In most systems of national accounts the GDP deflator measures the difference between the real (or chain volume measure) GDP and the nominal (or current price) GDP. The formula used to calculate the deflator is:
Dividing the nominal GDP by the GDP deflator and multiplying it by 100 would then give the figure for real GDP, hence deflating the nominal GDP into a real measure.
It is often useful to consider implicit price deflators for certain subcategories of GDP, such as computer hardware. In this case, it is useful to think of the price deflator as the ratio of the current-year price of a good to its price in some base year. The price in the base year is normalized to 100. For example, for computer hardware, we could define a "unit" to be a computer with a specific level of processing power, memory, hard drive space and so on. A price deflator of 200 means that the current-year price of this computing power is twice its base-year price - price inflation. A price deflator of 50 means that the current-year price is half the base year price - price deflation.
Unlike some price indexes, the GDP deflator is not based on a fixed basket of goods and services. The basket is allowed to change with people's consumption and investment patterns. (Specifically, for GDP, the "basket" in each year is the set of all goods that were produced domestically, weighted by the market value of the total consumption of each good.) Therefore, new expenditure patterns are allowed to show up in the deflator as people respond to changing prices. The advantage of this approach is that the GDP deflator reflects up to date expenditure patterns. For instance, if the price of chicken increases relative to the price of beef, people would likely spend more money on beef as a substitute for chicken. A fixed market basket measurement would miss this change.
In practice, the difference between the deflator and a price index like the CPI is often relatively small. On the other hand, with governments in developed countries increasingly utilizing price indexes for everything from fiscal and monetary planning to payments to social program recipients, the even small differences between inflation measures can shift budget revenues and expenses by millions or billions of dollars.
In recent years, some commentators have expressed concern that the national accounts may overstate spending on computer hardware because of the way the hedonic index and implicit price deflator are used. It is well-known that the prices of a unit of processor speed, a unit of memory, and a unit of hard drive capacity have declined very quickly since 1995. Therefore, the current-year (say, 2003) price deflator for an entire computer - using the hedonic method - is less than one relative to a base year of 1995. This means that when nominal spending on computer hardware is divided by the deflator to give real spending on computers, the number rises. (The "deflator" here is actually an inflator!) From the second quarter of 2000 through the fourth quarter of 2003, the government estimated that real tech spending rose from $446 billion to $557 billion, when nominal spending only increased to $488 billion. Some analysts feel that this overstates the "true" spending on computers by $72 billion. However, it is also true that this extra $72 billion captures the increase in value and utility of the computers that were purchased in 2003 as compared to 2000, due to the former's superior quality and capability for the same nominal price as the latter.