US Source Taxation of Dividends: Rational?

In discussing the Section 871(m) regulations, I asked the broader question as to why we impose withholding on dividends?  See discussion at the end of my post here.  After all, if there was no US source withholding on dividends, there would obviously be no need for Section 871(m) and the regulations in the first place.

US persons and the Economics and Taxation of Dividends.  Let’s approach this by first reviewing the taxation of dividend payments to US persons.  It obviously won’t surprise you that, very generally, dividends are subject to US net income tax.

So cash dividends must be income right?  Not so fast.  Let’s look at an example.

Let’s say you bought stock for $100 prior to its ex-dividend date and that you hold the stock through the dividend payment date.  Let’s also imagine that other than the dividend, the stock’s value would hold steady at $100.  After the ex-dividend date, you would expect the stock value to decrease to $95, an amount equal to the old value of the stock less the dividend.*  You’d eventually receive a dividend payment of $5 and you’d retain stock worth $95, a total value of $100.

Let’s review the tape:

  • Before purchasing the stock, you had $100 of cash.
  • After purchasing the stock (but before the ex-date), you had $100 of stock
  • After purchasing the stock (and after the ex-date), you had $95 of stock and a $5 dividend receivable.
  • After purchasing the stock (and after the dividend payment date), you had $95 of stock and $5 cash.

Throughout the entire timeline, you retained a net value of $100.  Okay, so where’s the income?  Well, in an economic sense there is no income.  At this point, we may be tempted to dive into the rabbit hole of “what is income?”  However, I don’t intend to do that. Maybe later we can think about that in the context of the 16th Amendment and cases like Eisner v. Macomber, 252 U.S. 189 (1920)

At this point, I’m more interested in asking whether taxing a dividend is “fair” when the US person has no income in the economic sense.  I know “fair” is a loaded term but we can perhaps agree that it is unfair to tax someone if their economic situation is unchanged.  But, since we already know that the dividend will generally be subject to tax, can we conclude that this is completely unfair.  Well, not completely.

In the above scenario, the investor would generally have to report $5 of income associated with the dividend on their tax return.  However, this is offset by a $5 unrealized loss associated with their stock, which is the difference in their purchase price ($100) less the post-dividend value of the stock ($95).  So, this appears fair to the US investor in the aggregate because the income and the loss are equal.  At least in the aggregate, this reflects the fact that the US investors economic position has not changed.  Of course, this ignores the potential timing and character differences in the tax treatment of the dividend income and the loss on the disposition of the shares.  Notwithstanding that giant caveat, at least the treatment of the US taxpayer is fair in the sense that it will ultimately recognize a $5 loss or reduction in gain.

Non-US persons and the Economics and Taxation of Dividends.  Are non-US persons similarly treated?  Generally no, unless their income is effectively connected (in which case they are generally treated similarly to US persons).

A dividend from US sources (e.g., paid by a US corporation) is subject to US source withholding at a 30% rate under Section 871(a) of the Code.  In some cases, this rate can be reduced by a treaty.  US source withholding is a gross basis tax.  That is, a non-US investor generally cannot take any deductions, related or otherwise, against the amount of the dividend in order to reduce the amount of the tax.  One theory for this form of taxation is that it is impossible administratively to calculate the deduction of the recipient that net-based taxation would require.

Therefore, in the above example, the non-US investor would be subject to a 30% tax on the dividend received in the absence of a treaty.  But what about the resulting capital loss?  From a US tax perspective, it’s effectively useless.  Capital gains and losses of non-US persons are generally not US sourced.  Moreover, as indicated above, even capital losses that result from the dividend cannot be used to reduce the amount of the dividend.

Based on this example, the treatment of non-US persons is not “fair” in the same sense that the treatment of US persons is “fair”.  The US tax rules eventually get around to treating the US person as if they have no net economic income, but treat the non-US person as having income of $5 without the benefit of the corresponding loss on their shares.

Does this make sense?  Not to me—at least not from a purely economic point-of-view.  On this point, I’m going to keep it simple even though the issue is, without a doubt, more complex.

With that in mind, one argument for a 30% withholding tax on dividends is that the 30% is a stand-in for higher rate that applies to net-based taxation; that is, it is a rate benefit given in lieu of a deduction.  However, in the example above, the dividend is a 1-for-1 match in the reduction in the value of the stock.  Therefore, the argument that the 30% withholding rate is meant as a substitute for what would otherwise be a permissible deduction is unsatisfying at best.

Of course, the next possible argument is that capital gains are generally not subject to US source taxation except in certain cases, such as effectively connected capital gains and real estate gains.  Therefore, any reduction in value of the stock as a result of the payment of the dividend is effectively deducted anyway because the gain is not taxed in the first place.  However, this logic fails in two respects.  First, it doesn’t hold true when there is an ultimate loss with respect to the stock since that would afford the non-US person no US tax benefit.   Second, it’s a haphazard justice at best since high dividend paying stocks would penalize non-US persons more than low or no dividend paying stocks.

Another argument is that this is simply a reflection of the apparent historical bias for retention rather than distribution of earnings, similar to the special rate that has historically applied to capital gains in the US context.  As to this last point, however, it should be noted that the treatment of dividends and capital gains in the domestic context has moved much closer with the application of the capital gains rate to the certain qualifying dividends. In most cases, the rate is equal.  It seems weird that this bias should continue to exist for non-US persons when it’s being ameliorated in the domestic context.

Does this mean anything?  Depends I guess.  Personally, I don’t think there should be any real difference in the treatment of dividends and capital gains.  I think making the treatment consistent would better reflect the actual economic relationship between the dividend and the value of the stock, it would simplify the tax code in general, and it would make irrelevant any arbitrages that may exist where the two are subject to different rates.  Now, this doesn’t necessarily mean that I think that dividends and capital gains should both receive a preferential rate; rather, that they should be treated the same whether it be the same net rate or same withholding rate.

On a last note, I wanted to tie this into my earlier post about Section 871(m).  Section 871(m) and the associated regulations are effectively applying the same sorts of withholding rules to derivatives that apply to US stock (albeit in a more complicated way because of Delta and the other Greeks).  But the same sorts of economics discussed above apply to derivatives.

Consequently, Section 871(m) can be seen as doubling down on the same conceptual error that is already inherent in dividend withholding.  Of course, this ignores the underlying purpose of Section 871(m) which is anti-avoidance.  To the extent that Section 871(m) and the regulations are properly tailored to address avoidance of dividend withholding on stock, then the new regime would apparently make sense.  To the extent that it goes beyond that and is more akin to establishing a method of international taxation of derivatives, it seems a backward step in applying a rationale system of taxation to a class of assets.

Other things that are boring.  Of course, as I mentioned above, my analysis is fairly simplistic.  There are many more scholarly analysis of US international taxation out there.  I read portion of them and, quite honestly, found them rather boring.  Not everything can be as exciting as this stuff.


Many of those discussions revolve around capital export neutrality, capital import neutrality, and national neutrality.  If you’d like to look at this more in-depth, a good place to start is in this Joint Committee report here.  Or you can just google one or more of the neutralities above and find a bunch of stuff online.

In order to whet your appetite, capital export neutrality describes a tax system where an investor residing in a particular locality can locate investment anywhere in the world and pay the same tax.  If capital export neutrality is achieved, the tax system neither encourages or discourages capital export (ie., the tax system does not motivate an investor to locate their investments in any particular jurisdiction).  The goal of this tax system is capital efficiency.

Capital import neutrality describes a tax system where an investment located in each country is taxed at the same rate regardless of the residence of the investor.  This arguably improves “competitiveness” as a non-domicile is taxed at the same rate as the domicile.

The US tax system seems to be some sort of a mish-mosh of the above.  In general, the taxation of US domiciles looks to be largely capital export neutral given its worldwide system of taxation and foreign tax credits.  This capital export neutrality seems to also spill over into the treatment of non-US persons who are engaged in a US trade or business.  I find it a bit more difficult to identify where US source taxation fits on this spectrum, however.  Anyway, if you want to explore this topic more, get to it!

*For this purpose, I’m ignoring more complex theories of valuation of stock.  I’m keeping this simple and primitive—your share of $5 of cash inside the company is worth the same as $5 of cash in your hand.  This may not be true, but it’s close enough for this blog.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s