The budgetary allocation process in public sector institutions: the case of the University of Nairobi

Abstract

This study sought to establish the various relationships existing between the budget allocation in any year and each of the following variables: (i) The budget request for year t (ii) The budget allocation in year t-l' and (iii) The actual expenditure in year t- 1· .in a bid to establish the best predictor of any year's budget allocation. For the research, data from forty University of Nairobi (UON) departments were gathered and a series of simple regression analysis run using the budget allocation as the criterion variable and each of the variables (i) to (iii) as the predictor variables. It was found out that budgeting in the UON is a hierarchical process which starts at the University departments with departmental requests and ends up in Parliament with the approved estimates. Evidence gathered from the data analysis suggests that each of the three predictor variables is highly related to the budget allocation in any year. It specifically, provides evidence to indicate that the budget requests that are sent by the UON to the Treasury through the Ministry of Education, conforms to a large extent, to the guidelines stipulated by the Treasury such that any subsequent revisions are minimal. For this reason, budget requests at this level, tend to act as good indicators of what the budget allocations will be, although they need not necessarily reflect departmental needs. Evidence also suggests that the vote head does appear to have an effect on the strength of the relationship between the budget allocation as the criterion variable and each of the predictor variables. In this regard, it was found to be the Equipment and Furniture vote which had the greatest effect, i.e. it was distinctly different from the other votes. Finally, the findings of this research should be understood and evaluated in light of the limitations of the study, which were mainly twofold. First, the study restricted itself to only a subset of the variables which may have an effect on the budget allocation in any year. Secondly, there were several cases of missing data