The Paradox of Thrift is a very old concept in macro-economics discovered by Keynes, whose essence is the following:
Suppose investment in any period is autonomously given. Now, actual (ex-post) investment and savings are always equal. Therefore, given this investment expenditure, there will be an equivalent amount of savings in the economy. If the planned (ex-ante) savings and investment of the household and the capitalists are equal then this is an equilibrium situation for the economy, in the sense that nobody has any incentive to change from this state of affairs.(This is shown in the figure as Y1, which is the output given this level of investment which is shown by the horizontal line). Now, suppose due to some reason, the households want to save more. This results in a shift of the savings curve from S to S1). In other words, the planned (ex-ante) savings increases. Given, the autonomously given investment there appears a disequilibrium in the system. This is because, at the output Y1, planned savings (ex-ante) is greater than the autonomous level of investment. How is this disequilibrium resolved?
With an increase in the ex-ante savings, there arises a problem of demand in the economy because instead of spending on goods and services people want to save more. Now output in any period is given by what is called the multiplier. This is given by Investment expenditure divided by the savings propensity. Therefore, with an increase in the value of savings propensity (higher desired savings) the value of the multiplier declines denoting a lower level of output for the same level of investment. As a result, with a desire to save more, the economy ends up with less output. This is the paradox of thrift. The story can be said in an alternative manner as well. Since, investment is autonomously given, ex-post savings must equal this amount of investment. Now, even when the households want to save more, the final savings can be no more than the given amount of investment. But, with an increase in their ex-ante savings, what happens is that there arises a disequilibrium in the system. This disequilibrium is removed through a fall in the output which results from the problem of effective demand. In other words, in order to make ex-post savings equal ex-post investment, given an increase in ex-ante savings, the output level comes down. It is easy to see that if we assume the investment curve to be positively related with income then there occurs not only a fall in output but also a fall in savings and investment.
A legitimate question can be raised at this point. Why should investment be taken as autonomous? Investment is taken as autonomous here for three reasons. Firstly, the investment decisions on the part of the capitalists depend on their assessment about the future, which is uncertain. Therefore, for simplicity it is assumed autonomously given since it is difficult to predict the exact motive of investment expenditure of capitalists given that the future is uncertain. Secondly, investment once implemented cannot be changed suddenly. So it is autonomous in the short run. In the long run too, investment depends on the demand situation in the economy. Capitalists invest when they feel that there are prospective returns from that investment. Savings play no role in determining this investment. Thirdly, and more crucially, autonomous investment expenditure in Keynesian and Kaleckian approach serve another very important purpose which is to differentiate between savings decisions and investment decisions. In neo-classical models there exists no autonomous investment function. Therefore, whatever is saved is automatically invested through adjustments in the interest rates. However, this need not be the case at all. In fact, the economics of Keynes and Kalecki argues the exact opposite which is to say that investment is never savings constrained and generates an equivalent amount of savings to finance itself. In other words in the identity I=S, the causality runs from left to right and not the other way round. This is precisely how, the paradox of thrift works by bringing savings in equality with investment, as shown in the above figure.
All this is basically Keynesian macroeconomics but is very easily misunderstood. As Paul Krugman says in his blog,
“In normal times, we believe that more saving, private or public, leads to more investment, because it frees up funds. But for that story to work, you have to have some channel through which higher savings increase the incentive to invest. And the way it works in practice, in good times, is that higher savings allow the Fed to cut interest rates, making capital cheaper, and hence on to investment.”
In this paragraph, it is clear that Krugman wants to say that savings determines investment (‘saving ...leads to investment’). On the other hand, he also says that in order to increase investment with an increase in savings what is required is a fall in the interest rate. In other words, he believes that savings and investment are equated through the flexibility of the interest rate. Both these ideas, however, in the light of Keynes and Kalecki are wrong.
The first part in terms of identifying the causality between savings and investment as savings causing investment is wrong. This is because, investment generates demand in the economy. In the presence of unemployment and un-utilized capacity, output increases. Given that a fixed proportion of output is consumed, this results in an increase in savings. In other words, it is savings which is determined by investment. Even when there is full-employment, an increase in investment demand increases the price level relative to wages. This results in an increase in profits. Since profit earners save more, this increase in profits results in an increase in savings. In this case too, it is savings that is determined by investment and not the other way round. To say that more savings, in the presence of downward flexible interest rate leads to more investment is therefore not correct.
Savings is nothing but a decision not to consume in the current period. This decision is essentially what is called a flow decision. Once this decision is made, there remains however another important decision. How does one allocate the savings amongst different assets? This is a stock decision. If the households decide to use the entire savings to invest in capital goods, then of course there is no problem and all savings are used to purchase capital goods resulting in further flow of income and hence no problem of effective demand. But the moment we introduce money as a form of holding wealth, the problem of effective demand comes to the surface. But the question then is why should anyone hoard money which gives zero rate of return while other assets gives at least the prevailing rate of interest. This is fundamentally due to uncertainty that exists in the market.
The rate of return on any asset depends on the expected change in price of that asset, the own rate of interest, the risk premium and the carrying cost. The first two terms are positively related while the last two terms are negatively related. In other words, suppose it is expected that the price of an asset will increase in the future then, ceteris paribus, people will hold on to this asset more and more. This will create an excess demand for this asset and hence will increase the expected price even further. This has happened in the USA, with the housing boom. However, once the boom is busted, the expected change in price falls and the risk premium rises to a great extent. This results in people leaving that asset and holding on to money. For money, rate of return cannot fall below zero. Therefore, on the face of tremendous uncertainty and falling expectation about asset prices, the only safe refuge becomes money, increased demand of which results in the problem of effective demand and hence recession or low growth of output.
Therefore, the real question is not why with an increase in savings, investment is not increasing in the US economy. The real issue is why is investment not increasing in the US economy. This is not solely because of a failure of the Fed to reduce the rate of interest any further. Rather it is a result of tremendous uncertainty in the market which had greatly increased the liquidity preference (demand for money), causing a decrease in effective demand. Therefore, the key to recovery of the US economy is essentially two fold. Firstly, if by chance there arises again a bubble which results in people shifting from money to a particular or a set of assets fueling another boom. The other way is to increase public expenditure which takes on the job of investment and meets the shortfall. The first remedy is currently not taking place in the USA. Moreover, if the economy gets revived again through another bubble, surely there will be another crisis in the future. Additionally, public expenditure in the US has not increased to the requisite extent. In other words, the economy faces a bleak future, completely in line with Krugman’s argument. But the theoretical underpinnings of Krugman’s ideas are problematic, as shown above.