Chapter 4: The materials and tools of subsidy analysis

Basic data
 
Ultimately, all subsidy analysis depends on data, and most of these data are collected and provided (not necessarily published) by governments. The usual primary source for expenditure data is government financial statements. Some government departments also helpfully prepare summary tables on expenditure under programmes for which they are responsible.

Another source of information is national accounts. While the data in national accounts capture only a subset of budgetary subsidies (no tax expenditures), the background reports for them can be enlightening. Canada, for example, as part of its annual national accounts exercise, publishes the names of companies or individuals receiving C$ 100,000 or more under a particular programme in a given year.
 
But many forms of subsidies, particularly tax breaks and credit sub-sidies, do not make it into the official accounts. According to the experts at the World Bank, only about a dozen countries regularly report estimates of tax expenditures. The U.S. Government publishes two sets of estimates of the tax expenditure of federal tax breaks, but only for tax expenditures worth US$ 50 million or more a year.

Information on subsidies at more local levels of government, which can be crucial in influencing investments in plants and buildings, is much harder to find, in part because the packages of incentives are unique to each recipient. In the United States, a few States (e.g., Illinois, North Carolina and Minnesota) now make some information on corporate subsidies available on the Web. The European Commission requires that its Member States notify new "state aid" programmes. This information also can now be accessed via the Internet.

Non-governmental organizations and journalists have been successful in some countries in extracting subsidy data from governments that previously had not been made public. So-called "Freedom of Information" laws have been critical in this regard.
 
 
Sectoral subsidy accounts
 
The bulk of what the world knows about the amounts and types of subsidies provided to specific sectors and sub-sectors comes mainly from intergovernmental organizations, such as the OECD.

The values of subsidies, or of support more widely, provided by OECD countries derive mostly from the OECD itself, or in the case of coal (until 2001) from the OECD's sister organization, the International Energy Agency (IEA). Estimates for the rest of the world come mainly from various one-off efforts by analysts working for the World Bank, the International Monetary Fund or one of the United Nations agencies, the IEA and from a few dedicated individuals.
 
As a result, when researchers combine the aggregate estimates from the sectoral accounts into global estimates of subsidies they are combining numbers based on different starting assumptions, different estimation methods, different policy coverage and even different time periods. Some normalization could be attained with a more careful approach to aggregation than some have used in past studies, although the level of comparability between these accounts is nowhere near that for corporate financial accounting.
 
 
Large computerized models and their ilk
 
In recent years, sophisticated economic tools have been brought into service to help understand the effects of subsidies, particularly agricultural subsidies, on trade and welfare at the global level and within individual countries or groups of countries. Most of the large-scale efforts to date - by the World Bank, the OECD, the Institut national de la recherche agronomique (IINRA), the Carnegie Endowment, and a few independent analysts - have involved the use of computerized general equilibrium (CGE) models.
 
Although the specification of these models differ, they all share the same source of their information on subsidies: the database of the GTAP (Global Trade Analysis Project) consortium. The GTAP database, in turn, draws on data generated by the OECD (relating to its own member countries' farm support) and by WTO Members in their notifications of domestic support.

An important caveat of any recent CGE-based analyses that purport to examine the effects of "subsidy reform" on trade and welfare is that they consider only a sub-set of subsidies, namely subsidies to primary agriculture. Because the databases used for these models do not contain information on subsidies to energy, manufacturing, transport or even fisheries, the effects of such subsidies, or their reform, are not analyzed.
 
Most of the effects of trade reform that these models measure, therefore, are driven by changes in border measures, namely tariffs and tariff-rate quotas. When it is reported that only a small percentage of the benefits of multilateral trade liberalization generally, or from a specific trade deal, would stem from the elimination or reduction of subsidies, it is vital to understand that the outputs of the models largely reflect their inputs, at least as regards subsidies.