Under inflationary conditions, a shift in relative shares of total income in favor of property income recipients will also result from what may be called the compounding effect. That which is compounded is savings, and the compounding occurs through its reinvestment. Actually, this effect operates independently of price rises during any period in which operating losses and capital losses of firms are at a minimum. However, such a situation is characteristic of an inflationary period. Further, as entrepreneurial income expands, relatively, in an inflationary period, the tendencies toward compounding will be reinforced, if it may be assumed that the marginal propensity to consume out of entrepreneurial income is smaller than out of labor income. As long as inflationary conditions continue, with operating and capital losses at a minimum, those who have enjoyed the enhanced returns may further improve their economic position by earning income on income. The check on the compounding effect, is, of course, the appearance of operating and capital losses. A reduced rate of return on investment will limit its operation but not offset it.
The importance of each of these factors in the relative growth of property income under inflationary conditions will vary with the rate and duration of the inflation. Presumably, lag effect will be most pronounced when the rate of price increases is most rapid. Compounding effect, on the other hand, is likely to be of greatest importance when inflationary conditions have been sustained for long periods of time. The capacity effect will depend on the level from which increases in economic activity proceed and will be most important when increases in output are initiated from a point where many firms are operating close to break-even.
The operation of three factors—lag, compounding and capacity effect—will alter labor and property shares in accordance with changes in the level of economic activity and the level of prices. Lag and capacity effect, in particular, may be expected to be important in year to year fluctuations.
These three factors are independent of inter-industry shifts which may alter labor and property shares in total income, particularly over a period of years. Most writers in the field have stressed the importance of these shifts in altering labor shares. In examining the upward movement in relative labor income between 1919 and 1938 Kuznets found that relative increases in employee compensation were traceable almost wholly to production shifts from industries in which the ratio of wages and salaries to property income was low to industries in which the ratio of wages and salaries to property income was high. Kuznets did not find a measurable tendency for particular industries to vary their ratio of wages and salaries to property income over this whole period. The Department of Commerce has also stressed the importance of inter-industry shifts in the explanation of the larger employee shares in 1950 as compared with 1929.
Unfortunately, it is not possible to disentangle statistically the year to year effects of the three factors noted above from the effects of interindustry shifts over a period of years on labor and property shares in total income.
Finally, it should be noted that the general conclusion which emerges from the examination of labor and property shares in accordance with the concepts developed here is that they are by no means stable, either from year to year or over a period of years. This conclusion differs rather
sharply from that reached by analysts in the Department of Commerce who have stressed the stability in income shares, particularly in the employee compensation share of national income. The explanation for these differing conclusions lies in the concepts which are employed.
Summary The major conclusions which may be drawn from the foregoing investigation of labor and property shares in total income are the following: 1. Since the Twenties labor shares in total income have tended generally to increase. The recovery period of the Thirties shows labor shares higher than in the Twenties, and higher in World War II than in the Thirties. However, the postwar period has tended to reduce labor shares very nearly to their position in the late Thirties.
2. The available data on the distribution of income by size class, when compared with changes in labor and property income shares, would seem to indicate that increased labor shares are generally associated with reduced concentration in the distribution of income.
3. As economic activity declines, labor shares tend to increase. As economic activity increases, labor shares tend to decline. A decline in labor shares (except for World War II) seems to be accentuated by inflationary conditions.
4. The measurement of income on the basis of claims, rather than on the basis of primary distribution, alters somewhat the relative change in labor and property income since the Twenties, but does not alter the relationship between changes in economic activity and changes in labor's relative share.
5. Where the measurement of income includes capital consumption allowances, relative shares appear to be considerably more stable between prosperity and depression.
The conclusion which is of most general theoretical interest is the third—that increases in economic activity are associated with a reduction in labor shares. This point is particularly significant if a reduction in labor shares is linked to increased concentration in the distribution of income. For this will mean that, if the marginal propensity to consume of property income recipients is lower than the marginal propensity to consume of labor income recipients, increases in economic activity will tend to increase savings ratios and, therefore, increase the volume of investment necessary to maintain the higher level of income. Since inflation appears to accentuate the shift to property income, it may be said that the inflationary process itself makes it more difficult to maintain prevailing levels of national income. Unfortunately, our present knowledge of consumption ratios by income class does not permit a final assessment of the significance of this point. We do not now know the range of variation in the marginal propensity to consume by income class or by type of income.
It is also of considerable interest that the "normal" pattern of relationship between labor and property shares and changes in economic activity was substantially altered during World War II. The increase in economic activity during this period was associated with an increase, not a decrease, in labor's share of total income. A comparison of this experience with the experience of the years 1946-48 and 1950, when labor shares were declining, suggests that the effective stabilization program of the war years may be responsible for the large increases in labor shares of total income in these years.