Populations of Priority for Potato and Poverty

This study is adapted from: Theisen, K. and Thiele, G.. 2008. Implementing CIP's Vision: Impact Targeting.
International Potato Center (CIP), Lima, Peru. Working Paper 2008-4.

The original working paper analyzed the comparative importance of both potato and sweetpotato. Only potato is included here.


In 2003-04 the International Potato Center (Centro Internacional de la Papa, CIP) conducted a study, published as The CIP Vision, to examine how CIP's research program is focused on the primary goal of reducing poverty in areas of the world where potato or sweetpotato is an important crop. Part of the study included a global targeting analysis to better identify populations by area that are of highest priority to CIP.

Since poverty is a complex subject which cannot be adequately measured only by monetary income, the study utilized several indicators to obtain more comprehensive measures of livelihood, including:

  • Income: the percentage of a population with incomes below US$1 per day, adjusted to purchasing power parity (PPP), a widely cited indicator useful for global analysis;
  • Malnutrition: the percentage of a population classified as chronically malnourished, based on estimates of the amount of food available in each country and a measure of inequality in distribution derived from household income and expenditure surveys;
  • Child Mortality: the number of children who die before reaching five years of age, per 1,000 live births annually;
  • Maternal Mortality: the number of women who die during pregnancy or childbirth per 100,000 live births.

Data for these indicators were obtained from the United Nations Millennium Development Goals site (MDG), combined with cropping data for potato and sweetpotato, usually from the United Nations Food and Agriculture Organization (FAO).

The 2004 CIP Vision study was updated in 2008, and revised in two ways:

  • The four MDG livelihood indicators are combined into a more robust composite score for each country;
  • Crop production data have been analyzed at a more geographically refined level, sub-nationally wherever feasible and appropriate, and classified according to production per capita.

Composite Scores of Livelihood

In order to provide a comprehensive measure of livelihood which can be displayed on one global map, each of the original four indicators is classified into six categories, "0" to "5." Classification of each of the four indicators is determined primarily by natural breaks, where differences are greatest between adjacent observations. Since the entire population of any country is assigned the same value for an indicator, it is not feasible to classify countries by quantiles, so that each class would contain a roughly equivalent number of people. Much of the world's population of highest priority to CIP is contained in only two countries, China and India.

Composite country scores based on these four indicators also range from "0" (highest indicator of livelihood, of lowest priority to CIP) to "5" (lowest indicator, and highest priority to CIP).

A composite score for a country is generally based on the sum of the scores of the four indicators, with a highest possible sum of 20 where all four indicators are reported. However, if one indicator is strongly divergent from the average, the higher score(s) is given greater weight. This provision is intended to account for the possibility that an indicator is either inaccurately reported or is not locally relevant. If, for example, the four livelihood indicators for a country are scored as "0," "1," "4," and "5," the higher scores (indicating lower livelihood) would be more heavily weighted.

Nigeria, for example, is reported at a very high 70.8 percent of the population living at income below US$1, but a much lower nine percent chronically malnourished. This wide disparity of data is especially doubtful considering that Nigeria's population is estimated to be 47 percent urban (PRB), and thus likely to be cash-dependant for subsistence needs. A classification based only on nutritional status would therefore incorrectly label Nigeria as a lower priority country. In this case, the income score is effectively given greater weight.

A single indicator can be essentially correct, but nevertheless provide an incomplete account of livelihood. Although the US$1 per day indicator is widely used for global comparative analysis, it is not always equally relevant across regions. Several countries report low scores of incomes below US$1, but significantly higher scores on other indicators. Armenia, for example, reports two percent of the population living below the US$1 level (for a score of "1"), while 29 percent is reported as chronically malnourished (for a score of "4").

The four indicators are not available for every country. Where the data set is incomplete, reporting three or fewer of the four indicators, the final composite country score is based on those which are reported. In these cases, the higher scores are weighted still more heavily, to account for the possibility that the factors indicating lowest livelihood conditions are those least likely to be reported.

Tables 2 and 3 display the analysis used to determine the composite livelihood score and the actual score for each country.

Although the data used to determine livelihood scores are reported by country, the real concern of this study is with populations. For this reason, the global distribution of populations, classified according to composite livelihood scores as indicated in Table 3, is displayed cartographically below.

     Map 1:  Population by Livelihood Indicators
     (To view this map, click on the thumbnail and expand the image to full screen display.)

Per Capita Crop Indicators

Indicators of the importance of potato and sweetpotato were derived in the 2004 targeting study by an analysis of both crop production and land use at the national level, except for China, India, and Russia, where sub-national data were utilized. Since then, CIP has further investigated both crops using more detailed sub-national data wherever possible. Utilizing more complete and current crop and population data, this study rates the local importance of potato relative to population, not to land, classified according to production in kilograms per capita for a given area.

Some countries can be adequately represented by a single value of production per capita, but where there is considerable variability within a country, finer estimates based on additional sources are used wherever possible. In most cases, these estimates are reported by agencies of the respective government and are consistent with FAO data. In a few countries, the analysis is based on more detailed crop surveys which are not consistent with FAO. (Examples include Cameroon, Ethiopia, and Malawi. See "Cropping Data Notes by Selected Countries" for more details.)

More specific reasons for representing a country by a single per capita value, or several, are indicated below.

1. One per capita value is displayed for the entire country, for one of four reasons:

1A.  Potato is produced for countrywide consumption. 

Potato is in some cases grown in a few areas of intensive production, as in the United States, but it is widely marketed and consumed within the country (and in some cases exported internationally). One value of average per capita potato production for the entire country is therefore appropriate. Other countries within this category include Canada, Japan, Australia, and all countries of Western and Central Europe.

1B.   The distributions of people and potato crops generally coincide. 

Potato might be produced either for local consumption or for commercial markets, but crops and people are generally concentrated in the same areas, so that the local importance of potato is likely to be fairly consistent across populations within the country. Examples include most countries of Central Asia and North Africa.

1C.   Within-country data are not complete or reliable. 

This category includes countries where sufficient and reliable data are not available to estimate variable per-capita production. Russia is an obvious example, where production data are available via FAO only for the entire country. Several countries in Africa are likewise assigned to this category.

1D.   The country is very small in area and/or population. 

Given the limitations of cartographic display for a global map, one value is reported for small countries, such as Bhutan and Albania.

2. Values are estimated and displayed at variable rates within the country for one of two reasons:

2A.   Potato is produced for a mix of distant markets and local use. 

In these countries, demand for potato by urban populations is often a factor of growing importance, but the local importance of the crop (as reflected in marketing and/or consumption) is highest in areas of high production, often far from population centers. Estimates have been made based on the literature for each country, especially those included as atlas chapters. Annual per capita use of potato by people living in major cities of Asia is generally estimated to be around 20 kilograms. For people living in major cities of Latin America, per capita potato use is typically 50 to 60 kilograms. Countries in this category include China, India, Turkey, Morocco, South Africa, Kenya, and all countries with a significant potato crop in South and Central America.

2B.   Production is mostly for local consumption and/or local markets. 

Where potato is significant, but rarely travels far after harvest, per capita production is estimated from the best available within-country data of production and population. Potato can be a very important crop in one region of a country and unknown elsewhere, accounting for per capita production values as low as zero in some cases. Countries in this category include several in Sub-Saharan Africa, such as Ethiopia, Cameroon, and Democratic Republic of Congo.

To estimate values of production per capita, population data provided by the Environmental Systems Research Institute (ESRI, the creator of ArcGIS/ ArcMap software) have in many cases been revised, usually upwards, based on a consensus of other sources, including the Population Reference Bureau (PRB), United Nations data reported by the Food and Agriculture Organization (FAOSTAT), and Population Statistics (Populstat). (See "Data Sources" for links.)

These data are summarized by country and displayed by population in the table and map below:

     Map 2:  Population by Per Capita Potato Production
     (To view this map, click on the thumbnail and expand the image to full screen display.)

Areas of High Priority

Areas of highest priority to CIP, determined by a combination of livelihood indicators and the importance of potato, are displayed below. Areas characterized by lowest priority ("0" values) for either factor are not displayed. Areas rated "1" through "5" in both factors have been classified into one of 25 possible combinations, the most important overall being those where livelihood indicators are most severe, and potato is locally very important. Areas displayed in vivid red or orange are those of overall highest priority.

     Map 3:  Population by Priority for Potato
     (To view this map, click on the thumbnail and expand the image to full screen display.)
The countries identified by the 2004 CIP Vision study generally remain those of highest priority in this updated study, but analysis at a finer geographic scale can reveal large populations of high priority that would be obscured by data reported only at the country level. For example, potato is generally a minor crop in Cameroon, in most areas not cultivated at all. However, in the western highlands, home to over six million people, per capita production ranges from around fifty to over seventy kilograms (WPA, Cameroon).

Areas of high priority include, with a few examples:

  • Higher altitude areas of several countries of Sub-Saharan Africa (Ethiopia, Cameroon, Kenya, Rwanda, Uganda, Tanzania, Malawi, Angola, Mozambique, and Madagascar);
  • Andean South America (Bolivia, Peru, Ecuador, and Colombia);
  • The Indo-Gangetic basin of southern Asia (Bangladesh, India, Nepal, and Pakistan);
  • China, with high production found in several interior provinces;
  • Central and western Asia (Tajikistan, Kyrgyzstan);
  • The Caucasus region (Armenia, Azerbaijan).

In addition to better identifying populations of high priority at a finer scale, sub-national mapping can also reveal areas where data quality is doubtful, typically where sharp transitions conform to administrative boundaries absent other geographic factors. In China, for example, the coastal province of Fujian is reported to be a high producer of potato, but very little production is reported from adjacent provinces. It is possible that Fujian is over-reported, or that other areas are under-reported, or that some combination of both factors occurs.

Further Applications

This study is limited to actual production data of potato, however tentative such data often are. What has yet to be considered is the unmet potential for potato to provide an important role in food security and economic development in areas which are physically suitable, but where effective research services and other infrastructure are not available. For example, potato could be a very significant crop in the higher altitude areas of the Kivu District of the Democratic Republic of Congo, but if farmers have not had access for several years to either new varieties or sources of seed free of viral infection, that potential is not likely to be realized. An assessment of the production possibilities for potato, not only its current production, could provide further guidance to CIP of where its research efforts could have the strongest effect.

Since this analysis is global and therefore highly generalized, it can provide only a first step toward identifying populations of high priority to CIP. This effort can be extended and refined via the use of Geographic Information Systems, utilizing data provided by more complete surveys and by more technically advanced methods, such as remote sensing where feasible. More refined and regularly updated analysis, displayed by global and more localized maps, provides a significantly improved framework to guide CIP's research and development program to those areas of the world where we may expect the greatest impact on the livelihoods of poor people.

Data Sources

Indicators of Livelihood

Indicators are provided by the United Nations Millennium Development Goals Indicators (MDG). These data are occasionally revised.

Databases utilized for this analysis include:

  • Goal 1: Population below $US1 (PPP) per day, percentage (data reported various years, none prior to 1997);
  • Goal 1: Population undernourished, percentage (data reported 2002);
  • Goal 4: Children under five mortality rate, per 1,000 live births (data reported 2005);
  • Goal 5: Maternal mortality ratio per 100,000 live births (data reported 2000).

Cropping Data

The primary source of cropping data for potato at the national level is FAO, with several exceptions and revisions for countries marked with an asterisk in Table 4. FAO data are frequently collected and reported by the governments of each respective country, hence are of widely varied quality in terms of accuracy and completeness. In some cases, national level data reported by FAO have been utilized, with the addition of other sources for sub-national estimated distribution.

More complete information of cropping data is available for countries included in the World Potato Atlas (WPA). Notes on other countries marked with an asterisk, but not included as atlas chapters, are listed below. (WSA refers to the World Sweetpotato Atlas.)

Cropping Data Notes by Selected Countries

Angola: FAO data have indicated rapid increases in area cultivated, from 11,500 hectares in 2001 to a tenfold increase of 115,000 hectares in 2006. National distribution has been estimated according to a July 2003 report of potato and sweetpotato production issued by the FAO and the World Food Programme. (See the World Sweetpotato Atlas Angola chapter.)

Burundi: Potato area has been reported by FAO at a consistent 10,000 hectares over the past several years, very likely well below potential and actual production, especially considering production under similar physical conditions in adjacent Rwanda. The country is, for the time being, classified as "no data."

Malawi: Potato production data are displayed as reported via the Famine Early Warning System (FEWS) for 2004, not as reported via FAO. FAO reports much higher production of potato and nothing of sweetpotato (grown much more extensively than potato), apparently reflecting a confusion of the two crops.

Nigeria: Data reported by FAO indicate a gradual but consistent national upward trend over the past five years, known from other sources to be concentrated in areas of higher altitude, centered on the Plateau and Kano States. Cropping data in general are very tentative for Nigeria. (See the WSA Nigeria chapter.)

Papua New Guinea: Potato production is not reported by FAO, but some potato cultivation is reported in other literature cited in the WSA Papua New Guinea chapter.

Rwanda: Sub-national distribution of potato production reported by FAO was estimated via literature cited in the WSA Rwanda chapter.

Uganda: Sub-national distribution is estimated by Government of Uganda reporting for 2003, updated with national production estimates reported by FAO.

Population Data

Population data provided by ESRI, the company that produces ARC Geographic Information Systems software used in this analysis, have been updated by composite figures provided by three other sources:

  • Population Reference Bureau/ Datafinder (PRB);
  • United Nations, FAO Data Archives/Population;
  • Populstat, especially for sub-national population estimates.
  • None