earlier theories. It is sufficient to make clear the fact that the theory of today is an outgrowth and development of earlier theories. Doubtless this marginal theory will prove but a stepping-stone to a more useful and penetrating wage analysis in a later period. Indeed, severe assaults have already been made upon marginal principles, and many authorities believe that they are already in a process of disintegration.

Marginal theory involves an application of the marginal concept to various factors in the wage contract. First, we may apply it to the laborers themselves. Marginal laborers tend to set the price below which wages cannot fall for all laborers. Marginal laborers may be viewed either as the least willing or as the least effective laborers. In all branches of industry are found laborers who are barely persuaded to stay in the industry by the wage rate they receive. If any further cut were made, these dissatisfied, reluctant, marginal laborers would detach themselves from the industry. Wage cuts drive out marginal laborers, and at the lowered rates, laborers who are marginal at lower levels appear. Willingness to work at a certain rate is one thing; ability to work effectively is quite another thing. Everybody is willing to work in jobs paying the most money and operating under most attractive conditions. But only a limited few are capable and qualified. The workers in any grade or occupation may be divided into descending scales of excellent, good, fair, mediocre, poor, and useless. The wage rate of the group will not be based upon the efficiency of the excellent, but upon the efficiency of the fair, mediocre, or even the poor. The least effective workers in the factory, the mill, and the mine tend to set the wage which must necessarily be paid to the more effective workers. Marginal workers are the least effective and the least willing workers in any trade and in all trades.

It should be noted carefully that the point where marginality falls depends closely upon the scarcity or plenty of laborers. If a huge supply of laborers is present, because of immigration or high birth rate, then the marginal laborer will be one who cannot afford to be unwilling to take almost any wage offered him. The marginal point may then be battered down to the bare level of subsistence. If labor supply is restricted, the scarcity of laborers will enable the least willing and least effective workers to hold out for high rates of pay. The marginal factor is fundamentally a scarcity factor. Marginal elements are supply elements. How far down the scale we will find the least effective and least willing laborers depends upon the supply of laborers at the given time and place. Scarcity is a basic governing factor in marginality of labor.

Secondly, the marginal principle applies to the status of the entrepreneur or employer. The marginal employer may be viewed both as the least willing employer and the least effective employer. In general, the least willing employers are those who expect to gain least from employing labor. They are least willing because they are least successful in making a profit. They are least willing to pay a given rate to labor because they can least afford to pay a given rate. In other words they

are least willing because they are least effective. Employers who just barely cover expenses are the employers who represent the wage rate which all employers must pay. More successful employers could afford to pay more, but they do not have to, since the marginal employers influence the wages of all employers. The marginal producer is the producer at whose hands wage rates are fixed. What he is able to pay is the high point in wage rates. Beyond that point it is unnecessary for better employers to go.

Marginal employees and marginal employers represent the battle lines where wage rates are won and lost. If wages are cut too low, marginal employees will quit the employment. If wages are forced too high, marginal employers will be driven out of business. These marginal factors cannot be adequately interpreted, however, without introducing a third marginal factor, namely, marginal productivity. The product of labor has a fundamental influence upon the price paid for labor.

The marginal productivity of labor is the productivity of the least effective units of labor in each given employment. The lowest units of labor add the least product. Any employer finds all degrees of efficiency among each kind of his employees. Let us assume that he grades each kind in the scale of their individual efficiency from highest to lowest. The productivity of each man in the scale is what the employer would lose by not employing him. Or we may say that the productivity of each man is what he adds to the product of the establishment. His product is the amount by which he can increase the output for the employer. The test of what each workman adds to product is what would be lost to product by his withdrawal from the work. If we take away one unit of labor, what diminution of product occurs? If we add one unit of labor, what increment of product occurs? The answers to these questions are the tests of the productivity of each unit of labor. Now, as we go down the scale of decreasing efficiency, we find that successive laborers turn out less and less additional product. And eventually we come to those units of labor whose product is barely enough to cover the expenses of turning out the product. When labor productivity is so low that product equals cost, and net gains are zero, the point of marginal productivity has been reached.

Any further employment of labor would then result in a net loss to the employer. Any less employment of labor would mean that the employer was not making the largest gain possible. Employers will tend to employ successive units of labor, even though efficiency is decreasing, until the point is reached where labor productivity is barely sufficient to cover expense. This tendency is only a tendency. It is at best imperfectly realized. It requires an assumed state of free and perfect competition. It is hypothetical. But as a tendency, nevertheless, it serves to fasten attention upon a fundamental force which is operative in wage determination. It emphasizes the direction in which a major influence is moving. Marginal productivity is the goal and resting point toward which wages tend in each plant, each branch of industry,

and in each national organization of industry taken as a whole. Every employer will seek to employ additional labor as long as it adds more value to the product than the cost of the labor. As long as the employer gets more for the product than he pays for it by adding more labor it is to his advantage to increase the labor factor. The marginal product of labor will be the dividing line beyond which no profit can be made from additional employment of labor.

The rate which is warranted by the marginal product of labor is also the rate which is paid to all laborers in the same group. In discussing the value of commodities, we emphasized the fact that marginal utility sets the price of all units. All buyers get the commodity for the price which the marginal buyer is willing to pay. Likewise in the case of labor, the wage which is put upon the labor producing marginal product is the wage for all labor in the same group. The marginal rate is the standard rate. Each unit of labor is worth to the employer what the last unit produces. This is in accord with the law of one price in one market. The law of one price has already been explained with reference to the price of commodities, and it may now be applied to the price of labor. The first units of labor, although more productive if they alone were used, bring the same pay as the last units of labor in the group. First units, middle units, all units in the group are priced at the same rate as the last or marginal units. Uniformity of wages at the level of marginal product for all laborers doing the same kind of work is the essence of the process of wage fixing.

The marginal principle is difficult to visualize with labor as a lump mass, but it is easy to visualize when labor is viewed in different grades. At the bottom of the labor scale is common labor of the humblest kind. At the top of the labor scale are high salaried managers and professional men. Between these lower and upper extremes are all grades and degrees of superior common labor, semi-skilled labor, and highly skilled labor. The men in the topmost grade receive high pay. But the least willing and least effective workers in that grade are on the borderline between staying in and getting out. They are at the margin of indifference between that grade and the next lower grade of labor. Some will be driven below the margin of indifference into the grade of highly skilled

same.

3 It should be noted that even though the successive units of labor be equal in willingness to work and ability to work, their product will not necessarily be the As successive units of labor of the same strength and intelligence are drawn into use, the additions of effort will gradually yield less and less product. Decreasing productivity of additional units of labor effort may be due to diminishing returns because of size of management or limitation of other factors. But more especially decreasing productivity of additional units of effort may be due to the fact that the added supply of the product may so alter the relation between supply and demand as to depress the price and exchange value severely. Additional units of effort yield increased total supply. Increased supply tends to depress prices. As the price of the product falls, the product of each additional unit of labor is worth less and less. The fall in the value of additional product is not due to any slacking in effort on the part of the laborer. He may be working harder than ever before, but his additional product will be worth less and less because increased supply has brought prices down.

workmen, of foremen, and the like. These workmen will be individuals who could not add sufficient product in the topmost grade to warrant their retention in that grade. But in the next lower grade, a similar development will also be taking place. Some of the skilled workmen will be ineffective and unwilling. They will not be able to add enough product to warrant their retention in that grade. Hence, they will be found at the borderline, or margin of indifference, for that grade. But as the sub-marginal skilled workers are driven out of the skilled grade, they will pass down into semi-skilled or unskilled grades. And there again will be found workmen at the margin of indifference. And finally will come the margin of indifference of the lowest grade of common labor. Now, men who are on the margin of indifference here have only the alternative of quitting work altogether. There is no other task to which they can turn their hands. If wages are cut so low as to drive them out of their jobs, they may emigrate from the country, or join the ranks of the unemployed, or restrict the size of their families, or die off, or start a revolution, or organize a labor union. But in any event, they are sub-marginal even for the lowest grade of labor. Their added product, if employed, would not be enough to cover the added cost of employing them.

These grades of labor have been termed "non-competing" groups. In a broad and general sense, they are non-competing. The common laborer does not compete with the manager for the same position. The unskilled does not compete with the skilled. Members within each group compete with each other, but the group as a whole does not compete with other groups as a whole.

The inability to compete, moreover, shows a pronounced tendency to persist. Even if wages in higher groups become very lucrative, it is with exceeding sluggishness that lower groups qualify themselves for the higher work in order to obtain the higher pay. The higher groups remain higher and the lower groups remain lower. The economist expresses this sluggish tendency by stating that labor groups are highly immobile. Their inertia by groups is immobility. If demand for labor in one group becomes intense, nevertheless the inertia of the lower groups will slow up the tendency for the new demand higher up to be satisfied. The immobility of non-competing groups accounts in large measure for the differences of wages between the groups. If members of one group could pass over with utmost ease into the ranks of a higher group, wages would be equalized throughout all occupations. But differences of wages persist, and the reason for their persistence is the immobility of the non-competing grades of labor.

We may ask, furthermore, why the grades show this quality of immobility? In part, the reason lies in differences of inborn abilities. Native capacity differs sharply as between the brilliant executive and the sodden laborer, between the sensitive skill of the mechanic and the crude ignorance of the lowest classes of labor. The inequality of inborn capacities is inferred by scientific tests of intellectual, emotional, and

physical qualities as well as by common observation. Only the rare individual shows the natural qualities which make for the highest success. But in part, the reason lies beyond native capacities. It lies in the stratification of opportunity which appears in society. The children of the common laborer's family are likely to be put to work at an early age, to be denied an education, to be exposed to a social environment which kills ambition and stifles aspiration. The children of the higher grades of workers have every opportunity to make the most of their talents. They are encouraged, inspired, educated. Doubtless the lower grades of labor possess an immense store of untapped ability. Their latent capacities have never been released because their economic and social environment has never been favorable. The smouldering and dormant powers in the common man are doubtless very great, but social stratification makes the release of these powers very slow and tedious indeed.

We may add a third factor to the two that have already been mentioned as accounting for differences in wages. In addition to inequalities in inborn abilities and inequalities in opportunity and environment, we may mention inequalities in scarcity of the given grades of population. The higher grades of labor show the smallest birth rate. They have small families. Consequently, there is a marked scarcity of the children of the high paid grades of labor. Population restriction limits labor supply in those grades, and enables them to keep wages up. Contrast with this the condition of the families of low paid workers. High birth rates, large families, over-supply of low paid labor, is the sequence of events. Population multiplies, and scarcity of labor is destroyed. When scarcity is undermined, wage rates are undermined. Inequalities of birth rate and family increase are fully as important as inequalities of native ability or of social opportunity. These inequalities among the grades and groups of labor largely account for the inequalities in wages paid the various groups. Speaking in marginal terms, these underlying factors determine the marginal product of labor in the given groups and grades. Marginal product, as previously explained, determines the wage rate for each kind of labor.

Inequalities of native ability, of social opportunity, of family increase, account for inequalities in marginal productivity, which in turn govern the inequalities of wage rates among the various groups.

The Dependence of Wages upon the Other Factors in Production.— The productivity of labor depends upon the use made of the other factors in production. Labor well supplied with efficient capital can produce much more than labor with scant capital. Not only the amount of capital but the technological quality of the capital influences the product which will be made from a unit of labor. The marginal product of labor tends to be highest when a good quantity of the best capital is available for use. Likewise, labor productivity depends upon the land factor in production. The same labor performed upon two different land areas, the one fertile, the other inferior, will result in sharply