International Saving Comparisons

Following up on yesterday’s post about the capital flow puzzle, I thought I’d examine savings and consumption patterns more closely, since differences in savings behavior are (as I argued) the crucial determinants of the odd pattern that we currently see wherein poor countries are lending vast sums of money to the wealthy United States. The table below presents a snapshot of the tremendous disparities in savings and consumption behavior between developing Asian economies and the US. All data is for 2003 unless otherwise noted.

Notes: Japan gross savings data for 2002; China household savings data from 2000.
Sources: Gross savings rates from OECD and ADB; household saving rates from OECD and sundry national statistics agencies; consumption and CA balances from IMF.

Rates of gross national saving – which include saving by households, firms, and the government – are double or triple the US rate in most of the countries in the table. Generally, much of the difference is accounted for by higher rates of savings by households. (Note that it is surprisingly difficult to come by household savings data for developing countries; I’d appreciate any data tips to help fill in my table.)

Another way of looking at the same thing is to examine how much consumption spending each country does. Unsurprisingly, the US leads the pack when it comes to consuming; by contrast, several Asian countries have extremely low rates of consumption spending. As I mentioned in yesterday’s post, the problem in Asia is too little consumption, if anything.

The last column in the table shows the effect of these divergent savings and consumption patterns: net international borrowing or lending, or the current account balance. Nearly every country in the world right now is a net lender, with almost all of those funds going to the US.

What explains these dramatic differences in savings and consumption patterns? It turns out that there’s no consensus on a single answer. Numerous possibilities have been examined in the literature to explain different savings rates across countries (see this World Bank paper for an example of a fairly thorough overview), including income levels and growth rates, financial sector development, credit constraints, fiscal policy, and pension institutions. Needless to say, different explanations seem to work for different countries at different times, and savings differences are usually due to a combination of several factors.

But my favorite theory, at least in this case, is that the high rate of saving in many Asian economies is primarily due to the inertia of consumption spending.

It is an increasingly well-established empirical fact that consumption displays a great deal of inertia – people’s consumption habits only change slowly, even when their income rises or falls dramatically. If a country is experiencing rapid economic growth, this could well explain why savings rates are high in those countries: households’ consumption has not yet caught up to their increased income, so they save the difference. Of course, to some degree the causation may run the other way (i.e. higher savings may lead to faster growth), but the evidence seems to suggest that the primary direction of causation is from higher growth to higher savings.

This helps to explain why some of the fastest growing economies like China have some of the highest savings rates. It might also help to explain why Japan, which has experienced a dramatic slowdown in growth over the past 10 years, has also experienced a dramatic fall in its household savings rate over the same period of time, from about 15% in 1990 to about 7% today.

At any rate, the upshot is that the US will continue to borrow from the rest of the world until it starts to save more, or households around the world start to consume more of their income. So the real question then becomes this: what will make either of those things happen?

Kash