Protecting the guardrails of democracy from the impulses of populist politicians…

Just over 11 years ago, when the world still believed that the veneer of multilateralism could be maintained, the United Nations General Assembly [UNGA] formally endorsed the Fundamental Principles of Official Statistics. This stamp of approval represented the culmination of a long journey aimed at promoting and cementing the idea that official statistics were an important part of public governance.

In the words of the UN, ‘official statistics – which must be reliable and objective information – are crucial for decision makers and citizens alike’. In democratic societies – where a politician’s popularity is fundamental to their ability to galvanise the vote – the tendency to commit to hyperbolic and ambitious targets is ever present.

In the 2024 general election cycle, the African National Congress – which had presided over a state where unemployment soared to over 30% and jobs were not being created in any meaningful manner – somehow declared that, if South Africans renewed its governing mandate, it would somehow create 2.5 million jobs and build more houses over a five-year period.

The Democratic Alliance, whose dalliance with governance had been primarily in the Western Cape and, tangentially, in the metros where it has formed part of coalitions on an on-and-off basis since 2016, boldly trumpeted a jobs number of two million if South Africans gave it the national governing mandate.

Other parties, whose ability to execute is untested simply because they hadn’t actually governed anywhere, also came up with numbers that sounded as fanciful as those touted by the main parties.

In societies where the ambitious ideals of political manifestos play a crucial role in swaying public opinion, the availability and accessibility of credible, reliable data is essential.

Statistical agencies who are beholden to the public mandate are able to produce and disseminate information that is immune to the political headwinds of the day. The integrity of such information – and of the agencies that collect it – remains one of the fundamental guardrails of stable societies.

Information used to map out the country’s resource planning and track population trends needs to be able to resist the temptation of political influence, which would generally favour the political incumbents of the day if they sought to leverage data to amplify their preferred political message. Threats to agencies of this kind can be subtle and implicit – as seen in the case of South Africa – or explicit and overtly political, as has been the case in the United States.

Read: Are you employed if you work an hour a week? Stats SA says yes

Stats SA under scrutiny 

Over the past few weeks, such threats have played out in the public arena. In the South African context, the legitimacy of data published by Statistics South Africa (Stats SA) has been called into question by various CEOs, who allege that the current model of determining the country’s unemployment statistics is flawed, as it leaves a significant part of the economically active population unaccounted for.

In the words of former Capitec CEO Gerrie Fourie, the country’s unemployment rate which Stats SA calculates at 33%, is more likely to be in the range of 10%, and the discrepancy can be attributed to the inability of the agency to capture the true essence of economic activity in the country.

Read: Stats SA rebuts claims of understated unemployment

Since this remark was initially made, various viewpoints expressed by political and non-political actors have forced Stats SA to defend the integrity of its methodology and continuously explain what the ambit of its mandate is.

As one would expect, politicians who are close enough to the problem – particularly members of the government that has presided over this escalating crisis – immediately lapped onto this theory and expressed support for it.

During a parliamentary sitting in the aftermath of the challenge to the unemployment data, Brian Molefe, a Member of Parliament (MP) for the MK Party, claimed that even the 33% figure was an understatement and the unemployment rate was closer to 41%. Whilst this might have been a reference to the expanded rather than the narrow definition of unemployment, it was the response from ANC MPs that raised alarm. They then alleged that the crisis was being overplayed, as it was premised on data that ignores the informal sector.

Not to be outdone, the Minister of Employment and Labour Nomakhosazana Meth published a piece claiming that we all had to engage with the sentiments of the Capitec CEO. In an article titled “Capitec CEO isn’t wrong: Unemployment data needs work,” Meth lent political gravitas to the idea that Stats SA is underestimating employment, and given the significant public policy considerations that are contingent on such data, her words were not to be ignored.

The net effect of these developments is that data from Stats SA is now faces renewed scrutiny, based on the anecdotal views of key political and other role-players. If such trends become entrenched in the public discourse and the data published is second-guessed, it could provide a new arsenal for political campaigners who have an interest in minimising rather than amplifying the scale of the crisis.

Perennial underfunding

While these attacks on the agency have been quite pronounced, it is the subtle undermining of the agency by political role-players that should also worry us.

Five years ago, the council that governs Stats SA threatened to walk out in protest over the perennial underfunding of the agency. The underfunding had been going on for years and worsened in 2015 ,when R160 million was stripped from Stats SA’s budget. The risk highlighted by the council and Stats SA was simply that under-resourcing an institution that has the responsibility of collecting a wide range of data, in a country as vast and complicated as South Africa, simply meant that the integrity of the statistical system would be under threat.

In 2022, when Stats SA had to finalise the national census amid significant funding constraints, it produced results that were subject to unprecedented levels of scrutiny, with the reliability of the data called into question. In 2024, the Statistician-General Risenga Maluleke informed parliament that the funding gap at Stats SA was R295 million. In the absence of adequate funding, vacancy rates remained high and perceptions of compromised outputs were escalating.

The de-funding of such an institution means it does indeed operate closer to the brink of a credibility crisis – one that would be catastrophic for the country’s ability to plan ahead.

Listen/read: Risenga Maluleke’s insights on rising SA unemployment

While South Africa’s problems are manifest, it still exists in a much safer space than the US equivalent of Stats SA. The Bureau for Labour Statistics (BLS), which is tasked with publishing labour and employment data in the US, recently published data that seemed to be at odds with the political rhetoric of the Trump administration. In presenting data that reflected the current headwinds faced by the US labour market, the agency executed on its mandate to account to the public – only to find the political backlash swift and decisive.

US President Donald Trump – to whom much power has been vested and even more is unilaterally claimed – simply decided to dismiss the head of the agency, Erika McEntarfer, claiming the BLS had rigged jobs figures to make him and the Republican Party look bad. Since then, he has appointed a new head who presumably will somehow strike the right balance between statistical rigour and the more mythical mandate to avoid embarrassing the president and his party.

The danger with such a turn of events, is that much of the decision-making across the world hinges on the winds of change in the US labour market and the economy at large. The idea that its statistics are now subject to political influence, from a president whose commitment to statistical rigour is tangential at best, makes the process of relying on the data points tricky for key decision-makers.

Read: Beating unemployment: Can SA reform fast enough?

This undermining of the statistics agency – and seeking to repurpose it to serve the political agenda of the day – is precisely the type of crisis that the UN’s 2014 endorsement  sought to pre-empt by reminding us that such data must resist political headwinds by being reliable and objective. But when one considers Trump’s stance on the UN and the essence of what it represents, it’s not surprising that he would upend that playbook and seek to define what data the public should see.

So, while there is plenty to lament about the South African situation, at least our politicians have not vaulted into the explicit dismantling of the integrity of national statistical data.

Rather – and depressingly – they are simply incapable of engaging with the data and using it as a springboard for difficult decision-making. Since the release of the latest data on unemployment, which emphasised the depth of the crisis, political leaders have been defined by their absence from the conversation, and it feels like they simply live in hope that the numbers will one day fix themselves.

As tragic as it sounds, this passive approach it is still preferable to actively manufacturing their own data sets, like Trump seems committed to doing.

Follow Moneyweb’s in-depth finance and business news on WhatsApp here.