Hon Chairperson, hon members, ladies and gentlemen, today Parliament debates the Vote of Statistics SA as a stand-alone Vote for the first time since the establishment of democracy in this country.
For the past 16 years Statistics SA has always been part of something else. Initially it was part of the RDP Vote and was later incorporated into issues raised by the Minister of Finance. When the Ministry of Finance came to Parliament or to an Extended Public Committee, to debate the Treasury; the South African Revenue Service, Sars; the Development Bank of South Africa; the oversight of the Reserve Bank; the Public Investment Commission; and the Financial Intelligence Centre, three lines would be committed to Statistics SA. Because of the way in which this matter was dealt with, there have been unintended consequences.
The fact that this is being debated as a stand-alone Vote for the first time today is a cause for celebration, so I say "Cheers!", but it will have to be a virtual celebration because of the fiscal constraints and also because of what we've indicated the workload of Statistics SA will be over the next period.
In the context of the ash cloud generated by volcano Eyjafjallajkull in Iceland, flying and flight safety are exceedingly popular topics. Shall we fly? Can we fly? Where to? Why? The question is: How much information does a pilot need on the environment that he or she is flying into and, more importantly, about the response of the aircraft that they are flying?
Think about the important question that nobody has quite turned to of who generates the information that pilots need. And what are the consequences of such information that would indicate to a pilot the altitude, speed, wind direction, etc? What are the consequences of such information being incorrect? Who checks that the instrument gauges are accurately calibrated?
In truth, the work of a statistical agency is no different to that of the avionic engineers who provide pilots with certainty. It would indeed be strange if those who travel to Johannesburg on a weekly basis regularly ended up in places like Windhoek or George - although our luggage ends up there sometimes. We want the safety of knowing that the pilots have accurate and well-calibrated gauges when they fly.
For the statistical agency the client base is different: not pilots, but policy-making and not aviation safety authorities, but Parliament. And the information must inform decisions, primarily about the economy and people's lives. What is consistent is that the information must be readily available, usable and indeed reliable.
The opportunity presented by this debate is, in many respects, the start of Parliament's engagement with the statistical outputs in order that Parliament will know that these outputs meet the quality standards because this then becomes the heart of oversight, not merely of Statistics SA, but of all of the government's functions.
Parliamentarians need to know about the number of houses being built, learners in schools, visits to public health facilities and jobs created. These are very important parts of information that Parliament must know; so the quality of statistics becomes vital to answering those issues, not just in general but as part of parliamentary oversight.
It's accepted across the world that official statistics are used to inform Parliament and political assemblies about the state of the nation and provide a window on the work and performance of government.
Firstly, this allows these political assemblies to assess the impact of their policies over time and between different areas.
Secondly, it provides Ministers with a picture of the economy and society, enabling them to formulate economic and social policies and to monitor and evaluate their delivery.
Thirdly, statistics permit government and its agencies, at all levels, to carry out their business efficiently and effectively and to make informed decisions based on evidence.
Fourthly, it provides citizens with a view of society and of the work and performance of government. Statistics show the scale of government activity in every area of public policy, allowing the impact of government policies and actions to be assessed.
Fifthly, it furnishes businesses with statistical services that promotes the effective and efficient functioning of industry and commerce and assists analysts, researchers, scholars and students with their work and studies.
It's important that we acknowledge the challenge of the attitudes faced by Statistics SA in the process of compiling the data. The enemy being faced is the cavalier behaviour frequently of persons surveyed and public servants who are tasked with providing correct and current information. If people tell untruths in response to questions about their circumstances, Statistics SA will generate statistical series that are quite unreliable. Unfortunately, for the people providing wrong information, there are no consequences.
And then, there are public servants who provide incorrect information - either as a consequence of omission or commission. But today I want to invite Parliament to assist with remedying this unhealthy situation.
Firstly, we must all work together to ensure that all citizens appreciate that each of us has a civic duty, which includes providing accurate information when requested. This becomes a task for all organs of society, elected representatives in all spheres and all parties, religious leaders, trade unions and educational institutions together.
If we cannot measure either what has been done or what remains to be done, we cannot allocate resources or plan improvements. If the appeal to civic duty fails to yield the correct responses, we will have to come back to Parliament to explore with Parliament the means to insist on compliance.
As part of its mandate, Statistics SA has to collect information from both households and businesses. Often, it has to ask about similar issues and compile a view on the same issue from the perspective of business and another view from the perspective of households.
An example of this is the measurement of the labour market, which measures both the number of jobs in the form of the Quarterly Employment Survey and the level of employment in the form of the Quarterly Labour Force Survey. These are two closely related concepts that are measured using two different sets of instruments and directed at two different audiences. Now, those of us who are here as nonstatisticians battle with these things.
One of our deputy directors-general last week in the discussion said that there are statisticians and then there are the rest of you, the rest of us, the rest of the people, who battle with these fine distinctions between things like the QES and the QLFS.
The measurement of jobs is directed at businesses and asks about the number of persons they employ or the number of job opportunities that exist and are occupied. To measure employment levels, data is gathered directly through household surveys. This information not only gives an indication of the number of employed persons as in the QES, but it also better measures the rate of unemployment.
What should become clear is that the different surveys must correlate with other information. But sometimes we see that the QES and the QLFS actually throw up results that are quite different. It's quite important that we understand how these things work.
An area that will improve the compilation of our economic statistics and enhance better service delivery is the business registration reform. In committing to speed up the country's economic growth and transforming the economy to create decent work and sustainable livelihoods, President Zuma identified in his state of the nation address, on 3 June 2009, the reform of the business registration system as a key priority when he said and, I quote:
In another intervention to create an enabling environment for investments, government will move towards a single integrated business registration system. This will improve customer service and reduce the cost of doing business in South Africa.
Statistics SA, Sars and the Department of Trade and Industry have set up a joint project and will work together to achieve this government object. The first appointments to the project team were made at the beginning of this financial year. This project will not succeed without the fullest collaboration of all the government partners, including the Companies and Intellectual Property Registration Office, which must maintain a current and reliable business register.
If you don't have a business register, you can't sample, if you can't sample; accurately, you can't expect anything that approximates trends in the labour market.
While it is estimated that the project will take three years to complete, Parliament will have to monitor the delivery by political and administrative heads of these institutions to ensure that the targets are indeed met.
To get a measure of both quantitative and qualitative public services, government and Parliament alike need dependable statistics. We know what funds are allocated, after all. Parliament votes these funds in the Appropriation Act. Part of the process in discussing different Budget Votes is dealing with this in some detail over a fairly lengthy period of time.
The independent audit system provides, or ought to provide information about the general utilisation of these financial resources. But we do not yet know enough about how the funds are disbursed nor do we have an adequate measure of the actual services that the financial resources pay for. We require sound administrative record-based sources for the data to form the basis of the national statistics system. The system should begin to provide a complete picture of government delivery and in the process identify gaps.
If the education statistics point to a low percentage of matriculants with maths and science passes, government may want to consider providing incentives for specialised training for teachers in these subjects. Colleges may need to consider providing bridging courses.
If the statistics from a specific health clinic point to a preponderance of a particular disease, we may want to investigate the environmental factors that cause this particular predisposition.
However, if we want this, then every cost centre must generate data that will inform an information management system within the function of the course of its work. So, your education management information must tell you how many learners there are on any given day. It must be able to indicate how many educators are there. We must know the state of the infrastructure where learning and teaching takes place.
It's also important that we know how many books are purchased, when the orders are placed and when they are delivered. This kind of information becomes vital to knowing the quality of an education system. You don't just wait until about 5 or 6 of January when you have matric results. You must know these things in the course of what happens.
Similarly, I think in the health system we need to know how many visits there are, what people are visiting health facilities for, if they are issued with pharmaceuticals by dispensaries and what they are issued for. You must be able to verify these kinds of things. If we are spending the amount of money on ARVs that this country does, then we must know which patients are actually receiving them.
If you can't verify these kinds of things, then all of Parliament and all of the allocations that we make through the Appropriation Act must have a serious question mark hanging over them. But collecting this information must be done in the course of what health workers, educators and others do in the system, and the national statistical system must then bring all of these together.
This is a challenge that confronts us. We must become much better at delivering democracy and we will not be able to attain this without a competent and trusted statistical system. This information also becomes important for a statistical system that informs evidence-based policy- making to drive future changes.
This brings me back to the point made earlier with regard to the role of public servants and the provision of accurate information. The responsibility for accurate administrative records should not only be a civic duty, but should also be linked to performance measurement of officials. We must accept then that in some cases there are challenges in this regard in terms of the administrative burden placed on some already overworked service providers.
At the same time, however, we must also acknowledge that public servants need to understand that it's imperative that accurate information is provided. Statistics SA is only able to compile an accurate reflection of services if the data provided to them is accurate.
Once again, we appeal to the Members of Parliament to take this matter seriously when performing oversight and holding government to account. It's important that Parliament and the public begin to understand the strategic role of statistics and the importance of evidence-based decision-making. In terms of the Statistics Act, the role of the Minister is limited to determining what is measured. How it's measured is left to the Statistician- General and his team.
The role of the SA Statistics Council is to provide us, the public and Parliament alike, with the assurance that the methods are scientific and comply with national and international standards such as the United Nations Fundamental Principles for Official Statistics; the Special Data Dissemination Standards, agreed to in the framework of the International Monetary Funds; and the eight dimensions of quality specified in the South African Statistical Quality Framework.
The consumer price index is an exceedingly important measure of both wellbeing and change. It is also the most widely used of the statistics series. It is used by the Monetary Policy Committee of the Reserve Bank in their setting of the repurchase rate, which is a key determinant of interest rates. It is also used as a key measure in wage bargaining.
It is therefore essential that the CPI approximates as closely as possible the impact of price changes on the living standards of families. In fact, it's for that reason that the CPI was quite significantly retooled just a few years ago and the process completed last year.
Amongst the changes introduced, there was a change to the basket of goods measured - what do we eat, what do we spend our wages on? Previous measures excluded things like cellphones, the Lotto and so on. These now needed to be brought into the basket. There was a simultaneous change to the weights of the various goods and services in the basket.
To compile this, the income and expenditure survey was instituted. It asked 30 000 families to record an income and expenditure diary over a period of a month. Once the baseline was changed, the updates were then captured in a modern form. This innovation puts our CPI on a par with the best in the world.
It is envisaged that retooling is expected to be done more frequently, at intervals of three years, as products in the basket change and as people experience greater social and geographic mobility. Then a more intricate process of calculation of products by amounts spent and the number of people consuming the products generates the basket for the CPI. Against this base, prices on selected goods are collected monthly from outlets and price changes are recorded to calculate a monthly CPI figure.
The related price index is the producer price index, PPI, which to date is not as robust in its product coverage. In Australia, for instance, the PPI is treated as a lead indicator to the extent that it acts as a harbinger for indicators downstream.
The PPI in South Africa is in the process of being retooled so that we too can begin to see the linkages between factory gate prices and store prices. Another product to be presented before Cabinet and Parliament soon is the Poverty Line. The measurement is designed to produce statistical information on poverty and inequality from a multi-dimensional perspective. And we will do so at regular intervals. Again, relevant, good-quality data is a key element of the Integrated Antipoverty and Inequality Framework.
The poverty line is an analytical tool using multiple sources of information such as the consumption expenditure basket; the income and expenditure survey, IES; and information from dietary protocols provided by the Medical Research Council, amongst other sources. This provides an indication of a food line across the different income brackets, which is then used to determine a representative food basket and a per capita food requirement. Basic requirements such as clothing, shelter and transport are added to provide a poverty line. It is envisaged that the line will be updated annually by the CPI.
Having noted the strategic importance of statistics earlier, it's therefore vitally important that the communicators of information are here - see how full they've made the gallery today, all of these wonderful journalists we have, who have an interest in our lives and who are part of the Parliamentary Gallery Association. They are here in their numbers and the place is crowded. It's probably because we are so "snoep" [stingy] with the food or something! [Laughter.] If our journalists are not up to date and don't understand the value of their information, ordinary South Africans will never appreciate this work. Then there's no measure and everything is left to guess work. You may as well treat the statistical outputs in the same way as you treat, I don't know what. When do you play the Lotto - Wednesday, Thursday? Whatever day. You treat it like playing the Lotto. You just click on numbers and, clearly, it's a poor indicator of everything.
Now, what Statistics SA has done, and this is part of the return they are getting, is that they've started a series of workshops, seemingly well attended; seemingly the journalists were fed on that occasion - to begin to create areas of expertise in the news so that at least a basis is being provided for statistics going forward. [Interjections.]