Identifying agricultural expenditures within the public financial accounts and coding system in Ghana: Is the ten percent government agriculture expenditure overestimated?

This paper is part of four country case studies that take a detailed look at public expenditures in agriculture, and at how the data on expenditures are captured in government financial and budget accounts. The objective of these studies is to unpack the black box of public expenditure statistics reported in various cross-country datasets, and ultimately to enable the use of existing government accounts to identify levels and compositions of government agriculture expenditures, with better understanding of what these data are in fact accounting for.

Since the endorsement of the Maputo Declaration in 2003, many African countries have reported larger government agriculture expenditure (GAE) shares. Do those reports reflect changes in actual allocations to the sector or changes in the accounting system? The 10 percent allocation of total government expenditure (TGE) to the agriculture sector that the government of Ghana claims has come under increasing scrutiny, particularly following recent budget statements that show less than 2 percent allocation to agriculture. This discrepancy reflects the confusion in defining the agriculture sector and in distinguishing expenditures of the government from expenditures of public corporations, which together makes up expenditures of the public sector. This is exacerbated by the lack of clarity in the public financial accounts and coding system for identifying and aggregating budgets and expenditures of the Ministry of Agriculture and other public departments, agencies, and corporations that carry out agricultural functions.

This paper analyzes how GAE may be consistently identified and aggregated within the new public financial management and accounting system to generate more reliable information on the amount and patterns of GAE. Having accurate GAE data has important implications for agricultural development beyond just monitoring progress toward the goals of the Maputo Declaration. It is necessary for generating more reliable estimates of the relationship between outcomes and GAE, and in turn for making better evidence-based planning and implementation of agricultural policies and programs that lead to greater and better-distributed outcomes.

The paper uses data from the 2012 consolidated accounts of the government of Ghana and from the 2011 accounts of public corporations in Ghana to analyze the share of GAE in TGE using different definitions of agriculture expenditure and different estimators of the share. The share of GAE in TGE is estimated at 1.7 to 3.4 percent when only expenditures of the government are considered in estimating GAE. These shares are much lower than the 8 to 11 percent reported in previous public expenditure review studies. In those studies, expenditures of the Ghana Cocoa Board (which is a major public corporation in the agriculture sector) obtained from other sources were added to GAE obtained from the consolidated accounts to derive public agriculture expenditure (which was used as the numerator for estimating the share). That should have been balanced by adding the expenditures of all public corporations to TGE to derive total public expenditure (which should have been used as the denominator for estimating the share), because the consolidated accounts of the government already reflect net transactions between the government and all public corporations. Otherwise the numerator of the estimator may be overestimated and the denominator underestimated.

Different patterns of GAE for the different definitions of agriculture are presented. Then, implications for data standards for an integrated financial management and information system to enhance program-based budgeting are drawn.