23 June 2006

What Should Estate Tax Reform Look Like?

Sometime, probably in the next two weeks, the U.S. Senate will vote on H.R. 5638, the estate tax reduction bill passed in the U.S. House of Representatives yesterday.

I hope that the bill doesn't get the 60 votes it needs to pass. Outright repeal won 57 votes, and this bill still gives away the farm.

Of course, nothing prevents Democrats, should they secured control again, if H.R. 5638 is passed, from restoring some of the estate tax cuts to raise revenue to help close the defict.

A better proposal is set forth below. I have put my recommendations (in addition to some subheadings) in bold, for those who like to save the analysis for later (if ever).

Towards A Reasonable Compromise

A compromise worth shooting for would be to freeze the estate tax at the levels set forth for 2009 under existing law, with some slight modications.

In 2009, the tax rate will be 45% and the exemption amount will be $3,500,000.
Under current law, the rate jumps to 55% (with a 60% bubble rate) in 2011, with an exemption that falls to $1,000,000. The increase in the exemption from $1,000,000 to $3,500,000 isn't outrageously costly in terms of revenue collected and eliminate the need for large number of people to have to file estate tax returns or do tax planning. This increases revenue relative to compliance costs and reduces the political pressure to eliminate the tax. The rate drop is fairly modest.

One would want to make some minor changes at the same time:

Gift Taxes and Generation Skipping Transfer Taxes

* The gift tax exclusion, currently $1,000,000 and scheduled to remain there indefinitely should be harmonized with the $3,500,000 estate tax exclusion.

* Gift taxes should be calculated on a tax inclusive basis, just as income and estate taxes are now. Right now, while the gift tax rate in 2009 will be a nominal 45%, because of the "tax exclusive" way that it is calculated, this is equivalent to an estate tax rate of 31%. There is no principaled reason for gift and estate tax rates to be different.

* Likewise, the generation skipping transfer tax rate (which covers transfers made to grandchildren directly in an effort to avoid estate taxes in the children's estates), which is now identical to the gift tax rate, should also be made the same as the gift tax rate.

* The generation skipping transfer tax exemption, unlike the gift and estate tax exemption, does not need to be increased. While the gift and estate tax exemptions are integrated in a way that makes unifying them into a single exemption useful, the generation skipping transfer tax exemption is an entirely different beast which has no linkage to the other exemptions.

Current law creates a strong incentive, especially in large estates, to make large gifts during life, and to make large gifts to grandchildren, even when they don't really want to do either, in order to benefit from the favorable tax rates that flow from the way these taxes are calculated.

Incidentally, there are almost no direct revenue impacts to changing gift tax rules, many people choose to limit their lifetime gifts to an amount that won't cause them to pay actual tax during life.

Indexing Exemption Amounts

The unified exclusion should be indexed. Almost all other comparable amounts in the tax code, such as the personal exemption, tax brackets, standard deduction and gift tax annual exclusion are already indexed. Inflation shouldn't change effective tax rates.

The political drive to repeal the estate tax was largely a product of the fact that the original tax free amount, $600,000, established in 1976, didn't change for twenty-five years, many of which were marked by high inflation. By the time the $600,000 was finally increased to $1,000,000 in the Clinton Administration, this once tidy sum barely covered a medium sized single family house in California.

Inheritability of Exemptions

The ability of a surviving spouse to inherit a deceased spouse's unified credit should be implimented in some form. This proposal is part of H.R. 5638 and its purpose is to dramatically reduce compliance costs. H.R. 5638's right of a surviving spouse to inherit any unused exclusion amount of a decedent should be adopted.

Most married couples want to leave everything to a surviving spouse, but this is an idiotic thing to do if the couple has a combined net worth in excess of an individual's estate tax exclusion.


The estate tax basically applies to everyone you own at death that is not given to a spouse or a charity. The estate tax exclusion is use it or lose it, and is on a per person per lifetime basis.

For example, if a husband and wife each own $2,000,000, the husband leaves all to his wife at his death, and then wife dies, she has an estate of $4,000,000, and no estate taxes were due at the first death. But, if she dies in 2009, she has only $3,500,000 of exemption and pays a tax on the remaining $500,000.

In contrast, if husband had left his $2,000,000 to a trust for the benefit of his wife, his exemption would have made that trust funding tax free, and the wife has a taxable estate of only $2,000,000 at her death, so she would also pays no estate tax, producing a savings of about $225,000 under current law.

Setting up trusts to accomplish this objective cost about $2,500 (your mileage may vary). Just about everyone with a net worth over the exemption amount, or for that matter, even close, needs one. This means that about three million families need them, which means that an aggregate $7.5 billion of investment in legal paperwork to avoid estate taxes needs to be in place at any time to deal with the non-inheritability of the exclusion. Realistically, this means that people need to spent about $200 million a year or so, as newcomers to the wealth level for estate taxes are a concern replace those who have died or seen their wealth dissipate. This is a dead weight loss to the tax system.

The simplest way to eliminate the need for this kind of tax planning is the one found in H.R. 5638, in which a surviving spouse inherits all unused exemption of a decedent. But, this isn't the only available approach.

The main alternative would be to limit inheritance of an exclusion amount to the amount given to the surviving spouse by the decedent, rather than to the amount unused at the decedent's death. This would be administratively more complex, because it would make it necessary to assign a value to assets at the first spouse's death, even if everything was left to the surviving spouse. But, some people would prefer it because it encourages "estate equalization."

Under either H.R. 5638, or an inheritance of exemption system based on the amount received by the surviving spouse from the decedent, the decedent has to leave something to the surviving spouse to avoid wasting the surviving spouse's exemption at the second death, although this can be done in the form of a QTIP trust, which requires that all income be paid to the surviving spouse, but allows the decedent to control who receives the property at the surviving spouse's death.

But, the couple can benefit taxwise from H.R. 5638 if the poor spouse is the first to die, but, husband and wife must first equalize their estates to benefit if the poor spouse if first to die if inheritance of exemption is limited to the amount received by the surviving spouse.

Under current law, if one husband has lots of money in his name, and wife has little or none, or visa versa, there is a strong tax incentive to change this by equalizing their estates. The trust for the surviving spouse system only works if the decedent has assets to fund the trust with at death. Normally, this is accomplished by simply transferring property from the rich spouse to the poor spouse.

Estate equalization can also be accomplished with a trust known as an "inter vivos QTIP" which pays income annually to the beneficiary spouse, and is treated for estate tax purposes as belonging to the beneficiary spouse, but leaves the rich donor spouse in control of who gets the property at the beneficiary spouse's death. QTIP trusts are normally used in second marriages, to perserve assets for children from a first marriage, and an inter vivos QTIP costs another thousand or two dollars to set up.

If estate equalization is an important value, it makes sense to limit inheritance of the exclusion from a surviving spouse to the amount actually inherited. But, in the case where the beneficiary spouse is the first to die, the QTIP income is paid at a time when the beneficiary spouse probably already lives in the same household as the spouse and hence usually benefits from the spouse's wealth, regardless of how it is titled, so the benefit's to the poor spouse from this requirement are modest.

The other benefit of not requiring estate equalization is that it doesn't disadvantage couples who don't do it simply out of lack of familiarity with the specialized area of tax law pertaining to estate taxes, something that is particularly common among wealthly people who have not grown up in wealthy communities where these taxes are more widely known.

The estate tax shouldn't be simply a tax on those too ill informed to plan for it. The best taxes, indeed, have little unintended impact on economic behavior at all. Estate taxes are designed to encourage people to make large charitable gifts. The fact that they encourage elaborate trust and entity arrangements is merely an unintended side effect which is wasteful and serves no valid purpose.

Given the limited benefits involved in linking estate tax exemption to the amount received by the surviving spouse, the H.R. 4638 method of allowing a surviving spouse to inherit the unused exemption of a decedent is the better choice.

State Tax Credits

Ten percent of federal taxes should be designated as a state tax credit, to be allocated amongst states where the decedent was domiciled and states where there are assets with a strong connection to a particular state like real estate and closely held businesses located in that state, if it agrees to not impose an estate tax of its own. This would discourage states from adding to complexity by enacting their own estate taxes with a different structure, or increasing the total estate tax burden. It would also provide revenues cut from state budgets without consultation from Congress in the lastest round of estate tax reforms.

Closing Loopholes

There are a few other issues that drive an immense amount of costly estate planning, which drives up compliance costs while reducing estate tax revenues. Closing some of these loopholes would make the estate tax simpler, would make up for much of the revenue lost by increasing exemptions and lowering rates, and would make the system more fair. Loopholes related to closely held businesses, life insurance and "split interest trusts" are particular important.

Closely Held Businesses

Privately held businesses are hard to value. This is inherent in the nature of the beast. It is entirely legitimate to note that privately held businesses may be worth less than publicly held businesses because shares are more difficult to buy and sell, because the death of the company founder or CEO makes the business less valuable, or that earnings have been inflated by a CEO who took artificially low compensation because he or she knew that it would be reflected in higher share prices for him or herself and other family members.

But, one huge tax valuation discount found in the current law has nothing to do with actual value. It flows from the fact that there are no tax rules requiring valuations to be done on a consistent basis across multiple gifts, or across both gifts and estates. Each gift is valued in isolation, without regard to what either the donor or donee owns. For example, it is possible to obtain a discount for the reduced value of owning a minority interest in a company when one receives 1% of a company's shares, even if the shares are received by the controlling owner of the company. It is possible to give all of the shares of a company to heirs at a valuation that values those shares as if they were minority interests in the company (typically 35%-50% less than their pro rata share of the company's asset value), even if all the shares end up going from a father to a single daughter, for example. These discounts have even been obtained when there is little or no underlying business other than inventment assets in the business, and the plan has been developed in the shadow of asset owner's serious health problems which ultimately lead to that person's death not much later on.

This is nonsense, results in massive compliance costs, and is very costly in terms of tax revenues. There is nothing wrong with intentionally giving a tax break of some kind to closely held businesses. There is something wrong with making those tax breaks contingent on expensive disingeneous hypertechnical appraisals that ignore the reality on the ground and elaborate tax motivated transfers.

A simple rule, that reflects the fact that a majority of closely held businesses operate on a consensus basis in which voting rights are largely irrelevant, could solve this problem. This rule would require the value of interests in business entities to be calculated by first valuing the business as a whole, and then allocating that value consistently on all gift and estate tax returns involving the business among the owners on a pro-rata basis proportionate to their economic rights in the company. Thus, while lack of marketability discounts would be allowed, minority interest discounts and control premium analysis would be eliminated.

But, one of the reasons that these ridiculous tax breaks have been allowed by Congress to persist is a strong concern about the impact the estate tax has on family owned businesses.

The legitimate concern of closely held business and farm owners is that they don't want estate tax issues to disrupt business operations, and that unlikely a publicly held company, fair value to pay estate taxes cannot be obtained from the underlying illiquid assets simply by selling shares on the open market in the nine month time frame normally allowed for payment of estate taxes. Congress doesn't want to destroy business value by forcing fire sales.

Another concern is that the estate tax encourages farms to sell out to developers, because farming is often not the most valuable use of land in the short run, and requiring a farm to finance taxes based on development values is unfair.

It is also legitimate for closely held business and farm owners to complain that existing law protections for closely held businesses, such as special use valuation, the family owned business exclusion (currently a dead letter, but something that will again be relevant if current law remains in place in 2011), and special installment payment arrangements for closely held businesses are themselves complex and to some extent micromanage a business.

But, the desire of closely held business and farm owners to owe less tax even when they are just as wealthy as the owner of a portfolio of stocks and bonds is not legitimate.

There are better ways to solve the legitimate concerns of closely held businesses and farms.

One would be to eliminate almost all restrictions on special use valuation, which allows closely held businesses and farms to value property based on its value in a going concern, even if this is less than development value. This currently has a stingy dollar cap, elaborate requirements related to the importance of the business in the estate and which family members must be active in the business, and the number of years that the business must remain family owned.

Instead, special use valuations could be made available without restriction to any closely held business or farm, so long as the heirs agree, in a manner that binds them and any successor in interest by gift, to pay the estate tax rate, rather than the capital gains rate, on any gains in excess of the special use value when the property is ultimately sold or was no longer used as a closely held business or farm.

Thus, for example, if an estate with $90,000,000 of other liquid assets has a ranch worth $1,000,000 as a ranch, but $10,000,000 developed into condominiums, the ranch could be valued at $1,000,000 for estate tax purposes, so long as a 45% tax rate was paid on any gain in excess of $1,000,000. So, for example, if the ranch were sold two years later for $10,000,000 for development into condominiums, a tax of 45% on the $9,000,000 gain would be due, producing only a couple of years of tax deferral from what would have happened if no special use valuation was involved. But, if the property was held as a ranch until the heir's death, the excess development value never realized, would never be taxed.

Similarly, the favorable financing and installment arrangements now available only to estate with a large percentage of closely held business or farm assets, could be made available to any estate, to the extent that their estate tax liability comes from closely held business or farm assets, and would be secured by a lien on those assets, to prevent them from being sold with the deferred estate taxes unpaid.

Life Insurance

Under current law, life insurance is included in a decedent's estate, at death benefit value, if the decedent has any control over naming the insurance beneficiary or access to the policy's cash value. As a result, there is a strong incentive to spend money to create trusts that benefit the insured's family to own the insurance and keep it out of the decedent's estate for estate tax purposes. Frequently, life insurance proceeds are the primary factor causing an individual to have a potentially taxable estate. But, because the planning work around is relatively easy to accomplish, the rule including death benefits doesn't generate too much tax revenue.

Still, existing law does impose some limits, because premium payments from the insured or the insured's spouse are treated as gifts, either to a trust, or in a properly crafted "Crummey trust" to its beneficiaries. Simply excluding life insurance proceeds from taxable estates, which would seem the simple solution, is a problem, because a terminally ill person could, for example, buy a $100,000,000 life insurance policy for $98,000,000 and shield vast amounts of otherwise taxable assets from the estate tax. This already happens to some extent. A large majority of cash value life insurance is purchased by wealthy people aged 55 and up for tax planning purposes, even though they don't need the "insurance" protection against an unlikely event that can't be covered with existing assets for which families with dependent children buy life insurance.

One could limit so called "Crummey trusts," which are named after the fellow who won the court case which made them possible, by tailoring a law narrowly to the holding of the Court case that made the possible. Under current law, gifts to a trust can only escape gift taxation if they are under $12,000 per beneficiary, and the beneficary has a right to withdraw the funds for a reasoable period of time which then lapses and becomes a part of the trust. It would be a simple matter to provide that gift taxation is escaped by a Crummey trust only if the money is actually withdrawn, something that never happens in practice, and this is probably a reform worth encouraging.

But, eliminating Crummey trusts doesn't really solve the entire life insurance problem because this would only encourage wealth insured to give money outright to their children, while strongly encouraging them to buy insurance on their parents with the money, knowning that gifts might not continue otherwise, in the absence of a legally binding trust, and would encourage families with many children to either enter into complex life insurance policy co-ownership arrangements, or to purchase multiple policies on the same person for tax reasons.

The cleanest solution to the life insurance problem would probably be to require that any proceeds from insurance purchased with funds traceable to the insured (made after the effective date of the law, so as not to harm people who reasonably relied on prior law) be included in his or her estate for estate tax purposes. This theory was considered in a few court cases, but ultimately rejected based on the language of the estate tax statute. This would virtually eliminate the incentive to create life insurance trusts for the purposes of estate planning, would partially curtail a largely tax motivated cash value life insurance industry (although income tax benefits beyond the scope of estate tax reform would leave it alive and breathing), and would modestly increase estate tax revenues, which is appropriate if the theory of the estate tax is, as it should be, as a tax in lieu of an income tax on heirs, so that the focus is on what is received by non-charitable and non-spouse beneficiaries, rather than what is actually owned at death.

Split Interest Trusts

A third major strategy to reduce gift and estate taxation is the split interest trust. These trusts have one person who is an "income" beneficary, who receives a fixed dollar amount or fixed percent of trust assets each year, and another person who is a remainder beneficiary.

Under current law, the remainder beneficiary's gift is considered complete and taxed for gift tax purposes, when the trust is created. This is done using a formula based on prevailing interest rates and the current fair market value of the assets in the trust.

The purpose of a split interest trust in an estate planning context is to undervalue the interest of your heirs. For example, if the value of a remainder interest in the trust (usually after a term based on a fixed number of years or life expectency) using the IRS formula is $100,000, but you have good reason to believe that this particular asset will actually be worth far more than the formula predicts at that point, you can reduce the effective gift and estate tax rate on the gift.

This is basically like playing blackjack against a house dealer who has inflexible rules on how to play the game, when you are allowed to count cards to know what your odds of winning are in reality in each hand.

In addition, any appreciation in the asset during the term of the trust is shifted to the remainder interest owners, even though they don't actually have asset to the asset until the trust terminates. So they are treated for gift and estate tax purposes of owning something that the donor, through the terms of the trust, still controls.

Again, a very simple rule could eliminate the elaborate actuarial calculations, complex trusts and gamesmanship that goes into these arrangements, increasing estate tax collections to make up for rate reductions and exemption increases, and reducing the compliance costs associated with trying to jump into Congressionally created loopholes that encourage purely tax motivated behavior.

The solution would be to treat split interest trusts as incomplete gifts from the donor until their term expires. "Income" interest beneficiaries in split interest trusts would be deemed to have received gifts directly from the donor in each year when they get the money from the trust in an amount equal to the fair market value of what was actually received in that year. Remainder interest beneficiaries would be deemed to have received a gift from the donor who established the trust in the year the trust ends in an amount equal to the fair market value of the trust when it terminated. If a donor died prior to the end of the trust term, the trust's assets would be included in the donor's estate.


An estate tax based on proposed 2009 law, with a 45% tax rate and $3,500,000 per person exemption amount, indexed to inflation, with the tax modified to reduce the estate planning driven by quirks of the current law, some of which are true loopholes that reduce tax revenue, and others of which are mere pitfalls in the way the tax is structured, would be a good way to put the issue of the estate tax to bed.

It would:

* Remove a vast number of upper middle class people from the group who has to pay estate taxes or engage in estate planning compared to current law, and make the top marginal tax rate less ominous (45% instead of up to 60%) for those who do pay the tax.

* Continue to impose a tax in lieu of federal, state and local income taxes on unrealized capital gains of decedents, and on inheritances which look no different to heirs than lottery proceeds upon which they would ordinarily pay high ordinary income tax rates, with a minimum of complexity or administration costs. This would also discourage the creation of an American aristocracy and continue to encourage charitable giving of great wealth.

* Reduce revenues from current law by less than existing 2009 law, because it would also close loopholes. The 2009 law involves a cut of about 50% from currrent law, and the proposal above might net out at a cut of 25% or so from current law.

* Greatly reduce compliance costs for taxpayers by closing loopholes that are now the basis for most of the routine tax planning by almost all wealthy families. This also has the effect of reducing dead weight loss from the tax to irrational economic behavior and of increasing the fairness of the tax amongst those who pay it with different levels of sophistication, planning and asset mixes.

* Protect the legitimate interests of closely held business and farm owners, in a greatly simplified manner, without making major concessions on ultimate tax levels for these taxpayers based simply on the form in which they own their wealth.

No comments: