The case for a professional public service

In September 2023, I wrote about how the lack of a capable state destroys infrastructure at the local level. In the run-up to national and provincial elections, it is worth revisiting this topic. 

It doesn’t matter what policies the post-29 May governing party or coalition has if the state is incompetent. I have heard some parties talk of “professionalising” the public service. Even the ANC in the past has used this rhetoric. That’s all well and good but what would it take? Cadre deployment is routinely attacked but the actual reasons for its existence and failure are not interrogated.

There are two core issues: concentrating experience and accountability at the top and effective consequence management. Both of these are wrong in the current approach to public service. Top jobs are short contracts and the consequence machinery is far too slow. These two factors interact to ensure that the public service is shambolic at best, deeply dysfunctional at worst.

The top jobs at all levels — municipal or city manager at municipal level, director general at provincial and national level — are five-year contracts, sometimes shorter, to coincide with the end of the term of the elected body. 

With only five years in office, the top official barely has time to learn the ropes and take full charge of their municipality or department. The excuse for this approach from the ANC, which sets the rules as the governing party, is that they like working with a top official who is close to them. 

In the early post-apartheid era, there was some merit in recycling officials steeped in apartheid culture. However, 30 years on, we have to examine whether this practice is working.

In some long-established democracies, it is the practice to have permanent heads of departments. For example, in the UK, a government department is headed by a permanent secretary. 

In the US, the top public service level is somewhat diluted by political appointments but there is a strong layer of permanent employees who know how things operate, called the senior executive service. So it is possible to have a professionally-run public service with some political appointments at the top but you need a strong core group of non-partisan professionals who constitute the hands-on management.

Where cadre deployment goes badly wrong is when the short-term appointments at the top result in a revolving door between public service, politics and procurement (whether suppliers or contractors). 

In one municipality that I know of, the municipal manager is a former mayor and the director of infrastructure is a former acting chief executive of the implementing agency for major water contracts. With such close relationships between what should be independent entities, there is potential for corruption. 

Another problem is that public servants who have high positions in the governing party’s own structures might believe they outrank their political principals.

Professionalising the public services should start with three reforms: making the top positions permanent, making it unlawful for public servants to hold office in political parties and requiring a gap of at least five years between working for any service provider, or contractor to a government entity, before being employed in the public service. 

Equivalently, the last provision could be enforced by disallowing service provision or contracting by any entity that has employed a public servant for at least five years after they parted company with that employer.

All of that however does not address consequence management; it does however create the possibility that a senior official is still there to account once wrongdoing is exposed. The other missing piece of the puzzle is speeding up accountability.

The 2018 amendment of the Public Audit Act permits the auditor general to issue a certificate of debt when an accounting officer fails to recover fruitless, wasteful or irregular expenditure from the responsible person. The problem with the process is that it takes 22 months from discovering a material irregularity to issuing a certificate of debt.

When I first came across the flow diagram for the process on the auditor general’s website, I thought it was Elon Musk’s plan to get to Mars. Why does it have to be so complicated? All that needs to be done is to give the responsible person a fair opportunity to respond. 

There are two big problems with such a slow, cumbersome process. First, audits are carried out after the event, so the timeline for accountability could start up to a year after the cause. With a 22-month process, an accounting officer with a five-year contract in practical terms cannot be called to account for any wrongdoing past their first two years as they could be out of the door and untraceable if any material irregularity is found after that.

Another issue is the slow and tedious processes for getting the Special Investigating Unit (SIU) to act. Before a criminal investigation or civil cost recovery can start, a presidential proclamation has to be signed.

In one case I have been following, it took more than two years after the SIU became aware of the case before it was put on the president’s desk to be signed — where it sits at the end of a long queue.

And even once the SIU springs into action, miscreants can fight off consequences through the courts for years, taking advantage of the highly congested rolls of our courts.

While this is going on, criminals are not sitting idle. So what is to be done? 

First, the certificate of debt process must be streamlined. Even if this could result in some cases being thrown out by the courts, recovering funds from those responsible will send a clear message that stealing has consequences. 

Likewise for the SIU. The time from being aware of the need for an investigation to signing a presidential proclamation needs to be radically shortened. Allied to this is reducing the backlogs in the courts and this brings me to my final reform — introducing a corruption court. 

In South Africa, we already have specialist courts that have the same status as the high court. The labour courts have exclusive jurisdiction of labour disputes; the electoral court has jurisdiction over disputes with the Electoral Commission of South Africa (IEC) and within parties, though not exclusive of the high courts. 

A corruption court could have the same status as the electoral court  — a court that is designed to expedite cases that would otherwise be heard by a high court.

How would a corruption court differ from the regular high court? As with the electoral court, it could have rules that expedite hearings, with short timelines. The electoral court is, by legal standards, extremely nimble, as it has to be if it is to ensure adherence to the electoral timetable. 

A corruption court would not have to be quite so nimble (the electoral court allows three days to take an IEC decision on review, for example) but could nonetheless be configured to hear cases expeditiously. 

Two factors could aid this, aside from short timelines: specific requirements on the nature of the case, so that only cases that could be heard quickly would apply, and judges specialised in such cases, so they can decide them quickly. 

Allied to this is funding prosecutors with accounting skills or backed up by accountants. While this could be expensive, so is corruption.

In summary, I advocate professionalising the public service by doing away with short contracts at the top level, doing away with the revolving door between public service, procurement and political office and making accountability more nimble.

Philip Machanick is an emeritus associate professor of computer science, Rhodes University, and a Makana Citizens Front councillor.