3. Vouchers and Inflation
Let us now return to the question whether voucher privatization will trigger off a rapid inflation. As I have already stated in the first part of my Lecture voucher privatization has been often criticized for the reason that allegedly people will immediately sell the acquired property in exchange for consumer goods. As a result the distributed property would turn into "hot money" which would start hyperinflation. This argument is of a rather different nature than the danger of the financial collapse discussed in the preceding part although some consequences may be very similar. It was argued in the preceding part that if the market value of property acquired for one voucher book falls below the level of guarantees issued by some privatization funds their shareholders will naturally start to claim the guaranteed amounts and the funds will be forced to sell some of the shares acquired by privatization at the depressed price.
The presently discussed argument is more general, it states that people will not only withdraw from the funds but they will also sell the shares acquired through privatization regardless of whether their market price is higher or lower than the issued guarantees. This view is usually supported by the reference to the experience of Great Britain, Canada and other countries where a significant number of people sold the shares soon after privatization, indeed.
This seemingly convincing criticism of voucher method of privatization is based on two logical lapses. The first lapse consists of the confusion of two quite different economic notions, and namely the value of property and the amount of money in the economy. Voucher privatization redistributes the existing property from the ownership of the State into the hands of population, however it does not create new money by itself. The money supply in the economy is regulated by the Government through the agency of the State Bank. Even though we all shall most probably agree that the inflation is caused by the growth of the quantity of money, it certainly does not imply that the distribution of state property by vouchers must increase the inflation rate. It depends on the Government monetary policy during privatization: will it be restrictive or expansive.
Another logical lapse is the implicit assumption that if somebody sells shares of joint stock companies or shares of investment funds he or she will spend all the proceeds for the purchase of consumer goods. But this is far from the truth. People usually behave rationally and as we shall see in a moment, the immediate spending of the entire property even if fallen "out of the blue" is not rational. If this were the case we would observe people spending immediately all winnings in lotteries, gifts or inheritance from their relatives. Instead we can see that a normal household will use only a small part of the unexpectedly acquired wealth on the instantaneous increase of consumption while it will save and invest the rest to make the growth of future consumption possible.
There is no reason to expect that after the end of the first wave of privatization all the people will try to get rid of their shares and spend all the proceeds on consumption. On the contrary, it can be expected that people will spend only a small part and that they will want to invest the greater part in profit generating assets. It is true, though, that people have different attitudes toward risk. Only a minority of them, let us say 10 to 20 per cent are willing to tolerate high risk in order to achieve higher expected returns. Majority of population is risk averse which means that it is content with lower but also much safer returns. In the British and Canadian privatization many people were getting rapidly rid of shares in order to transform the acquired property into a less risky form and not to spend it all at once.
For the same reason the majority of people in Czechoslovakia put their vouchers in investment funds which provide lower but also less risky returns. But for many people even investment funds may be too risky. These people will sell shares and their interests in investment funds in order to reinvest their property in the form better suited to their preferences. Some of them will buy a family house or an apartment, others will invest in their business, others will purchase government bonds or other securities.
Now we face the task to estimate what part of the property acquired in the first wave of voucher privatization will people most probably spend immediately on consumption and what part they will save and invest in some form of assets. Two economic theories dealing with this issue on the macroeconomic level may serve us as a guidance. They are the so-called "permanent income hypothesis" of Milton Friedman and the so-called "life cycle hypothesis" of Ando and Modigliani. For several decades both of these theories have had a deserved place in macroeconomic textbooks and have been many times empirically tested using the data of numerous countries. Most of these empirical tests as well as new theoretical developments have confirmed their validity. These two theories can give us an approximate answer to our question.
Friedmans "permanent income hypothesis" is based on the assumption that people distinguish between permanent and transitory changes in their incomes. The permanent income is a capacity to generate certain level of income in the long run. In contrast to that the transitory income involves temporary and mostly random variations up and down from the permanent income. Transitory income may be, therefore, either positive or negative. Friedmans theory asserts that people determine their consumption expenditures exclusively according to their permanent income. Transitory income does not cause any change in the consumption expenditures. It only increases or decreases savings. This, however, does not imply that all the proceeds from voucher privatization will be considered by people a transitory income 100 percent of which they will save. Wisely invested increment of property will permanently generate for them returns equal at least to the real interest rate. These permanent returns will become part of their permanent income and will result in a proportional increase in consumption expenditures. If we assume that the real interest rate is 6 percent then 14 thousand Kcs which will come "out of the blue" should raise the permanent income and consequently the annual consumption expenditures at most by less than 1 thousand Kcs while the remaining 13 thousand Kcs or more should be saved and invested.
Andos and Modiglianis "life cycle hypothesis" is based on a similar assumptions about rational behavior of households but it formulates the model in a somewhat different way. The result is again very close to the result of the Friedmans model. The life cycle hypothesis starts from the fact that in each year of its life a typical household owns a certain amount of previously saved or inherited property. The household estimates the expected future flow of labor income in the remainder of its life cycle and also the expected future changes in the real interest rate. Using this information and its preferences between the present and future consumption the household decides what amount it will consume each year and what it will save. Consumption may be in some years lower than incomes and as a result of positive savings the household wealth would grow. In other years, e.g. at the very beginning or at the end of their career the consumption expenditures may be higher than current incomes and the household wealth would decline.
Thus the life cycle hypothesis explains how a rational household spreads optimally consumption over its life span so that it can build up its wealth by savings during the years of high labor income and draw from it in the time when its labor income is low or zero. Naturally the optimal strategy must include such a volume of wealth to be left at the end of the life cycle which the older generation wishes to bequeath to its descendants. From this model Ando and Modigliani derived macroeconomic consumption function according to which aggregate consumption expenditures are a linear function of an after-tax income and an accumulated wealth. Empirical studies have shown that the regression coefficient at the wealth variable equals approximately the mentioned 6 percent. This again implies that from a sudden and unexpected increase of the wealth of the households acquired as a result of voucher privatization about 6 percent will be spent for immediate consumption and the rest will be saved and invested.
For the better understanding of this point I have performed a number of numerical calculations on the computer with a simplified model to demonstrate the impact of voucher privatization on consumption expenditures of households in typical Czechoslovak conditions. Although these calculations are merely for illustration and are not based on actual empirical data I believe that the parameters were chosen sufficiently realistically so that the results show convincingly how small is the probability that the consumption expenditures would increase by more than 10 percent of the value of the privatized property.
The model of a dynamic optimization of a typical household is based on the maximization of utility function
(4) U(t) = St ln(C(t))/(1+d)s-t
subject to the budget constraint
(5) St C(s)/(1+r)s-t = A(t) + St Y(s)/(1+r)s-t - A(T+1)/(1+r)S
T ... length of life of the household
t ... age of decision making (today),
s ... remaining years of life (future)
U(t ).. total utility during the rest of life
C(s).. annual consumption expenditures at the age s
Y(s ).. annual labor income at the age s
A(t).. value of assets at the beginning of the year
r ... expected interest rate
d ... annual rate of time preference
^ ... operator of power
The model assumes that the household makes repeatedly at the beginning of each year of its life decision on how much out of its total income it will spend on consumption and how much it will save in each of the remaining years. In this decision making it has the knowledge of the value of its present assets and estimates the expected flow of its labor income (including pension) for the rest of its life. For simplification I assume that the household considers interest rate r and its time preference constant during the rest of its life even though it is not too difficult to allow for a change in both of these parameters. The decision is made in such a way that at the end of the life there are assets A(T+1) left, and the utility function (4) is maximized.
The constraint (5) states that in each year t the present value of future consumption flow must equal the value of the assets at the beginning of the year t increased by the present value of future flow of labor income and decreased by the present value of the assets to be bequeathed to the future generation at the end of the life. This constraint ensures the balance between incomes and expenditures of the household. It is assumed that assets A are invested and yield profit (interest, dividend) at rate r. The model admits that A may be in some years negative, i.e. that in these years it represents debt on which interest is paid at rate r.
It follows from this formulation of the model that the household has at the beginning of each year at its disposal previously saved assets A(t) plus work income Y(t). As it will spend C(t) for consumption by the end of the year it will have A(t) + Y(t) - C(t) increased by interest. Thus the following formula must apply to the assets in the year t + 1
(6) A(t+1) = (1 + r)(A(t) + Y(t) - C(t)
which is a difference equation describing the dynamics of family assets. Eulers condition of optimality for our model determines that in order to achieve the maximum of the utility function (4) the following difference equation must hold for the dynamics of consumption in time
(7) C(t+1) = C(t)(1 + r)/(1 + d)
If we repeatedly insert (7) in (5) and rearrange the expression we shall get the formula for the consumption expenditures in the year t as a function of assets A(t), the present value of the expected flow of labor income, the present value of assets planned for heirs and discount factor depending on the coefficient of time preference.
(8) C(t) = [A(t) + S Y(s)/(1+r)s-t - A(T+1)/(1+r)T+1-t ] / S 1/(1+d)s-t
This is the sought formula which will allow us to calculate the effect of voucher privatization on the consumption expenditures of households. Let us consider a household with, say, one adult, which at his/her age of t will suddenly acquire additional assets by voucher privatization. Thus A(t) in the numerator of the right-hand side of the equation (8) will be increased and we can easily calculate the amount by which consumption expenditures of this household will increase in the year t. Using equation (7) we can then calculate how the consumption expenditures of this household will grow in the subsequent years.
The following table presents an example of the household which for simplification consists of one adult, a female wage-earner. This woman will complete her education at the age of 18 and start to work in the following year. Her initial salary is 2500 Kcs which is 30 thousand a year. This salary will grow by 2 percent annually so that at the age of 63, i.e. in the last year before retirement it will achieve the level of 71.7 thousand a year. Her pension will be 43.02 thousand Kcs and will remain unchanged until her death at the age of 78. This woman inherited 50 thousand Kcs when she was 18 and at the time of her death she intends to bequeath to her relatives 200 thousand Kcs as inheritance. In the course of her whole life the real interest rate is constant at the level of 6 percent, and her coefficient of time preference is 3 percent.
On the basis of these data we can calculate an optimal path of consumption (column C1 of the table) and the optimal path of the assets resulting from inheritance and savings (column A). Now let us examine how the optimal consumption is influenced by the acquisition of additional assets in the amount of 14 thousand Kcs as a result of voucher privatization. Columns C2 to C7 show the changes in the optimal path of consumption if privatization occurs at different age of the examined person. C2 represents the case when privatization occurred at a young age of 20. Each subsequent column postpones privatization by ten years.
The table shows that a rationally behaving household will spend for consumption in the first year only a small portion out of the 14 thousand Kcs acquired by privatization. The rest it will invest and to ensure for itself higher consumption in all subsequent years. The following table shows the amount by which the optimal consumption in the first year after privatization will be increased.
C O N S U M P T I O N
According to this calculation a young household will spend on consumption only approximately 500 Kcs, i.e. about 3.5 percent out of the acquired assets. This share grows with the age but only in the last decade of life it will exceed 10 per cent. It could seem that this result is so low because the chosen household was not very poor. However, that would be a mistake. Equation (8) shows clearly that if labor income and assets planned for inheritance remain unchanged then the consequence for C(t ) of the change in A(t) depends only on the age t and on the coefficient of time preference d. This means that with d = .03 even a very poor household would spend for consumption equally low share of the assets acquired by privatization.
Let us present one more table which summarizes the dependence of the examined share both on the age and the coefficient of time preference.
SHARE OF CONSUMPTION EXPENDITURES IN THE PRIVATIZED PROPERTY (percent)
Usually the coefficient of time preference is assumed to be about 3 percent. Values of this coefficient higher than 10 percent would be very unusual. It is quite understandable that time preference of very old people similarly as of some groups of poor may be higher than normal. However, at least in the case of poor people it is an open question whether they have high time preference because they are poor or whether they are poor because they have high time preference, i.e. a strong urge to consume immediately rather than save. In any case this table indicates that for the most numerous strata of population the examined share will range between 3.5 and 10 percent, and that it will be greater than 10 percent only for marginal strata.
In the end of this part let me mention briefly inflation and restrictive policy. We have just estimated that with constant prices and constant interest rate the demand for consumer goods will increase as a result of voucher privatization by less than 10 percent of the value of privatized property. This relatively small, however, not negligible increase of aggregate demand may be at least partly compensated by the increased aggregate supply. If this is not the case the Government can apply restrictive fiscal or monetary policies. Restrictive fiscal policy would compensate the increased personal consumption expenditure by reduced government spending. Restrictive monetary policy would increase the interest rate and that could push down the consumption demand to a balanced level. Our equation (8) shows that the increase of the real interest rate will decrease consumption expenditures of households and increase their savings.