Economic Development, Urbanization, and the Food System in East Africa

I am a co-author on a new article that can be found here. It is titled "The Rise of the Middle Class in East and Southern Africa: Implications for Food System Transformation". The paper makes a few points that have policy implications:

• Urbanization and economic development is occurring rapidly in many Sub-Saharan African countries
• There is a growing urban middle class which is associated with a rapidly shifting structure of food demand - away from staple grains and towards processed and fresh perishable foods.
• The current system is not sufficient to support this demand.
• While many people predict that much of this growing demand will be provided by imports, we show that the share of imported food has actually decreased with rising incomes
• However Africa still imports more than what would be predicted relative to other developing countries, suggesting much more room for improvement in the development of the local food industry
• More research and investment is needed in the development of the food processing sectors and fresh perishable food supply chains in order to 1) ensure food security and decrease reliance on imports, 2) support small farmers, and 3) provide sources of employment for the burgeoning youth population. 

Thesis update

So I am in the last phase of my thesis, getting ready to send out 2000 surveys, 1000 each to the homes in two different clusters of Albuquerque. While I am only looking at two areas, I have to send out so many because probably only 10-15% will respond. It is quite a bureaucratic and time consuming headache to do this. First I had to get Internal Review Board (IRB) approval which required a mountain paperwork, an online course, and three weeks of waiting. This approval is required for anybody doing a study involving human subjects as a way to insure against possible exploitation. Now I am getting the bulk mailing worked out with the UNM campus mail. I will provide them with the materials (2000 surveys, 2000 consent forms, 2000 10" envelopes), and they will put together the smaller, business reply mail pre-paid envelopes that the responds will use to mail back their surveys. They will also sort and stuff all of the large envelopes and send them out. It will take about two weeks to get all of the details sorted and all of surveys ready to be mailed. Finally, It will probably take about 3 weeks to get all the surveys back.

I am saying all this because I am getting impatient. I want to finish my thesis, graduate, and get on with my life. While sometimes I wish that I didn't go the survey route, I think that in the end it will pay off. I decided to do it because it is rare in the literature to see an analysis that uses both qualitative and quantitative analysis. I am interested in a personal and descriptive analysis to complement all the data that has been processed

The goal of my thesis is to identify neighborhoods in Bernalillo County that lack provision of healthy (primarily meaning access to fruits and vegetables) and relatively cheap food. The quantitative side of the analysis involves two things. The first is to determine personal distance, meaning in this case the average network distance (aggregated on a block group scale) of people to various types of food retailers. Block groups are ranked based upon the aggregate median distance to supermarkets, grocery stores, specialty shops, etc. The second is to determine personal mobility, meaning the ability of people to get to these food retailers in a reasonable amount of time. While there is no index to do this directly, It can be inferred based upon a set of demographic and spatial characteristics, such as car ownership, distance to a bus network, unemployment, disability, poverty, education, etc. Finally, personal distance and personal mobility are compared to rank personal accessibility and possibly identify "food deserts" (areas where people clearly lack accessibility and risk facing health risks in consequence).

I have identified two clusters that could be construed as "food deserts" based upon my analysis. While they have equally dismal personal mobility characteristics, they are different in that one has decent access to supermarkets, but not grocery stores, and the other visa-versa. The purpose of the surveys is to determine if the people's perception of personal food accessibility confirms my methodology and results. In the literature, most people assume that supermarkets are the only real source of healthy and cheap food. I want to see if that is true based upon peoples perceptions.

I am hoping that this analysis will help to better target "food deserts" in Bernalillo County so that businesses and non-profits can better serve these areas. There is a push in many parts of the country to encourage local agriculture in urban areas that lack it. There are also other endeavors, such as food on wheels, which takes healthy food on mobile carts directly to the areas that are lacking.

I am also hoping that I have made a contribution to the spatial accessibility and food accessibility literature by improving upon some methodology. In addition to paring qualitative analysis with quantitative, there are some other methods that I feel are somewhat original.

1) Frequently I see people walking, carrying bags of food. For this reason I set up the network to account for the fact that many people do not own vehicles. bike paths and arroyos were included with rodes as plausible walking routes.

2) Network analysis is a way to calculate the distance from one point to another along a given network. In order to make distance information comparable to census data, it must be aggregated to a census scale, such as a block group. Researchers often estimate the distance of a population from the geometric center of a block group to the nearest facility. The problem with this (in addition to the fact that the resolution is bad) is that the population is not homogeneously distributed through each block group. The geometric center is therefore often a poor measure of average distance. What I did is to find the distance from every residential address point to the nearest food store, and then take the median of those distances within the block group. While the measure is still not perfect it better accounts for heterogeneous population distributions.

3.) Most of the literature assumes that supermarkets are the only source of food. It is widely known that local neighborhood shops are more expensive and have less selection, but they should not be neglected in the analysis. Farmers markets should also be considered, since they are tied to local agriculture and cultural traditions. That is one reason why in my surveys I want to compare people's perception of food provision in different kind of food environments.

This thesis has taken longer than I hoped, and has required more work than I would have liked. There have been many setbacks. At one point my hard drive was stolen containing weeks worth of work. A few times I have realized that something was done wrong a few steps back in a work-flow process, causing me to have to redo the entire process. On the other hand I have learned more doing this than I could have by taking any class. In classes, the data is already clean and the methodology is taught to you. In this I had to find new ways of doing things and work around intractible problems. It has been quite the experience

Initial Thesis Proposal

I am interested in hearing feedback on this. My ideas are still in development

Research Proposal

Background:

Poor diet is the cause of many adverse health effects, such as heart disease, diabetes, obesity, and many others (Must et al. 1999). There is a growing body of literature linking the lack of accessibility to healthy, diverse, and cheap food with bad dietary habits and food insecurity (Morland et al. 2002, Cummins and Macintyre 2002). Low accessibility areas, often coined “food deserts”, have been defined as areas that are both far from suitable retail food outlets, and are inhabited by those without suitable access to transportation, often those with low income, minorities, and the elderly (Apparicio et al. 2007, Zenk et al. 2005). There has been a growing social concern over the existence of food deserts in the last decade, as large supermarkets have moved out into the suburbs (Clark et al 2002), leaving many inner city and low income areas reliant on small independent stores (Eisenhauer 2001), many of which lack healthy food and are more expensive (MacDonald and Nelson 1991, Chung and Myers 1999).

Various indicators have been developed over the years to study the spatial variability in access to food retailers. These include simple provision indicators, accessibility indicators, and model-based indicators. An example of a simple provision indicator is retail grocery square feet per household. The more sophisticated model-based indicators attempt to investigate the level of population-store interaction by estimating the demand and supply of a given area (Clark et al. 2002). Spatial accessibility has most commonly applied to hospital access, and consequently this application is where many of the more advanced measurement techniques have been developed. These have been simplified into four general categories as “provider-to-population ratios, distance to nearest provider, average distance to a set of providers, and gravitational models of provider influence” (Guagliardo 2004). It has been observed that results can vary significantly depending on the measurement being used. This has prompted a recent study in Montreal that combines 3 different measurements, proximity, diversity, and variety, to measure spatial accessibility to food. Proximity is derived by using network shortest path distance, diversity is derived by measuring the total number of supermarkets within a walkable distance, and variety is derived by measuring mean distance to three different supermarket chains. The authors then go on to correlate this combined measure to a social depravation index using five variables to identify food deserts. These variables, based on percentage of the total population, include income, single parent families, unemployment, education, and immigration (Apparicio et al. 2007).

To my knowledge, there have been no studies locating the existence of food deserts in Albuquerque. It is an open question whether even the most sophisticated methodologies, particularly the recent Montreal study, would apply to Albuquerque in all respects. Albuquerque is a historically agricultural area, even though it has almost ceased to be a source of primary income for most farmers, especially the younger generation (Friends of Albuquerque 2006). Nevertheless, there is still an abundance of food sold in farmers markets and neighborhood food stands. Many of these agricultural neighborhoods might also be those that fit the common criteria as being food deserts.

Statement of Problem:

I want to contribute to the methodologies of identifying ‘food deserts’ by testing the common assumption that supermarket chains are the only indicator of healthy, cheap, and diverse food, especially for places similar in character to Albuquerque.

Objectives:

1.) I want to replicate the Montreal study, which appears to be the most sophisticated in assessing food accessibility, in Albuquerque.

2.) I want to revise/improve this methodology to reflect local agriculture by surveying where people get their food in areas that would be considered ‘food deserts’.

3.) I would then develop a set of metrics to reflect differences in diversity, health, and price between local and supermarket food retailers.

4.) I would use these metrics to possibly redefine ‘food desert’ boundaries in Albuquerque.

Method and Data Considerations:

1.) In identifying supermarkets, I will need to get access to a licensed database of food stores and geocode them into a map. I would also need to do some field checks to test the quality of the information.

2.) I will be relying heavily on census information to create a social depravity index in similar fashion to the Montreal study.

3.) I will need to develop a stratified random sample in order to survey the areas identified as food deserts.

4.) I need to do some more research on developing the metrics necessary for comparing the food characteristics of different food retailers. This is where I might need to get a committee member with a background in economics.

Discussion:

I hypothesize that in Albuquerque, the assumption that supermarkets are the only provider of food accessibility will be true regarding food diversity, but not always true regarding food health and price. If this is the case, then I believe that this study can help to more accurately define food deserts in Albuquerque and other urban environments that have similar characteristics. It might also contribute to the debate about how to alleviate food deserts. A few studies have pointed out the location of supermarkets in food deserts might actually exacerbate the problem, by putting many of the local vendors out of business and creating ‘micro-food deserts’ for those that cannot travel to shop, such as retirees, mothers with children, and the disabled (Clark et al. 2002). Although it might be outside the scope of my research, future research might investigate the possibility that a policy of supporting the creation of local farmers markets in ‘food deserts’ provides better food accessibility in some areas than that of attracting large supermarkets.

References:

1. Apparicio P, Cloutier M, Shearmur R: The case of Montreal’s missing food deserts: Evaluation of accessibility to food supermarkets. International Journal of Health Geographics 2007, 6:4.

2. Chung C, Myers S: Do the poor pay more for food? An analysis of grocery store availability and food price disparities. Journal of Consumer Affairs 1999, 33(2):276-296

3. Clarke G, Eyre H, Guy C: Deriving indicators of access to food retail provision in British cities: studies of Cardiff, Leeds and Bradford. Urban Studies 2002, 39(11):2041-2060.

4. Cummins S, Macintyre S: A systematic study of an urban foodscape: The price and availability of food in Greater Glasgow. Urban Stud 2002, 39(11):2115-2130.

5. Eisenhauer E: In poor health: Supermarket redlining and urban nutrition. Geojournal 2001, 53:125-134

6. Friends of Albuquerque: Friends of Albuquerque’s environmental story 2006

http://www.cabq.gov/aes/s2sv.html

7. Guagliardo, M: Spatial accessibility of primary care: concepts, methods, and challenges. International Journal of Health Geographics 2004, 3(3).

8. MacDonald J, Nelson P: Do the poor still payer more? Food price variations in large metropolitan areas. Journal of Urban Economics 1991, 30(3):344-359

9. Morland K, Wing S, Roux A, Poole C: Neighborhood Characteristics associated with the location of food stores and food service places. American Journal of Preventative Medicine 2002, 22(1):23-29

10. Must A, Spadano J, Coakley EH, Field A, Colditz, and Dietz WH: The disease burden associated with overweight and obesity. Journal of the American Medical Association 1999, 282:1523-29.

11. Smoyer-Tomic K, Spence J, Amrheim C: Food Deserts in the Prairies? Supermarket accessibility and neighborhood need in Edmonton, Canada. The Professional Geographer 2006, 58(3):307-326

12. Zenk S, Shulz A, Israel B, James S, Bao S, Wilson M: Neighborhood racial composition, neighborhood poverty, and the spatial accessibility of supermarkets in metropolitan Detroit. American Journal of Public Health 2005, 95(4):660-667.