II Importance of Data
- Marcello Caiola
- Published Date:
- August 1995
Before preparing a financial program or analyzing developments in a country, sufficient data on the economy’s past and expected performance are necessary. If the basis for the program or analysis is deficient, the program will probably not be well prepared and the objectives will not be achieved. All those data that may have an impact on the preparation of a program are needed. IMF missions are sometimes described as “hungry” for both statistics and data in general, but this hunger is logical, because before giving any advice or taking any decision, the mission must know what is happening in a country. The term “data” is used broadly here, meaning not only statistical information, but also details of pertinent legislation and of rules and regulations connected with the implementation of such legislation.
For example, in preparing a tax package to be included in a stabilization program, a mission requires reliable information to allow it to estimate the yield of the proposed tax, including details of the proposed legislation and the pertinent regulations. For instance, the actual yield from a new tax could be less than expected during its first year, if the tax is collected through agencies and/or banks that are authorized to retain the collected taxes for some time before depositing them into the treasury accounts. Similarly, in some countries, taxes on domestic sales of petroleum products are collected by oil companies, which may retain these funds for some time. Also, in a country that suffers an accelerating rate of inflation and that levies income taxes on the income generated in the previous year, the yield from income taxes is bound to decline in relation to GDP, even with satisfactory tax administration machinery.
Another example of discrepancies between estimated and actual tax collections may occur whenever the estimate has been based on insufficient information. For instance, in a given country the projected yield from an export tax may not take into account the fact that exporters are allowed to deduct sales taxes and other payments from their export tax liabilities. As a result, the actual yield of the export taxes would be well below the original projection.
If a program were prepared without taking into account the details mentioned above, actual revenues would be likely to fall short of the budgeted amount, and the expected overall fiscal result would not materialize.
Similarly, to prepare a good monetary program, reliable monetary accounts, with adequate details on individual accounts and legislation, are necessary; otherwise, the measures proposed may fall short of expectations. For example, in the case of legal reserve requirements, data on both legal and actual requirements are necessary, as well as information on the enforcement of these requirements. Thus, if the legal reserve requirements in a country are equivalent to 20 percent of the banks’ deposits, but actual reserves are equivalent to 30 percent of the deposits, any increase in the legal requirements up to 30 percent would sterilize little additional liquidity, since the banks already maintain excess reserves. State banks, which in several countries do not fulfill their legal reserve requirements, provide another example. Any increase in the legal requirements for those state banks would have little or no impact, because the state banks would show only larger reserve deficiencies. Also, if the legal obligations are calculated as average holdings during a month, banks could borrow from each other, thus technically fulfilling the requirement even though their reserves are lower than specified by the law.
But the information required by a mission is more than purely economic. After all, people are not numbers and they react differently, particularly to political developments. Thus, in estimating the yield of certain measures, one should be aware of circumstances peculiar to the country that may affect, positively or adversely, the implementation of those measures. For example, an increase in income taxes in a country suffering a civil war or in a country where there is widespread tax evasion is not likely to yield revenues equivalent to the theoretical calculation of the new taxes. Similarly, sharp increases in import taxes in a country with large contraband activity may only result in more evasion, and consequently in lower receipts.
The conclusion then is that if the data at one’s disposal are deficient, they will, at best, make programming more difficult and, at worst, they will result in the wrong advice or an over-estimation of the impact of the proposed measures.
Besides being comprehensive, data should be organized and presented in such a way as to facilitate analysis. Coverage of data may vary from country to country, and data presentation by the authorities of a given country may differ from “legal” or “conventional” formats. For instance, in some countries the operations of the central government are sufficient to give a satisfactory picture of the fiscal situation; whereas in others, data for the entire nonfinancial public sector may be required. Also, in some countries, financial developments are reflected in the operations of the banking system, because nonbank financial intermediaries are very small and have only limited operations; whereas in other countries, the nonbanking financial system is active and, therefore, data on the banking system alone would not be sufficient to give a full picture of the financial intermediation. Incidentally, the importance of the nonbank financial intermediaries may be enhanced if the central bank has little standing authority and/or control over the operation of these intermediaries. Thus, coverage of the data must be tailored to each individual country.
Regarding the presentation of data, a presentation that could be correct or required from a legal, accounting, or even economic point of view, may not be the most useful for the purpose of analysis. For instance, in most countries the government budget document includes among the revenues the accumulated surplus or deficit of previous fiscal years, as well as any form of gross borrowing. Similarly, a bank presents its accounts in the form of a balance sheet that would classify accounts as assets and liabilities. But, to be analyzed, these data would have to be reorganized into completely different formats. More generally, data must be presented on the basis of what we want to show. In the case of a program, one has to quantify the resource gap and its financing.
However, the quality of data in many developing countries is poor and may change drastically over time, so that statistics are not always comparable or comprehensive. Also, there may be exceptional factors that influence trends over time or create breaks in these trends. As a result, extreme caution should be used in drawing conclusions from certain past trends or relationships, particularly because in recent years trends have been distorted by exceptional factors, such as renegotiation of debts and the reluctance of banks and the private sector to invest in developing countries.