for the poor household. There is need for the national statistical office to reconsider the definition of employment so that the government works with practical figures, other than the inflated employment rates that are reported in the statistical year book.