Average crop yield (2001–2017) in Ethiopia: Trends at national, regional and zonal levels

This article presents average agricultural yield data per hectare for key cereal, legume and root crops from 2001 until 2017. Data was obtained from the annual Agricultural Sample Surveys of the Central Statistics Agency (CSA) of Ethiopia. We present data at national, regional (SNNPRS) and zonal (Wolaita) levels. The data shows that average yields for all crops, at all levels, show increasing trends during the time period. Data for the main cereal crops is consistent and aligns with literature relatively well, however we raise questions about the root crop data in an effort to encourage greater critical reflection of components of data from the CSA.


Experimental factors
Data used in this article were obtained from the Central Statistics Agency of Ethiopia, with reference to available literature.

Experimental features
Tables and graphic trends of analysis were employed.

Data source location
Ethiopia

Data accessibility
The data are with this article.

Value of the data
Average agricultural data are presented for key cereal, legume and root crops from 2001 to 2017. The data can be used by researchers and policy makers to analyze the implications of agriculture products on food security and poverty reduction.
Average yields for all crops, at all levels, show increasing trends, with cereals doing so progressively and root crops increasing rapidly in recent years.
Based upon some components of the governmental data, questions are raised about accuracy, encouraging researchers to be more critical when utilizing these data sets.

Data
The figures and tables of agricultural data were obtained from the annual Agricultural Sample Surveys of the Central Statistics Agency (CSA) [1][2][3][4][5][6][7][8][9][10][11][12][13], covering the time period of 2001 until 2017. The CSA is the only provider of data at this scale. Average yields for all crops, at all levels, show increasing trends, with cereals doing so progressively and root crops increasing rapidly in recent years (Figs. [1][2][3][4][5]. All the data is presented on a year-by-year basis in Tables 1-3, enabling ease of re-analysis. However, there are general concerns about the quality, methodologies, and politicization of data produced by central statistics agencies [14]. We present data at national, regional (Southern Nations, Nationalities and Peoples' regional state; SNNPRS) and zonal (Wolaita) scales. The data for the major cereals (teff and maize) is relatively consistent with the literature, whereas the shifts as well as contrasts with the literature in root crops raise questions about components of the agricultural data. For example, 1) In the 2012/13 season yields per hectare of taro and sweet potato tripled, according to CSA personnel this was due to methodological changes (Tables 4-5) [15];         Alternative surveys of the required scale do not appear feasible or realistic at this time. However, the questions above highlight the need for more research to assess the data provided by central statistics agencies. Often these data sets are utilized without critical reflection about quality, methodology or politicization.

Experimental design, materials and methods
Average crop yield data at national, regional (SNNPRS) and zonal (Wolaita) levels (see Map 1) were obtained from the CSA annual Agricultural Sample Surveys. The data is presented using figures to highlight trends and tables to allow for further analyses of the data. We have selected SNNPRS as an example region and Wolaita as an example zone primarily due to our familiarity with the areas respectively, and thus enhancing our ability to identify questions. The objective of raising questions about the agricultural root crop yield data is to encourage researchers to engage with central statistics data more critically. This does not suggest that the CSA data is inaccurate; rather it acts an encouragement for CSA data to be a subject of greater study.

Transparency document. Supplementary material
Transparency document associated with this article can be found in the online version at http://dx. doi.org/10.1016/j.dib.2017.12.039.