Date: 14/08/2009

Source: South African Reserve Bank

Title: SARB: Mnyande: Welcoming remarks by the chief economist of the SARB at the SARB-Irving Fisher Committee seminar

Governor Nielson (of the National Bank of Denmark)
Irving Fisher Committee Chairperson, Dr Marfán
Members of the Irving Fisher Committee Executive
The Irving Fisher Committee Secretariat and distinguished central bankers.

The South African Reserve Bank wishes to welcome the delegates assembled here to the South African Reserve Bank-Irving Fisher Committee (IFC) Seminar on "Economic and Financial Convergence en Route to Economic Integration: Experience, Prospects and Statistical Issues".

The Bank is honoured to co-host this seminar with the IFC. Durban is one of South Africa's most beautiful cities. The weather is good throughout the year, and the Durban beach along the coastline of the Indian Ocean - just like the people of this province - is warm and inviting. The hospitality of the people of South Africa is exceptional and you will see this for yourselves in the next few days.

Central banks and statistical challenges

Modern central banks are typically tasked with the formulation and implementation of monetary policy with a view to delivering price stability and with the maintenance of financial stability. These very critical macroeconomic functions require that relevant, reliable and quality statistics be available to enable central bank authorities to execute their functions properly.

Central banks across the world face numerous statistical challenges and the South African Reserve Bank is no exception in this regard. For example, the compilation of the balance-of-payments statistics - for which the Bank is responsible - is frequently under scrutiny due to the size of the number of unrecorded transactions. The Bank's experience of compiling the balance-of-payments statistics is that the exercise itself is not as simple and straightforward as often argued. Inconsistencies arising from estimating aggregates from various sources, the use of sometimes incomplete information and various reporting errors lead to this item. But, very importantly, in the case of errors and omissions in the balance-of-payments statistics, the size of the discrepancy does not necessarily provide an indication of the accuracy of the balance-of-payments statement. Furthermore, the fact that the unrecorded transactions in some instances reported together with the financial account does not mean that the unidentified counter-leg of a specific balance-of-payment transaction is necessarily part of the financial account; in fact, it could just as well be part of the trade or services account.

The accuracy of official balance-of-payments statistics, we have observed, relies heavily on the quality of data provided by external sources, of which the most important in our case are: the South African Revenue Service, which provides trade statistics; the formal exchanges, providing information on transactions in equity and debt securities; the South African banking and insurance sectors; and other private-sector non-bank companies surveyed by the Bank. Inaccurate and/or unreliable information released unintentionally by any of these entities will eventually be reflected in this item referred to as "unrecorded transactions".

Apart from the use of incomplete information that may have an effect on the classification, timing, coverage and valuation of transactions between transacting parties, it should also be mentioned that fluctuating exchange rates could also materially contribute to the size of the unrecorded transactions. Despite all these challenges, the Bank has worked tirelessly to minimise statistical discrepancies by always striving to improve its data-collection process.

The Bank's view is that its task should be to communicate the challenges it faces in this area effectively. It should constantly advise the public on the facts that, for example, complete surveys of certain sectors often only become available after a significant time delay, and that substantive revisions are usually published once the revised historical data have been thoroughly analysed. Time series are, therefore, revised with the utmost circumspection and mindful of their importance for analysis and for policy formulation. Revisions are not an admission of guilt, but a recognition that better, more comprehensive data become available after some time has elapsed. Our attempt to construct good and reliable statistics is but part of what the Bank does in the dissemination of the country's external-sector data.

Statistical improvements at the South African Reserve Bank

The Bank endeavours to improve the quality of the statistics it releases on an ongoing basis. Like all central banks represented here, it continuously tries to ensure that the concepts, definitions and standards set internationally as published in manuals such as the Systems of national accounts, the Balance-of-payments manual, Monetary and financial statistics manual and Government finance statistics manual are relevant, understood by compilers and internally coherent.

The framework used by the Bank in compiling its statistics is based on the recommendations made in these manuals. In 1996 South Africa subscribed to the International Monetary Fund's Special Data Dissemination Standard (SDDS), and the Bank, in co-operation with statistical agencies in South Africa, has managed to meet the rigorous requirements of this standard on an ongoing basis.

The Research Department of the South African Reserve Bank has recently launched a Data Quality Working Group to review the quality of statistics obtained from respondents systematically. To enhance the quality of the economic statistics the Bank thought it pertinent to identify and prioritise the focus areas related to statistical challenges in South Africa. The key strategic challenges are the mandates to collect and disseminate primary data on the expenditure side of the economy, balance of payments and financial intermediaries.

This is a huge undertaking that looks at a broad spectrum of issues, including the survey forms used to request data from respondents, international manuals, domestic requirements, the manner in which the data are collected, issues of periodicity and frequency. This is not a short-term project. The Bank felt that it was important to look at statistical requirements in a holistic manner and to take not just operational but also strategic steps to ensure that policy is informed by sound statistics.

The required outcome of the working group's activities is timely and relevant data capturing; essential dimensions of economic activity that is of verifiable quality, measured consistently and satisfies both the needs of the Bank and its clients. Here the Bank is of the view that statistical methodologies have to take cognisance of international best practice. One such international best practice forum, to gain insight in the application of recommended concepts and learn from others' experience, is the IFC

Benefits of being an Irvin Fischer Committee member

Since joining the IFC as an institutional member on 23 August 2006, the Bank and I personally, as a member of the Executive, have had an enriching experience. As a member of the committee, the Bank has been able to keep abreast of international best practice in the compilation of statistics.

Staff members of the Bank have benefited immensely from seminars, conferences and events organised by the IFC. As a member of the committee, the Bank has been able to gain from the insights of other central bankers in the areas of collecting, analysing and presenting statistics.

Recently, the Bank for instance participated in the creation of a handbook on debt securities statistics (the prepublication draft dated May 2009 is available co-ordinated by the Bank for International Settlements, the European Central Bank and the International Monetary Fund. This was a rewarding experience for the staff members of the South African Reserve Bank as they interacted with other colleagues from other central banks. It has also necessitated that the Bank makes a deeper analysis of the domestic securities statistics. The handbook will assist by providing a conceptual framework for the compilation and presentation of debt securities statistics globally.

Similarly, with the reweighting and rebasing of the South African consumer price index (CPI) in February this year, several changes were made. These included the adoption of the Classification of Individual Consumption by Purpose (COICOP) system for CPI compilation. Other refinements, including the introduction of owner's equivalent rent to cost owner-occupied housing, were also introduced. These changes had an impact on inflation targeting framework, and the Bank had to study the consequences carefully and how the implementation of monetary policy would be affected. Contact with other IFC members was of great value in doing so.

It is for these reasons that the Bank strongly encourages central banks that are not members of the IFC community to reconsider their positions. As members, they are presented with an ideal opportunity to strengthen and expand their statistical capacity by way of cross-pollination of ideas and processes. Meetings and seminars hosted by the IFC provide ideal platforms for sharing views and experiences, and discussing future prospects in central banking statistics. The South African Reserve Bank will, therefore, continue its active participation in IFC activities.

Contents of the seminar

Briefly on the content and theme of the seminar: Regional economic integration has never been a smooth process; we have observed that some countries benefit more than others in such a process. However, commitment and drive towards regional economic integration persist. Various economic integration initiatives are currently in progress throughout the world. It would seem that some form of convergence of selected economic indicators usually precedes the "deeper" forms of integration. But this can only be sensible if there is a reasonable degree of harmonisation of statistical standards.

From a statistical point of view, the process of economic integration poses many challenges. One such challenge is ensuring that cross-country statistical comparisons are valid, because economic indicators are often compiled differently across countries. Authorities need to be cautious in their interpretation of cross-country statistical comparisons.

This seminar aims to provide an analysis of international economic and financial convergence experience, which may be essential for assessing progress and obstacles in the way of achieving economically integrated states here in the Southern African Development Community (SADC) region and in the rest of Africa. Emphasis is put on the statistical dimensions of this experience. The strains which develop when shocks and turmoil hit the participating countries in an integration arrangement are also touched upon, along with the question whether convergence criteria should be treated with some flexibility under extraordinary circumstances.

The first session provides an overview of initiatives to promote convergence in the context of regional integration. The second session delves deeper into the prospects and statistical issues relating to convergence in SADC and the broader continental African economic integration process, and the third addresses issues of definition, measurement statistical improvement as far as key convergence variables are concerned.

The seminar should, therefore, provide a broad range of discussion on these themes. The diverse group of presenters and discussants all experts in their various field, should ensure that we all benefit from the discussions.

Thank you and all the best in the next few days.