Hon Chairperson, in his speech earlier, the Minister flew a kite on the political nature of statistics. In that regard, I came across this from a British judge, who was obviously much more cynical than I am, and many years ago had the following to say about statistics:
The government are very keen on amassing statistics. They collect them, add them, raise them to the nth power, take the cube root and prepare wonderful diagrams. But you must never forget that every one of these figures comes in the first instance from the chowky dar (village watchman in India), who just puts down what he damn pleases.
Thankfully, these days, I understand Statistics SA depends less on village watchmen and more on personalised surveys to provide our national statistics.
And because these statistics allow us to assess the needs of our people, underpin our policy-making and inform the division of revenue, it is difficult to overstate how important it is to accurately measure key statistics like population, GDP, inflation and jobs.
Significant errors by Statistics SA since 2011 have, at times, threatened to undermine public trust in official statistics. The 2001 census had a 17% undercount, and this was followed by the massive overestimation of a decline in manufacturing in 2005. The DA welcomes the fact that such errors have not recurred in recent years, although some valid questions are now being asked about methodology, which I will get to later.
First, let me turn to the issue of the Statistics SA budget. The one figure that is not clearly defined is the estimated final cost of the 2011 survey, so I ask the Minister to please clarify the figure for me. The DA gave its full support to the census and eagerly waits for the results to be announced.
We appreciate that the census required a massive increase in head count at Statistics SA and we also note the 28% increase in the line item for the compensation of employees in 2011. It is concerning, however, that this item declines by only 4% in 2012 and then increases marginally next year and the year after.
Surely, once the 150 000 field workers have completed their temporary work, there should be a related decline in the total compensation of employees? Instead, what we see is a marginal decline in the wage bill from R1,25 billion to R1,2 billion in 2012. By contrast, total expenditure this year declines by 53%, appropriate, I think, for a year following a national census. But it does raise the question of why employee compensation is only declining by 4%. I therefore ask the Minister: Where is that money going with no census on the go?
Furthermore, in their 2012 strategic plan, Statistics SA makes a request for above-budget increases of 12%, 23% and another 23% over the medium term. Its plan does not provide detail on why these increases are justified, and the engagement with the Statistician-General in the committee did not result in full satisfactory explanations. Perhaps the Minister could shed some light on why these increases are needed above what has been budgeted for by Treasury.
Turning now to methodology, we welcome the effort to rebase and re-engineer the consumer price index, CPI. We are also pleased with Statistics SA's decision to construct separate inflation indices for income groups. But we call for these findings to be more prominently published on the Statistics SA website and in press releases.
We do this because these indices show that poor South Africans have faced the highest rates of inflation for 2011. In our party's own desktop research into price increases of products bought by poor South Africans, that research shows that these goods may have increased in cost by as much as 16% in just the first three months of 2012.
This reality has real implications for policy-makers in South Africa who need to have a better understanding of price increases imposed on those least able to afford them. We believe Statistics SA has an important role to play in getting this information out there. Through you, Chair: Mr Minister, will you join me in asking the department to do this?
The DA would also like to see Statistics SA change its attitude to criticism from labour market analysts. The ongoing spat between the Statistician-General and Adcorp about the measurement of the informal sector and migration is unseemly. It lowers the tone of what could be a constructive engagement on the measurements of key variables.
In recent years the Statistician-General has labelled Adcorp's research as "ad hoc rubbish", "absolute nonsense" and "metaphorical hallucinations of an intoxicated institution". This is the wrong approach from a key government institution to a debate that could improve our understanding of the structure of the South African economy.
It is obvious to everyone that Statistics SA's measurements of the informal sector and migration are not perfect, because, I think, these variables are relatively difficult to measure. But it also seems clear to most that some improvements need to be made.
Since 1995 the gap between GDP growth and formal employment growth has widened. Statistics SA attributes this to increased unemployment, while Adcorp claims there has been an increase in the size of the informal sector, which they estimate is three times larger than Statistics SA's official numbers indicate.
Adcorp has also identified significant undercounting of key statistics across the board, averaging out at around a 30% undercount for everything from the number of taxpayers and businesses to the number of unemployed and those receiving social grants. They calculate these divergences by using other government sources like the SA Revenue Service, Sars, the Unemployment Insurance Fund, UIF, and the bargaining councils and comparing them to what Statistics SA is finding.
It seems reasonable to us that, rather than lashing out at these findings, Statistics SA should investigate whether it makes sense for them to weigh data from their surveys against other information on economic activity, such as tax returns, the amount of money in circulation, studies of the informal sector, the size of the small business sector, and estimates of unrecorded economic activity including that of illegal immigrants.
Regarding migration, it seems that the questions asked in last year's census might not generate sufficient data to assist us to better understand the true extent of immigration into South Africa, because they were based, as I understand this, on language preferences. That is all very well and we can understand why the survey was designed like that, but we feel this obliges Statistics SA to undertake a specific and extensive study, perhaps in collaboration with Home Affairs and even private-sector players, like the Business Trust or Adcorp, to get a better understanding of how many immigrants there actually are in South Africa, legal and otherwise. Through you, Chair, does the Minister agree?