Share This
March 2009 Cover Art

Search Finance & Development

Advanced Search
View this Issue

Archive of F&D Issues

Subscribe to Print Edition

United States and the IMF United States and the IMF

Write to us

F&D welcomes comments and brief letters, a selection of which are posted under Letters to the Editor. Letters may be edited. Please send your letters to:

Free Email Notification

Receive emails when we post new items of interest to you.

Subscribe or Modify your profile

Finance & Development
A quarterly magazine of the IMF
March 2009, Volume 46, Number 1

When Governments Intervene
James L. Rowe

New IMF reporting standards provide a better gauge of the costs of government financial rescues

The IMF has developed a new approach to reporting on government finance statistics that nearly everyone agrees is more comprehensive, more transparent, and a better measure of a government’s financial position. It’s also a more useful policy tool for IMF surveillance, part of the Fund’s mandate to oversee the health of the international financial system (see “Data to the Rescue,” F&D, March 2009).

The new approach, unveiled seven years ago, substantially augments the traditional government accounting approach of recording transactions only when cash changes hands. It records transactions when economic value is created, transferred, or extinguished (often well before any money is expended), and takes account of non-cash transactions. Moreover, it links transactions (flows) to the stock of a government’s assets and liabilities (both financial and non-financial), which makes it possible to determine the change in a government’s net worth for a given accounting period. That in turn gives clues as to how sustainable a government’s policies are.

The IMF’s Statistics Department and Fiscal Affairs Department, the main forces behind the accounting framework embodied in the 2001 Government Finance Statistics Manual (GFSM 2001), have completed studies that show that recasting fiscal data from the national presentations to the new analytical framework is feasible and can be done in a reasonable amount of time without significant cost to either the IMF or the countries. Full implementation of the new methodology, involving improvements to data sources and the adoption of accrual accounting systems, would require some resource commitments on the part of countries.

Gauging intervention costs

But the massive government interventions to offset the growing unemployment and falling production during the current global economic crisis have made the GFSM 2001 framework more pertinent. The approach will provide a better gauge of the ultimate costs and potential financial effects of those interventions—such as the U.S. government’s $700 billion plan to shore up the financial sector, the automobile industry, and its de facto takeover of the government-sponsored housing finance enterprises Fannie Mae and Freddie Mac, according to Keith Dublin, Advisor, Statistics Department, who has worked closely with the staff of the Government Division (GFD) on these issues (see box).

Issues in intervention reporting

There are a number of key issues involved in properly reporting interventions. Among them:

  • Valuation. In some cases the current market value of assets may be hard to determine. But the value is crucial in defining the nature of the intervention. If, for example, the government buys troubled assets at market value, then the operation is an exchange of assets (providing liquidity) with no expense to the government. But if it pays a premium for the assets, then the operation includes also a transfer (equal to the amount of the premium) that should be considered an expense.
  • Timing. In recapitalizing an entity there may be several years between the time losses occur (or before anticipated losses occur) and the government intervenes. On a cash basis the payment to recapitalize an entity would be recorded when the disbursement is made but on an accrual basis would be recorded in the period when the losses actually occurred.
  • Balance Sheet. After assets have been purchased, any gains or losses from changes in price (even if the government does not sell the assets) should be reflected on the balance sheet. If the assets rise in price there will be a positive effect on government net worth and if they fall, the government’s net worth would deteriorate.
  • Ownership. When the government takes control of a private corporation (as the United States did with government-sponsored enterprises Fannie Mae and Freddie Mac) then the corporation should be considered part of public sector even if technically ownership of its underlying assets remain in private hands. Control is defined as the ability to determine the general policy of the corporation.

Using the old cash accounting framework may lead to wrong conclusions, Dublin said. “All the criticism you hear” about the U.S. intervention, for example, “is that this basically is a government expense…. That need not necessarily be the case if the government acquires an equity stake in the companies.”

In many cases, the transactions are of a financial nature that do not affect the net worth of the government. The government “gives up an asset, cash, and in its place it acquires another asset,” Dublin said. “In the case of the automobile companies, it’s a loan, in which case [the government] has a claim on these companies.… If all goes well, [the government] might even make money.”

That’s not to say that every government intervention will make money. Some might be obvious money losers from the start and should be clearly considered an expense from the inception. If the government were to pay a premium for an asset (buying it for more than it is worth) then the transaction might have to be split between pure financial transaction and an expense (the amount of the premium).

When deficits are affected

If the governments “don’t acquire an asset of equivalent value, if these are assets that are compromised, and they are not what they say they are worth, that the market value is something less than what they government pays for them, then obviously it will affect the deficit. Because we treat this as a capital transfer,” Dublin said. Whenever a government acquires an asset that it knows will not be repaid, it is counted as capital transfer, not an asset, and that increases the deficit.

“But without that knowledge that these are impaired assets, we cannot make that judgment. What we are saying is that on the face value it looks as if the government is acquiring a financial asset equivalent to whatever loan they are extending.”

It may often be difficult to put a value on assets, according to Cornelis Gorter, Deputy Division Chief in GFD. In some cases, an asset is not publicly traded, so there is no market price. In other cases, when a government underwrites some sort of guarantee, the cost, if there is any, may not be realized for some period of time. Other times, the cost seems clear from the outset.

Moreover, things change. If the value of an asset appreciates or depreciates after it is acquired (because of, say, market conditions), then that has to be accounted for too—in an adjustment called “other economic flows,” which records changes in value of a government’s assets that are not the result of transactions.

Still a work in progress

Of course, the nature and purpose of government interventions—which are likely to grow and spread as the global economic crisis deepens—have been evolving, Gorter said. At first the U.S. Treasury planned to use $700 billion to buy bad assets from financial institutions. Then it decided to inject funds directly recapitalizing them. When the Bush administration decided to shore up the auto industry it used a portion of that $700 billion Congressional appropriation to make loans to General Motors Corp. and Chrysler Corp, loans that will be reviewed March 31.

“There is a lot of experimentation going on…A lot we need to nail down,” Dublin said.

In cash accounting, many of these types of transactions are treated as expense items, which have an impact on the fiscal deficit. In addition, changes in the value of assets are not captured. It is this type of deficiency in cash accounting—that it does not properly reflect what a government is doing—that led the IMF’s Statistics and Fiscal Affairs Departments to spearhead development of the GFSM 2001 framework.

Old framework has deficiencies

Deficiencies in the old framework had long been recognized. In many cases, IMF economists in annual country consultations have been making ad hoc adjustments to statistics to compensate for data problems—adjustments that often anticipated the 2001 manual’s approach.

Following a seminar on using theGFSM 2001 for fiscal analysis, the IMF’s Executive Board said that the 2001 framework would lead to greater transparency and consistency in the presentation of country fiscal data in staff reports than does the approach in the predecessor 1986 Government Financial Statistics Manual. There would be no need, for example, for economists to make those ad hoc adjustments, which often hindered making comparable cross-country data analysis.

The complete 2001 approach produces three interrelated tables:

An operating statement that records, on an accrual basis, a government’s revenues and expenses for an accounting period.

An integrated balance sheet that shows the government’s net worth (assets less liabilities) at the beginning and end of the accounting period as well as the flows that explain the changes in net worth during the period (see table).

A cash statement that shows how much money the government takes in and how much it disburses in an accounting period, classified by economic activity and the resulting cash surplus or deficit.

Costs of a new framework

 But there are costs to moving to the new framework, both for the countries and the IMF. One problem is data issues, which is why the IMF’s Executive Board in 2005 ordered studies of the cost of making the shift. Often, countries are capable of reporting cash data classified according to the GFSM 2001. But many still do not compile even simple balance sheets and need to begin the process of integrating accrual recording in their accounting systems. Usually countries have reliable data on their debt available, but statistics on their assets, particularly non-financial assets such as roads and land, may be lacking.

Another impediment to implementation occurs for countries that have an IMF program under way. It would be difficult to change the definitions of fiscal targets and performance criteria mid-stream. In other cases, if data produced using the 1986 format appear sufficient for country surveillance, there is less incentive to shift to the new format, even if it promises superior results.

IMF Senior Economist Isabel Rial, who coordinates work on the studies, said, however, that even a partial move to the 2001 framework can result in a substantial improvement in analysis.

Change in emphasis

The accrual basis of recording transactions supplements the information on cash flows and provides a better measure of the government’s financial position. Most critically, it requires governments to record transactions—such as wages or debt service (including interest) or pension obligations—when they commit to them, even if the related cash payments are made days, weeks, months or, with pension obligations, even years in the future. Transactions should be accounted for when the commitment is made because that is the point at which the economic impact is felt. Under cash accounting, even if a government has committed to a payment, if it is not made, it is not recorded. “Many countries accrue arrears on their debt servicing or even on their wages and salaries and payments for supplies and it’s very easy for them to meet certain targets by postponing payments,” Dublin said. The accrual approach permits a better measurement of the impact of government activity and the magnitude of the transactions in which a government is involved.

Revenue—such as taxes, royalties, and investment returns—less expense (including depreciation) yields the government’s net operating balance. When the net acquisition of nonfinancial assets is subtracted from the net operating balance, the result is the net lending/borrowing balance. That essentially measures the government’s impact on the economy. If the result is negative, the government pulled financial resources from other sectors of the economy (net borrower) and if positive it made financial resources available to other sectors of the economy (net lender).

The net operating balance is critical to determining the change in a government’s net worth during an accounting period. The year’s net operating balance and other economic flows are added to the net worth at the beginning of an accounting period to arrive at the government’s net worth at the end. In addition to valuation changes that affect the government’s asset position, such as exchange rate changes that could affect the value of bond holdings, other events can occur that could affect the government’s assets: a hurricane might wipe out building and bridges; oil reserves might be reduced or augmented.

“The evolution of net worth over a number of years will give you a good indication of the sustainability of government policies,” Dublin said. “If net worth is declining over time, the government cannot continue to pursue those policies indefinitely. If I had to put it in a nutshell, the data provided by the new system make it possible to add sustainability analyses to the liquidity analysis that was the focus of the previous system. In addition, the new system is harmonized with public accounting standards and with the other macroeconomic statistics.”

James L. Rowe is a Senior Editor on the staff of Finance & Development.