Logo Background RSS


GDP vs GNP: Relevant?.

  • Written by Syndicated Publisher 41 Comments41 Comments Comments
    August 16, 2011

    There is a lot of debate about the best measures of income and income growth. Is it better to use GDP or GNP? And does it matter?

    GDP measures “Domestic Production”. In other words, it measures the value of all that is produced inside the U.S.; which is also the sum of the income of those people who contributed to this production. Normally, if consumption exceeds income, this excess consumption has to be imported from abroad; and paid for with money borrowed from people living abroad.

    In the national accounts, all imports are “subtracted” from GDP as GDP = C+G+I+X-M. In the aggregate, all that is consumed in C which is imported (i.e. consumed in excess of income) is subtracted back in M; and thus does not contribute to GDP.

    It therefore follows that, starting from a zero trade balance, if a country’s consumption growth is larger than its income growth, its domestic debt will inflate. And, if the excess consumption is not eventually reversed, a chronic indebtedness problem lies ahead.

    Some claim that the U.S. government’s decision to switch from GNP to GDP to measure growth about twenty years ago is responsible for blurring the proper alarm signal that could have warned us about our debt addiction and our resulting current excess indebtedness.

    However, given the dynamics of national accounting explained above, how could using GDP have failed to capture that we were borrowing beyond our means?

    The answer is that it normally should have. But, in the case of the United States, it did not because of the particular situation of the country as it built large claims on foreign assets as a result of its post WWII reconstruction efforts and the expansion of its multinational corporations in the second part of the 20th Century.

    The difference between GDP and GNP is that GNP measures National Production (rather than Domestic Production). That is, it measures the income of Americans obtained from both their share of Domestic Production and their share of Foreign Production. Until 1980, most of the income that was generated from domestic production flew into American hands as we owned most of our domestic assets. However, because Americans also derived a lot of income from assets they held abroad, GNP was larger than GDP and Americains were thus able to spend more than they would have been able to, had they relied only on their domestic sources of income.

    Let us use an analogy to illustrate the process. Let us assume that you derive your income both from your salary as an employee of Company A and from dividends from your ownership of shares in a portfolio of stocks. GDP would be the equivalent of your salary income (your share of what is produced at Company A) and GNP would incorporate both sources of income as it would also include dividends paid by these other companies in which you own shares.

    To understand why using GDP as a growth measure could hide a borrowing problem, let’s push the analogy a little further. If you did not have any other sources of income but your salary, your bad borrowing habits would be immediately apparent as your consumption would exceed your income. However, having a portfolio of stocks lets you spend more than your salary income alone would. If you only spent the dividends that you pocket every year, this would not be a problem. After all, it is income to which you are entitled by virtue of owning the shares (probably the result of past savings that allowed you to invest to buy these shares).

    Where the consumption starts to become problematic is when you start selling shares in your portfolio to finance your current consumption. If this continues long enough, over time, the assets held in your portfolio will gradually shrink to zero as will the income derived from these assets. If you cannot curb your enthusiasm for excess consumption, you will soon find yourself in a precarious situation. Your salary income (GDP) may be still be growing but this growth may be hiding the fact that your total income (GNP) is shrinking as you are selling stocks (and thus foregoing future dividends) to finance your consumption.

    This is what likely happened to the U.S over the last twenty years or so. And the situation could become worse if we cannot get rid of our borrowing habit. As a result, we are probably already a net debtor nation. In other words, foreigners might have more claims on our assets than we have on theirs; a big reversal from the pre-1980 situation.

    When debt financing was insufficient, not only did we sell large chunks of our foreign portfolio of assets to finance our past trade deficits, we also sold to foreigners some of our domestic assets as well. This means that we are no longer entitled to all revenues from our domestic production. In other words, GDP may now be overestimating our total income (GNP) as we now owe more to foreigners than they owe to us. As the share of what we owe to foreigners continues to grow faster than GDP, our true income growth is obviously also overestimated.

    To continue with the analogy, you now have no more stocks in your portfolio and, every year, a higher and higher proportion of your salary goes to paying interests on money you borrowed from the bank. Even if your salary is still growing, this excess borrowing habits of yours is constantly reducing your “true” income growth. This might be a problem indeed!

    In conclusion, the switch from GNP to GDP might have concealed a problem that could have otherwise been identified early. But is this hypothesis verified in the data? I could not find strong confirmation of the phenomenon. In the graph below, you can see that the GDP/GNP ratio has been quite stable over a long period of time, at least since the late eighties. Normally, if what I said above truly happened, we should have expected the GDP/GNP ratio to have started below 1 in the fifties, remained there through the sixties and the seventies and to have slowly increased above 1 in the last twenty to thirty years.

    The blog Angry Bear offers an explanation as to why this may have be the case even though we have a strong current account/trade deficit since the early eighties. Moreover, the data is quite old and the situation might have deteriorated significantly during the last seven years.

    All this to say that, using GNP between 1990 and 2004, rather GDP, might not have send us as strong a signal about our bad debt habits as we might have thought. It did in the eighties (look at the graph) when we were still using GNP. But we seem to have missed the signal that it was clearly sending then. We decided instead to switch to GDP precisely at that time.

    It’s too bad that I could not find the data for the period after 2004 because this is when the real borrowing binge started and the GDP/GNP ratio probably now stands well above 1. Yet, like in the eighties, I doubt that we would have had the presence of mind and the political courage to use this indicator to get rid of our debt addiction.

    Shared under Licence From

    The Sceptical Market Observer: GDP versus GNP: Relevant?.

    Images: Flickr (licence details)

    About The Author – Luc Vallée

    Currently President of The Independent Market Observer. Chief Economist and Vice-President at the Caisse de dépôt from 2001 to 2008. Chief Financial Officer and Vice-President, Corporate Strategy for Mediagrif Interactive Technologies from 1999 to 2001 – MDF.TO. Deputy Treasurer at Canadian National Railways, 1997-1999. Associate Professor of Applied Economics at l’École des Hautes Études Commerciales (HEC) in Montréal, 1989-96. Consultant for the World Bank, the Canadian government, the Quebec government and the City of Montreal. Deputy director of the Center for International Studies, 1993-1996. Adviser to the investment banking division of Société Générale in Canada, 1996. President of ASDEQ in 2005-06. Member of the National Statistics Council of Statistics Canada, 2008-now. Ph.D. (1989) in economics from the Massachusetts Institute of Technology.
    Share on FacebookTweet about this on TwitterShare on Google+Share on LinkedInPin on PinterestShare on StumbleUponShare on RedditShare on TumblrDigg thisBuffer this pageFlattr the authorEmail this to someonePrint this page