a dizzy height and might have a great fall, and yet, after the stock dividend, would feel quite safe in buying two shares of the stock at $110 a share.

Even if, however, the market value of a stockholder's holdings is increased by a stock dividend, the government did not purport to measure the tax by this increase in market value. The tax was not a tax on the difference between the market value of a stockholder's holdings before and after the payment of the dividend; under the Revenue Act of 1916 it was a tax on the whole cash value of the stock issued as a dividend. Thus if we assume that the stock sold, before the stock dividend, at $200 a share, and that it sold, after the stock dividend, at $110 a share, the stockholder who had one hundred shares of a total market value of $20,000 and now has two hundred shares of a total market value of $22,000, would be subjected to a tax, not on $2000, but on $11,000. These considerations of market value, therefore, are not a legitimate basis for the tax which Congress imposed.

Assume that, in a particular case, there was no increase even in the market value of a stockholder's holdings (and this was the fact in Eisner v. Macomber). This brings us back to the argument that the stock dividend is not income, because its payment does not cause an increase in the stockholder's wealth.

Reflection will show that there is nothing in this argument.

A taxpayer is employed by a solvent corporation which promptly pays its debts. He receives his weekly wage of $50 at one o'clock on each Saturday, when his week's work is done. He is not $50 richer at one o'clock than he was at 12:59. He has been growing richer throughout the week as he was earning the $50. When he received the $50 he exchanged a right for cash. Possibly the cash is worth a little more than the right—it is comforting to have the cash in hand- but the difference in the case supposed is very slight. No one would claim that the government may tax as income only the difference between the $50 and the value of the right at 12:59.

Take the case of a cash dividend. If the corporation with a capital of $500,000 and a surplus of $500,000 pays a cash dividend of $100 a share, the book value of the shares drops from $200 to $100 a share. Instead of one hundred shares of a total book value of $20,000, the stockholder now has $10,000 cash, and one hundred shares of a total book value of $10,000. His wealth has not been

increased by $10,000, and yet no one disputes but that the Government may tax the whole $10,000 cash dividend as income.

There is one case in which a person's wealth is increased by the amount which he receives. That is when something is given to him. If $1000 is given to him, he is worth $1000 more the instant it is received than he was the instant before the receipt. But Congress itself has expressly declared that gifts are not income.

It is not profitable to multiply examples. It may well be that a stock dividend is income, although the taxpayer's wealth is not increased at the moment of its receipt.

In Pollock v. Farmers' Loan & Trust Co. the court held, inter alia, that taxes upon returns from investments of personal property were direct taxes and that Congress could not impose such taxes without apportioning them among the states according to population, as required by Article I, section 2, clause 3, and section 9, clause 4, of the Constitution. Thereafter the Sixteenth Amendment was passed providing that "The Congress shall have power to lay and collect taxes on incomes, from whatever source derived, without apportionment among the several states, and without regard to any census or enumeration." In the Revenue Act of 1916 Congress declared that "the term 'dividends' as used in this title shall be held to mean any distribution made or ordered to be made by a corporation . . . out of its earnings or profits accrued since March first, nineteen hundred and thirteen, and payable to its shareholders, whether in cash or in stock of the corporation which stock dividend shall be considered income, to the amount of its cash value."3 In Eisner v. Macomber the Supreme Court decided, four judges dissenting, that Congress could not constitutionally tax stock dividends as incomes. Mr. Justice Pitney delivered the opinion of the majority. Mr. Justice Brandeis delivered the principal dissenting opinion.

What is the proper mode of approach for the Supreme Court in dealing with the constitutionality of an act of Congress? We do not open any discussion of that great question. If the question were new, it might be urged that the court should interpret the Constitutional provisions to the best of its ability, uninfluenced by any interpretation which Congress may have made, and should

sustain or reject acts of Congress according as they conformed or did not conform to the Constitution so judicially interpreted. But the question is not a new question; the Supreme Court has repeatedly stated that, since it is dealing with the act of a coördinate branch of the government, it will pay great respect to the legislative interpretation of the Constitution, and that no act of Congress will be declared unconstitutional if it is consistent with any reasonable interpretation of the Constitution. Our question therefore becomes this: Was this provision in the Revenue Act of 1916 consistent with any reasonable interpretation of the Constitutional provisions mentioned above?

We should note at once the different methods of taxation which Congress has employed and now employs in taxing trusts, partnerships, and corporations. A trustee takes in the trust income, and distributes the net income to beneficiaries. Here are two acts, the receipt of income by the trustee, and the receipt of the net income by the beneficiaries from the trustee. Conceivably Congress might have enacted that the trustee should be taxed upon his receipt of income, and then that the beneficiaries should be taxed upon their receipt of net income from the trustee. In a business sense, this would obviously be double taxation, and Congress has not seen fit to make such enactment. As the law now stands, the trustee (under the kind of trust most commonly found, where it is the duty of the trustee to distribute income periodically) simply files an information return, no tax is assessed to him upon his receipt of income, but a tax is assessed to the beneficiaries upon their distributive shares of the net income, whether the income has in fact been distributed or not. Similarly with partnerships. The present law provides that "individuals carrying on business in partnership shall be liable for income tax only in their individual capacity. There shall be included in computing the net income of each partner his distributive share, whether distributed or not, of the net income of the partnership for the taxable year." That is, Congress does not treat the partnership as an entity separate from the partners, and tax it upon the receipt of income, and then also tax the members upon their receipt from that entity of their shares of net income. Congress adheres to the common-law conception that a receipt by a partnership is a receipt by the persons who are members of that partnership.

But it is otherwise with a business carried on by a corporation. Here there is double taxation. The corporation is taxed upon its income and then the stockholders are taxed (under the present law, for surtax purposes only) upon their receipt of income from the corporation. The stockholder is a different legal unit from the corporation, and Congress is "at liberty to treat the dividends as coming to him ab extra, and as constituting a part of his income when they came to hand."4 Thus, a dividend paid to a stockholder, even out of earnings which the corporation had made prior to March 1, 1913, may, if Congress sees fit, be treated as taxable income of the stockholder. If that which was part of the corporate earnings (remaining after corporate taxes have been paid) is passed on to the stockholder, the asset received is taxable income to the stockholder.

Note again the words of the Revenue Act of 1916. "The term 'dividends' as used in this title shall be held to mean any distribution made or ordered to be made by a corporation . . . out of its earnings or profits accrued since March first, nineteen hundred and thirteen, and payable to its shareholders, whether in cash or stock of the corporation." 5 (The words of the Revenue Act of 1918 are, "A dividend paid in stock of the corporation shall be considered income to the amount of the earnings or profits distributed.") "

But, when the directors of a corporation pay a stock dividend they do not distribute corporate earnings. No part of the corporate assets passes from the control of the corporation to the control of the stockholder.

The nature of the transaction is the very opposite to a distribution of earnings. Dividends can only be paid out of surplus, they must not be paid out of, and hence impair, the capital. So long as there is a surplus there is a fund out of which dividends may be paid at some future time in the discretion of the directors, and out of which a court may order dividends to be paid, if the directors abuse their discretion. So long as there is a surplus, there is a possibility of dividends. When a stock dividend is paid, that which was surplus becomes capital. A fund available for dividends is converted into a fund not available for dividends. The payment

4 Lynch v. Hornby, 247 U. S. 339, 344 (1918).

of a stock dividend is not only not a present distribution of earnings, but it destroys all possibility of a future distribution therefrom. The payment of the stock dividend does not give to the stockholder as against the corporation any right which he did not have before. On the contrary, the rights of the stockholder are diminished and the rights of the corporation are increased. By the capitalization of surplus, corporate control of the assets is perpetuated. If a court should order directors to distribute corporate earnings, it would be a mockery for the directors to pay a stock dividend.

Mr. Justice Brandeis, however, argued, in effect, as follows: The payment of a stock dividend, in business substance, includes two transactions: (1) the payment of a cash dividend by the corporation to its stockholders, and (2) the return of that cash to the corporation for further stock.

It is submitted that the constitutionality of the provision in question cannot properly be sustained on this ground. A corporation has no authority to pay a cash dividend out of surplus, and make it a condition that the cash be returned in payment for new stock. Once the asset has passed from the control of the corporation to the control of the stockholder, it is wholly for the stockholder to decide what he will do with it. It is true that the corporation may accompany its declaration of a cash dividend with an attractive offer to subscribe to a further issue of stock, but it must permit the stockholder to assign his right to take advantage of this attractive offer. The corporation must really let go of the cash. It may at the same time bid to get the cash back again, but it will be exposed to competition with all other legal units who seek the stockholder's cash. The stockholder, with the cash in hand, is master. He may prefer another investment, or he may feel the need of using the cash to pay his income taxes and debts, or he may spend the cash in the pursuit of happiness.

The financial history of corporations shows that under such circumstances there is usually a lively market in the "rights," and there could be no such market if there were not stockholders who did not choose to return the cash to the corporation. Even if the corporation gets back an amount of cash equivalent to that paid out, it does not at all follow that it is getting it back from the persons to whom it was paid (and sometimes it happens that the