Defence Expenditure of the TBVC States*

Markus Reichardt
Research Associate with the Institute for Defence Policy

*Many thanks to Mr W Botha, Development Bank of Southern Africa and Dr M Reichardt, formerly special advisor to Dr Simon Brand, formerly chief executive DBSA.

Published in South African Defence Review Issue No 13, 1993

It is a political reality that government statistics are meant to obscure at least as much as they reveal. This observation is particularly pertinent in the South African context where for years the Pretoria government sought to hide - among other things - the true extent of its financial support for the independent states.

In his budget speech of March 1993, Minister Keys announced that South Africa was providing for about R6 billion direct and indirect budgetary assistance to the TBVC countries. Of this, direct budget assistance amounted to R5,145 billion, whereas the combined budget assistance that these countries are planning on for 1993/94 is R5,216 billion, i.e. already some R70 million overexpenditure.

It is important that clarity regarding the finances of the independent states be improved since the public debate assumes that the amalgamation of homeland structures will yield a great financial bonanza - the so-called apartheid dividend. Disregarding for a moment that existing structures, such as military establishments, cannot be eliminated overnight, this debate suffers from a general lack of facts and figures. The statistics presented below are an attempt to provide some of this much-needed clarity; they represent for the first time, a historical picture of the cost of the TBVC military establishments.

A number of points need to be made a priori about these figures and their reliability. All data was taken from official TBVC publications - either the Auditor General reports or failing that, the Estimates of Expenditures; or received from the relevant governments on request. Internationally, Auditor General Reports represent the most authoritative picture of government finances since they detail the amounts of money actually received and spent, in this case those expenditures for which the Auditor General has seen the receipts and satisfied him/herself that the funds were indeed expended for the intended purpose.

Internationally, Auditor-General Reports tend to be published about 1 to 12 months after the completion of a financial year. In South Africa, and especially the homelands, the process tends to take longer due to the fact that the National Party has, since the early 1970s, found itself progressively unable to live within its means. As a result, the latter part of the financial year has regularly seen the tabling of a supplementary budget through which the government appropriates additional funding. This, together with the intentional confusion surrounding some aspects of the finances of the independent states, has resulted in Auditor-General Reports being delayed by up to three years or in some extreme cases 5 years.

More recent, though theoretically not so accurate figures, can be obtained from the Estimates of Expenditures which are in effect the projected expenditures for the next financial year. Such documents should be published prior to the beginning of the financial year, but again, this has not always been the case. The Estimates of Expenditures also contain figures for the previous year which are the so-called revised estimates. These tend to be the original estimates plus additional appropriations from the supplementary budgets. In the table below it has been indicated whether the figures are actual expenditures, estimates or revised estimates.

The accuracy of the figures varies from country to country. Based upon information provided by analysts who have examined TBVC budgets over the previous decade, the following comments can be made:

Transkei - Of the figures published by the Transkei administration only the total budget figure can be regarded as reliable. Due to the massive financial mismanagement of the Matanzima administration, the Auditor-General was forced to manipulate individual figures to 'accommodate' the misallocation of funds. Financial reporting standards did not improve under the Military Council; no other territory delays the publication of its financial reports for as long as the Transkei. At best, figures provide a rough approximation of actual expenditures which are generally believed to be substantially higher.



Bophuthatswana - Bophuthatswana's financial data have been plagued less by mismanagement than by lack of government accounting skills. Due to persistent attribution of expenditure items to the wrong portfolios, a 10% margin of error must be allowed for.



Venda - The figures published by the Venda government can be regarded, up until very recently, as generally reliable.



Ciskei - Ciskei's figures tended to give a reliable reflection of the territories finances up until the latter years of the Sebe administration. Since 1988 their value has declined drastically as a result of pervasive financial mismanagement of government funds. Under the military regime the figures have remained similarly suspect and should be regarded as conservative.

In summary, the totals given in the table below must be regarded as conservative and are unadjusted for inflation. Finally a comment must be made concerning an unusual characteristic in the composition of the expenditure: between 55% and 75% of total expenditure on security forces in the independent states has generally been spent on salaries. Given the job and salary guarantee to public servants in the TBVC states contained in the most recent draft constitution, this leaves little room for cost savings when the time comes for the integration of forces.

TBVC TOTALS: DEFENCE

For Transkei and Bophuthatswana, all figures are actual expenditure except 1986/87 and 1987/88 which are revised estimates.



For Venda, all figures are actual expenditure except those for 1988/89 and 1989/90 which are revised estimates.



For Ciskei, all figures are actual expenditure except those for 1990/91 to 1992/93 which are revised estimates.