Three Reforms to Make the Income Tax System Simpler, Fairer, and More Progressive


The main tax filing season is now recently over.  With the frustrations of an absurdly complex system still fresh, it is perhaps a good time to consider reforms which would simplify the system while improving its progressivity.

Conservatives have asserted that the only way to make the tax system simpler is to end its progressivity, with the poor then required to pay the same tax rate as the rich.  This is simply not correct.  The complexity in the system does not comes from the calculation of how much is due in taxes once taxable income is determined.  That is easy to do.  Rather, the complexity comes from the many different rules for determining what is included in and deducted from taxable income, and then taxing different types of income at different rates.  Tax all income, whether from wages or from wealth, at the same progressive rates, and the system would be far simpler as well as fairer.

This blog post will set out three basic reforms that would lead to a simpler, fairer, and more progressive personal income tax system.  It will focus on broader issues, and not go into the specifics of the many special interest loop-holes that clutter the system.  These loop-holes lead to tax breaks for a few while not only forcing everyone else to pay higher taxes to make up for this (in order to meet any given overall revenue need), but also forces us all to determine in each case whether the special provision applies to us (usually not, but you do not know until you investigate the details).  One can find many discussions elsewhere on the pros and cons of these many different special provisions.

Nor will this blog post focus on the corporate income tax system, which is also cluttered with an enormous number of unwarranted special provisions.  But the corporate income tax system raises a very different set of issues.

[Many specifics of the current tax code will be cited below.  Rather than reference each one, the source for all can be found at the IRS web site.]

Reform 1:  Equal Tax Benefits for Deductions at All Income Levels 

Suppose one had a system that led to the following:  There are three individuals, one from a rich household with an income of a half million dollars a year, one from a middle income household with an income of $100,000 a year, and one from a poorer household with an income of just $25,000 a year.  Each decide to give $100 to the same worthy charity.

To the rich household, the US Government says that was very generous, and due to this generosity, the government will give back $40 (more or less – it will depend on a number of special factors, but will be close), so the net cost of the contribution is only $60.  To the middle income household, the US Government also says that the generosity was welcome, but with their lower income compared to the rich household, will give back to them only $25, making the net cost of the contribution $75.  And to the poor household, making only $25,000 a year, the US Government might say that while it was extremely generous to have contributed $100 to that same charity, we will only give back $10 to you, leading to a net cost of $90.

Most would agree that this is enormously unfair.  But it is how our current system works.  Contributions to charities are counted as deductions from income, so the tax subsidy the government provides back to the household will depend on the tax bracket and whether the household itemizes deductions.  It costs the rich household, in a 40% bracket, only $60 to make a $100 donation; it costs the middle income household, in a 25% bracket, only $75 to make a $100 donation to the exact same charity; and it costs the poor household $90 to make a $100 donation to that charity.  (This assumes, for the sake of illustration, that each of the households are itemizing their deductions, rather than taking the standard deduction.)

Most people understand and recognize this.  It is perverse, but many believe there is no simple way around it.  But in fact there is.  Instead of making such payments a deduction from household income, on which one then determines the tax, one can instead make a percentage of such payments a tax credit, to be subtracted from taxes due household incomes without first taking deductions.

One would do this for all deductions as well as personal exemptions (credits for the number of household members, which will be $3,900 per person in 2013).  For all expenses that one determines should remain as deductions (they should in fact be simplified, but that is a separate issue), one would add up the expenses as one does now on Schedule A, which would not need to be changed.  There would then be a tax credit rate (20% perhaps, but the same for all households, rich or poor) applied to the sum of the former deductions and personal exemptions.  One would determine taxes due on the full income (what the IRS calls the adjusted gross income) with no subtraction for deductions or exemptions, based on the standard tables, and then subtract as a tax credit 20% of the total former deductions and exemptions.  And as is the case now, one could have the equivalent of the standard deduction by applying the 20% rate (or whatever rate is chosen) to the current standard deduction amounts and apply it as a tax credit.

If someone had been in a 20% tax bracket before, then their tax would remain unchanged.  But now all households would receive the same tax benefit from a $100 contribution to some charity regardless of their tax bracket.  Poorer households would receive more (relative to what they are receiving now) while richer households would receive less.  This would not only be fairer, but also simpler and more progressive than is the case under the current system.

Reform 2:  Equal Tax Rates on All Forms of Income

Under the current income tax system in the US, the tax rate you pay depends not only on the level of your income, but also on the type of income.  For so-called “ordinary income”, the tax rate can be as high as 39.6% in 2013 (and in fact somewhat higher, as will be discussed below).  Such ordinary income includes wages, interest income from savings accounts or from long-term bonds, income received as rents, and income from capital gains on sales of assets held less than one year.  But if the exact same asset is held for a day more than one year, the tax rate on any captial gains from its sale drops to only 15%.  Dividends may be taxed at ordinary income rates or at the long-term capital gains rates depending on whether the stock paying the dividend had been held for at least 60 days during a period straddling when the dividend was paid (unless it was a dividend on a preferred stock, in which case the requirement is 90 days).  And for those with lower incomes, the long-term capital gains tax rate might be 0%.

There are additional complications on top.  Different rates apply on “ordinary income” if the household is subject to the Alternative Minimum Tax, with tax rates then (on a different definition of what is taxable income) of 26 or 28% (depending on income).  The long-term capital gains tax rate is 15% on many assets, but is 28% on “Section 1202 qualified small business stock”, 28% on “collectibles” (coins or art, defined somehow), and 25% on “unrecaptured Section 1250 gain from selling Section 1250 real property” (whatever that is).  Even once one determines whether any of these apply to your own situation (normally not), one must still navigate tax forms that have to be complex to allow for each of these options.

And in addition, starting in 2013 there will be complications from a number of additional tax provisions.  These are in effect higher tax rates, and the revenues they will raise could have been achieved by modestly higher tax rates, but since Congress did not want to appear to be approving higher tax rates, they instead “hid” them.  The most clear-cut was a decision to phase out the tax benefits for personal exemptions and itemized deductions for high income households (those with incomes of over $300,000 if married filing jointly).  In effect, this will mean higher effective tax rates for the moderately rich, which will then fall back to lower rates for the extremely rich.  There will also now be a Net Investment Income Tax of 3.8% on investment income of all kinds (including short and long term capital gains, interest, dividends, and more), but only on income on the lesser of net investment income or some “modified adjusted gross income” that is in excess of a some threshold amount.  And there is now a additional 0.9% tax for Medicare that applies to wages and self-employment income above a threshold amount (of $250,000 for those married filing jointly).

This is absurdly complex.  But perhaps even more importantly, these widely varying tax rates on different forms of income create an incentive for firms and individuals to shift the incomes being paid to categories subject to lower tax rates.  This mostly benefits the well-connected insiders of firms.  The biggest gap, and hence the biggest incentive, is to shift what are in fact wages to forms of income that are taxed at the low long-term capital gains rate.  One has therefore seen a shift in recent decades to payments in the form of stock options which, if structured correctly, are subject only to tax at the long-term capital gains rate.

An even more egregious example is the taxation of “carried interest”, which is the payment of the fees to fund managers that are linked to the success of the investment.  The fund managers pay tax at only the low long-term capital gains rates on such fees.  Lawyers or accountants working on the same deals, who might also be paid based on the success or not of the deals, receive income from such deals that is taxed as any wages are.  And these fund managers can earn a lot.  Forbes has estimated that one such manager personally earned $2.2 billion in 2012 alone, while four more earned over $1 billion each in 2012.  How can one justify that such managers pay tax at a rate of only 15% on such earnings, while households with ordinary incomes of $72,500 or more pay from 25% up to over 40%.

The basic principle of taxation should be that a dollar of income is taxed the same whether that dollar is earned as wages, or as capital gains (long or short term), or as interest, or as rental income, or as dividends, or as carried interest, or as whatever.  Widely different rates are not only not fair, but add needless complexity as well as incentives to manipulate the system by shifting incomes from one category to another.

The one adjustment that should be made is for inflation, as one is seeking to treat all dollars in terms of the same prices (that of the current year).  This affects capital gains.  Gains that arise only due to inflation do not reflect any real gain in income.  To illustrate, suppose one purchased a house ten years ago for $100,000.  Suppose inflation averaged 3% a year over those ten years.  If the house were now sold for $134,400, one would be receiving back cash that is the same value now as $100,000 was ten years ago.  There would be no real gain.

Some have argued that the low long-term capital gains tax rate of 15% is warranted to compensate for such inflation.  But if so, it is an absurd way to do this.  In the example above, one would still pay capital gains tax on the $34,400 nominal gain.  And one would pay this same capital gains tax if the asset had been held for a day more than a year, or for ten years, or for thirty years.  Yet the cumulative impact of inflation would be widely different depending on the length of time the asset was held.

Much more rational would be to adjust for inflation directly, rather than assign different tax rates for capital gains than for ordinary income.  It would not be difficult to do.  The IRS would publish each year a table showing for each year (going back 100 years if one wants) the adjustment factor to be applied to the cost basis of the asset that had been sold in the current tax year.  In the house example described above, with inflation averaging 3% a year for the past ten years, the table would indicate that the cost basis of an asset purchased ten years ago should be marked up by a factor of 1.344.  That is, the cost basis  of the house sold now would be $134,400 (in terms of current year dollars).  If the house now sold for $150,000 say, the capital gain (in current dollars) would be $15,600, which would be added to current year wages and all other current year income, and taxed at ordinary income tax rates.

A similar inflation adjustment would be made for interest earned.  Suppose inflation is 2% this year, and you have money in a savings account earning 3% a year.  The real return is only 1%, and taxes (at ordinary income tax rates) should be charged only on this 1% return, not (as now) on the 3%.  A way to see this is to note that in earning a 3% return, one has lent to the bank $100, say, last year, and then received $103 back this year.  But that $100 from last year is now worth $102 this year in today’s dollars, due to the 2% inflation, so one is receiving a net return of just $1 in today’s dollars.

With such adjustments for inflation, which would put all sources of income into terms of today’s dollars, there is then no reason to tax different categories of incomes at different rates.  The special long-term capital gains tax rate of 15% should be ended (as should the 28% and 25% rates on other forms of capital gains).  There is also no reason to create hidden increases in marginal tax rates by introducing phase-outs of exemptions and deductions.  Reform 1, discussed above, presents a fair and simple way on how to handle exemptions and deductions.  Finally, there is no logic in having an Alternative Minimum Tax.  If there are loop-holes that lead rich people to have a legally low level of taxable income, one should get rid of the loop-holes.

Reform 3:  Ensure All Sources of Income Are Taxed Similarly Over a Life-Time 

Even with equalized tax rates on all forms of income (as well as equalized tax credits for deductions and exemptions, regardless of income level), taxes are still assessed only on incomes that can be readily measured each year.  Thus taxes are assessed on wages, dividends, and interest, and on capital gains when assets are sold.  If the assets are not sold, the tax system does not assess taxes on what the gains might have been from an increase in the market value of the assets.  Such gains accumulate tax free until the assets are sold.  And if they are never sold, but instead are passed along to heirs at the time of death, the heirs inherit the asset at a cost basis equal to the value on the date it passed to them.

Furthermore, and as was noted in a column by Larry Summers (former Treasury Secretary and Obama advisor), the really wealthy have developed many mechanisms to receive income and accumulate wealth over their lifetimes and even upon death, without ever paying significant taxes on what can be truly massive accumulations.  The accumulation of unrealized capital gains is only one such mechanism.  And it would require a specialized lawyer (which I am not) to explain many of the tricks.  But in recent decades one has seen massive accumulations of wealth with apparently limited tax consequences.

The principle to follow should be that all income and earnings from accumulation of wealth should be taxed equally at at least one point in a person’s lifetime, or upon death.  Income is taxed generally as it is earned, whether as wages, interest, dividends, rents, or other forms.  Capital gains are taxed if assets have been sold.  All other wealth, in whatever form, should be taxed upon death according to the capital gains that would have been generated based on their valuations on the date of death.

The estate tax (combined with the unified gift tax) can be seen as a poorly designed substitute that tries to make up for this.  While total estate assets are taxed at a relatively high rate (reaching 35% currently), and the tax is based on the gross value of assets upon death and not on the unrealized capital gains on such assets at that time, no estate taxes are due at all until estates reach a sizeable $5,250,000 (as of 2013, and this can effectively be doubled for a married couple though some straightforward legal measures), with the 35% rate applying only to estates over $5,750,000 (which can also effectively be doubled for a married couple).

Both through the high thresholds before any estate taxes are due, and through other mechanisms, little is actually paid in estate and gift taxes each year.  As Larry Summers noted in the column cited above, only about $12 billion in such taxes are paid each year.  A conservative guess would be that at least $1.2 trillion is passed down in inheritances each year, the vast bulk of which is concentrated among a wealthy few.  Thus current estate and gift taxes are generating taxes equal to only 1% of this, and the bulk of estates are being passed along with small or no taxes being paid.

The reform would be to have much lower thresholds (perhaps $1 million per married couple, meaning $0.5 million per person, with unlimited spousal inheritance if the couple so chooses), but taxes then based only on the capital gain as of the date of death instead of the gross value of the assets at that time.  Ordinary income tax rates would be applied for estates above these limits.

It has been argued that those gaining the inheritance would not know the cost basis, but this can be easily remedied.  For almost all households except the very rich, the assets to be passed along are either in financial accounts (where the financial firm holding the asset will know the cost basis, which is now being reported to the IRS when assets are sold), or are homes.  The cost basis on homes (and indeed other major property) can be recorded in an annex to the will.  And as was proposed in Reform #2 above, the cost basis would then be adjusted up to reflect inflation between when the asset was first purchased and the date of death.  For the very rich with a diversity of highly valuable assets, the cost basis could be recorded as part of the will or other document in which inheritance decisions are recorded.

IRS enforcement would focus on the extremely large estates.  These are where there would be major losses from underreporting of the value at the time of death of the assets being transferred.  As under the current system, such assets should be reported at their fair market value, but sophisticated operators have often reported much less.  One way to enforce this might be that the value of very expensive assets (say those over $10 million) would have to be publicly declared, where anyone would be able to bid for those assets at a price of at least 10% above the declared price.  A few firms would likely develop that would specialize in valuing such assets and then bidding for those being undervalued.  Such a market mechanism would enforce discipline.

Note also that with this new system, there would no longer need to be the complication of separate gift tax accounting.  Gifts of up to $14,000 (currently) per person per year can be given with no gift tax consequences.  Gifts above this must be recorded and then effectively added to the estate to determine estate tax.  This is needed so that the rich cannot transfer all of their estate to their inheritors tax free prior to their death.  But with an estate tax based on capital gains, one could allow gifts of any amount provided they are made in cash (as there is no capital gain involved).  If assets other than cash are given to inheritors, then one could require that any such gifts above the minimum ($14,000 currently, but it could be adjusted) would be subject to tax at the amount of capital gain involved in that asset, with this tax to be paid by the one providing the gift.  It would be taxed as if the asset were sold, the capital gain tax paid, and then the resulting cash being provided as a gift.


The reforms presented above would lead to a simpler tax system, as well as a fairer and more progressive one.  The focus has been on structure.  Decisions would still need to be made on the specific tax parameters to be applied.  But these would be far fewer in number than is the case under the current system.  In the current system there are different sets of tax rates for ordinary income taxes (which now vary from 10% to 39.6%), for capital gains (15%, 25%, 28%, or the ordinary income tax rates if short term), for the Alternative Minimum Tax (26% and 28%), for estate taxes (0% to 35%), for the new Net Investment Income Tax (3.8%), for phase out rates for deductions and exemptions, and for more, as well as the multiple floors and ceilings which apply to each rate.  In the new system, one would have one set of ordinary income tax rates (and the income ranges to which they would apply), and one rate of uniform tax credit to be applied to what are now tax deductions and personal exemptions.  There would also be thresholds to determine for the new Estate Tax.

A determination would need to be made on what the new income tax rates, tax credit rate, and thresholds would need to be.  This would require data and models that I do not have, but which are readily available at the US Treasury, at the Congress (Joint Committee on Taxation), and at the private non-profit Tax Policy Center.

But I would note that there is no logic in requiring that such a tax reform be revenue neutral, generating only as much revenue as the current convoluted system currently leads to.  Rather, the tax parameters should be set so that over the course of the full business cycle, the tax revenues generated and the government expenditures that are warranted lead to a path of deficits and surpluses such that the public debt to GDP ratio follows some desired path.

Note that what makes no sense is to pre-specify (as was been done in the original Bowles-Simpson proposal, or the budget proposals of Congressman Paul Ryan) that federal tax revenues should be set to some level or share of GDP, with expenditures then cut to whatever extent necessary to match those revenues.  Under such an approach, even Congressman Ryan as well as Bowles and Simpson would agree that it would be nonsense to spend more if revenues surprisingly turned out to be higher than anticipated.  It would be equally nonsensical to say expenditures should necessarily be cut, no matter how worthwhile and efficient the program, if revenues turn out to less than anticipated.

Rather, for long-term budgetary purposes, a determination should first be made of the level of government expenditures that are warranted and worthwhile, and reflect an efficient use of resources.  Revenue targets should then be set to meet these goals.  The specific tax rates of any system, including the one proposed here, would then be determined to reflect this.