GDP versus GNI: a brief look at the "Atlas Method" of the World Bank

[Jurriaan Bendien]

In recent years, the World Bank like other international financial
institutions have been using GNI more frequently as a measure of national
income in international economic comparisons. Why ?

In order to understand this, we need to know first of all what GNI is. Well,
basically it is just a new acronym for the good old Gross National Product
we used to know, which in the 1970s was largely abandoned in favour of GDP.
The rationale given for the change in wording is that GNI is considered more
a concept of "income" rather than a "product measure" (sic.).

Officially, GNI equals GDP (the sum of value added by all resident
producers in the sphere of production) PLUS any product taxes
(less subsidies) not already included in the valuation of net output, PLUS
 net receipts of primary income (compensation of employees and
profits) from abroad. (To smooth fluctuations in prices and exchange
rates, the "Atlas method of conversion" used by the World Bank then
applies a conversion factor, which averages the exchange rate for a
given year and the two preceding years, adjusted for differences in
rates of inflation between the country, and through 2000,
the G-5 countries (France, Germany, Japan, the United Kingdom,
and the United States). For 2001, these countries include the Euro Zone,
Japan, the United Kingdom, and the United States.)

The extra net income added to GDP thus refers specifically to the income
received from labour and capital owned overseas by residents of the domestic
economy, MINUS similar payments made from the domestic economy to
non-residents overseas. Conceptually, these incomes must be related to the
social accounting concept of "production". In other words, it has to be new
income generated by the application of factors of production overseas, which
are owned by domestic residents, income which represents a fraction of
new value added.

So far, so good, but now I want to know, just what difference does this
component make to the GNI and GDP totals ?

In order to do that, I have taken the British statistical data for an
example. I have done so, because these data are supposed to be of high
quality. By subtracting GDP at market prices from GNI at market prices, I
obtain the "net income from abroad". In the British case, the recent
official figures were (in current values, British pounds):

1996: minus 2.6 billion (negative balance)
1997: 1.3 billion
1998:  9.1 billion
1999:  0.4 billion
2000:  5.6 billion
2001: 13.2 billion
2002: 18.3 billion

If we look at the historical time series, we discover that the "net income
from abroad" component was, in the British case, actually negative for most
of the 1980s, and up to the mid-1990s. But, since that time, it has always
been positive. So whereas previously it lowered national income totals, now
it raises national income totals. So if we use the GNI measure, national
income nowadays will always appear larger, than if we use the GDP measure.
In 2002, for example, British GNI was 1.7 % larger than British GDP.

But what, you ask, is the cumulative magnitude of this increase due to the
contribution of "net income from abroad" ?

Let's calculate, for the last five years of annual data, the percentage
which a positive "net income from abroad" actually contributes to GNI, and,
just to compare, what fraction of GDP this net income represents, using
official current values provided by the National Statistics Office:

1998: 1.05 % of GNI and 1.06 % of GDP
1999: 0.04 % of GNI and 0.04% of GDP
2000: 0.59 % of GNI and 0.59 % of GDP
2001: 1.31 % of GNI and 1.32 % of GDP
2002: 1.72 % of GNI and 1.76 % of GDP

As you can see, the proportions are almost identical for GNI and GDP, but
the important thing is, that GNI INCLUDES "net income from abroad", whereas
GDP EXCLUDES it.

In addition, we can see that, in recent years, "net income from abroad" has
actually increasingly raised the British Gross National Income (GNI)
aggregate, i.e. added an increasingly larger amount to it. That is, the
value of British GNI is increasingly higher than British GDP, purely because
of "net income from abroad".

My next question is, what is the relative proportion of capital (property)
income and labour income (compensation of employees) in the "net income from
abroad" aggregate ?

In order to answer this, I refer to the current account data in the UK
Balance of Payments Statistics, from which I can obtain data on the net
income receipts representing "compensation of employees" from overseas
(pre-tax wages and salaries, plus social security levies). By relating this
labour income to "net income from abroad", I obtain the percentage which
"compensation of employees" represents out of the total  "net income from
abroad", as follows:

1998: (a negative balance of -10 million pounds)
1999: 56.6%
2000:  2.5%
2001:  1.3%
2002:  1.6% (my own estimate)

As you can see, in recent years except 1999, the vast bulk of "net income
from abroad" must have represented gross profits from overseas production,
and the proportion represented by wages & salaries actually declined since
2000.

In 2001, for example, we know that "net income from abroad" amounted to 13.2
billion pounds, but net income representing compensation of employees was
just 180 million pounds. Thus, even if "net income from abroad" includes
some product taxes (VAT etc.) it is clear most of the aggregate must be
gross profits. And, of course, "compensation of employees" conceptually
includes higher management salaries as well.

We know that the official British GDP figure (at market prices) for 2002 is
about 1.04 trillion pounds and the official figure for British GNI (at
market prices) is about 1.06 trillion pounds, in current values. Using a
mid-range 2002 conversion rate of 1 x 1.535, we can convert the
pounds to US dollars, in which case British GDP is equivalent to
US$ 1.6 trillion and British GNI is equivalent to $ 1.63 trillion.

As you can see, having done the conversion, GNI works out to be
about 300 billion US dollars LARGER than GDP, which is what we
would expect.

Let us now compare this result, however, to the latest World Bank figures
for
the United Kingdom, looking at the same year, using the "Atlas method".

Strangely, here we find, at the World Bank website, that following the
"Atlas method", GNI in current US$ is valued at 1.5 trillion whereas GDP in
current US$ is valued at 1.6 trillion !!!

In other words, according to the World Bank, not only does a conversion to
US dollars LOWER British GNI, but also, suddenly, British GDP is not 300
billion US$ dollars smaller than GNI, but US$ 100 billion dollars LARGER
than GNI.

Somewhere along the line, a few hundred billion dollars got lost. Can you
trust bankers like that ?

Jurriaan

PS - according the the World Bank, the Brazillian GNI (Atlas method) is
about 9 percent larger than Brazilian GDP whereas e.g. in the case of
Indonesia, GDP is more than 13 percent larger than GNI (Atlas method).













http://www.statistics.gov.uk/STATBASE/tsdataset.asp?vlnk=205&More=N&All=Y

The current account records international transactions in goods and
services, other net income from abroad and current transfers to and from
abroad. If your country is running a current account deficit with the rest
of the world then you need more foreign currency than foreign countries need
your domestic currency. The transactions underpinning these dealings as well
as other international capital market activity are in turn recorded on the
financial account, simply described as international transactions in
financial assets. The financial account therefore offers detail on the
international flows of financial assets and liabilities.

The current account deficit shows the extent to which a country has been
spending relative to its earnings, important to understanding trends in
domestic economic behaviour. If expenditure exceeds income on current
account,  then (1) rising domestic demand stimulates imports,  or (2)
Domestic producers cannot sell enough domestically or abroad; (3) foreign
goods are outcompete local goods, (4) overseas demand for domestic goods
shrinks while domestic demand remains strong.

http://www.statistics.gov.uk/downloads/theme_economy/Pink_Book_2002.pdf