Options for Improving Cost Recovery for Structures

October 11, 2017

Last week’s Republican framework for tax reform included a proposal known as “full expensing,” which would allow businesses to immediately deduct the cost of certain capital investments. This proposal is aimed at reducing barriers to investment in the U.S. tax code. However, the framework would have excluded structures, such as buildings, from being eligible for full expensing.

It is not entirely clear why the Republican framework excludes structures from its full expensing proposal. After all, commercial and residential structures make up about three-quarters of the private capital stock in the United States, and are therefore a key part of the U.S. economy. Lawmakers seem to realize this intuitively: politicians often talk about getting companies to build factories in the United States, but perhaps they don’t realize that factories generally count as structures for tax purposes.

The question of the tax treatment of structures is particularly relevant because the current system of tax depreciation is especially unfavorable to structures.[1] When a company purchases a structure, it is required to deduct the cost of the asset over a period of up to 27.5 years (for residential buildings) or 39 years (for nonresidential buildings). There is reason to believe that requiring businesses to deduct the cost of their investments over such long periods discourages them from investing in the first place. After all, a deduction 39 years from now just isn’t worth very much for a business today.

eHere are some options that lawmakers could explore:

Revenue and Economic Effects of Options for Changing Cost Recovery for Long-life Structures
Billions of Dollars and As a Percentage of GDP
  Change in revenue, 2017-2026 Change in revenue, 2027-2036 Change in long-run annual revenue Economic effect
Note: For each of these proposals, “long-life structures” indicates all investment property that is subject to the 27.5- and 39-year schedules under the current tax code.
Note: Each of these estimates assumes that policies are enacted on January 1, 2017. This is a modeling convention, not a legislative expectation. All revenue estimates are static.

1. Allow full expensing of long-life structures

-$998 billion -$1,116 billion   2.4% increase in long-run GDP
-0.44% of GDP -0.33% of GDP -0.23% of GDP

2. Move to a 20-year depreciation schedule for structures

-$119 billion -$417 billion   0.7% increase in long-run GDP
-0.05% of GDP -0.12% of GDP -0.08% of GDP

3. Move to a 20-year schedule for residential structures and a 25-year schedule for nonresidential structures

-$76 billion -$265 billion   0.5% increase in long-run GDP
-0.03% of GDP -0.08% of GDP -0.06% of GDP

4. Allow 50 percent immediate expensing for structures

-$500 billion -$559 billion   1.2% increase in long-run GDP
-0.22% of GDP -0.17% of GDP -0.12% of GDP

5. Allow depreciation indexing for new structures (with an index rate of 5%)

-$16 billion -$137 billion   2.2% increase in long-run GDP
-0.01% of GDP -0.04% of GDP -0.33% of GDP

1. Allow full expensing of long-life structures

One option would be simply to extend full expensing to structures as well.[2] This would allow companies to immediately deduct the cost of their investments in buildings, and would remove the bias against investment in structures from the federal tax code. The Tax Foundation’s economic model estimates that enacting full expensing for long-life structures would increase the long-run size of the U.S. economy by 2.4 percent.[3]

One potential obstacle to enacting full expensing for structures is that the policy would come with large up-front costs. Over the first decade, full expensing for long-life structures would reduce federal revenue by $998 billion, or 0.44 percent of GDP. However, much of this cost would be transitional: in the long run, full expensing for structures would only reduce federal revenue by 0.23 percent of GDP.[4]

The revenue loss from full expensing would be larger in the short term because, for some time, companies would be taking two sets of deductions simultaneously. Companies would be able to fully deduct the cost of new investments, while still deducting a portion of the cost of old investments they had made years earlier. Once companies finish deducting the costs of their old investments, the cost of full expensing for structures would be significantly less.

2. Move to a 20-year depreciation schedule for structures

This option would preserve the system of depreciation for structures, but would shorten the time periods over which companies are required to deduct the cost of their structural investments. Under this option, companies would be allowed to deduct the cost of their investments in structures – both residential and nonresidential – over a 20-year period, using a straight line method. As a result, companies would be able to deduct a greater share of the cost of each building every year.

Because this option would increase the total present value of deductions for investments in structures, it would make it less expensive to purchase additional buildings in the United States. This would lead to an increase in the U.S. capital stock, and therefore the overall economy. The Tax Foundation’s model estimates that this option would increase the long-run size of GDP by 0.7 percent.

Over the first decade, this option would reduce federal revenue by $119 billion, or 0.05 percent of GDP. The revenue loss would rise slightly in the second decade of the policy, to 0.12 percent of GDP, before falling to a long-run cost of 0.08 percent of GDP.

3. Move to a 20-year schedule for residential structures and a 25-year schedule for nonresidential structures

Like the previous option, this proposal would shorten the time periods over which businesses are required to deduct their investments in buildings. However, it would preserve the current tax preference for residential structures over nonresidential structures, by allowing the former to be deducted over 20 years and the latter to be deducted over 25 years.

Because this option would do less to lower the tax burden on nonresidential structures (such as factories, malls, and office buildings) than the previous option, it would also do slightly less for the U.S. economy. The Tax Foundation’s model estimates that this option would increase the long-run size of GDP by 0.5 percent.

However, this option would also come with a lower revenue loss than the previous one. Over the first 10 years, this option would cost $76 billion, or 0.03 percent of GDP. In the long run, the policy would reduce revenue by 0.06 percent of GDP.

4. Allow 50 percent immediate expensing for structures

Under this option, companies would immediately deduct 50 percent of the cost of each building they purchase, and then deduct the remaining portion according to the current 27.5- and 39-year depreciation schedules. This is analogous to the current policy treatment of equipment and other short-lived assets, which are generally eligible for 50 percent “bonus expensing.”

Because this option would allow half the cost of structures to be expensed, it would provide half the economic benefit that moving to full expensing for structures would. According to the Tax Foundation’s model, it would increase the long-run size of the economy by 1.2 percent.

This option would also have half the revenue effect as full expensing for structures, lowering revenue by $500 billion (0.22 percent of GDP) in the first decade and 0.12 percent of GDP in the long run.

5. Allow depreciation indexing for new structures

One final option for improving the tax treatment of structures is “depreciation indexing,” which is also sometimes known as “neutral cost recovery.” This approach would preserve the current system of depreciation deductions for buildings, but would scale up the size of the deduction for each year a company is required to wait to take it.

The problem with the current system of depreciation stems from the fact that companies don’t value deductions in the future as much as deductions in the present. There are two reasons for this. First, depreciation deductions lose value over time, due to inflation: $100 in five years can’t buy as much as $100 can today. Second, even after adjusting for inflation, companies prefer having money in the present, because they can do productive things with it.

“Depreciation indexing” would correct the existing depreciation schedules for these two factors. For each year that a company is required to wait to take a depreciation deduction, this policy would scale up the deduction by a factor reflecting both inflation and a normal return on investment.

For the purposes of this option, we use an index rate of 5 percent. To take an example of how this would work: imagine a company builds a new structure that, under current law, would yield a $10,000 deduction a year from now. Under depreciation indexing, the company would be able to claim a $10,500 deduction instead (or $10,000 × 1.05^1). If the company would be able to claim a $10,000 deduction five years from now under current law, it would be able to claim a $12,762 deduction in five years under depreciation indexing (or $10,000 × 1.05^5).

Depreciation indexing would have roughly the same positive economic effect as full expensing: it would increase the size of the U.S. economy by 2.2 percent in the Tax Foundation model. This is because the policy would increase the present value of deductions for investment, removing barriers to building more factories and other structures in the U.S.

One key advantage of depreciation indexing is that it would have much less revenue loss in the short run. Over the first 10 years, the policy would only reduce federal revenue by $16 billion, or 0.01 percent of GDP. This is because companies would not start taking significantly larger deductions until far into the future. But ultimately, in the long run, depreciation indexing would reduce revenue by 0.33 percent of GDP.

Timing of revenue

As shown above, there are many ways to make the U.S. tax code less burdensome on investments in structures. Notably:

  • Proposals to increase expensing of structures lose the most revenue in the first decade.
  • Proposals to shorten the depreciation schedules of structures lose the most revenue in the second and third decades.
  • Proposals for depreciation indexing lose the most revenue in the very long run.

In principle, there’s no reason why lawmakers couldn’t adopt a combination of several of these policies, to achieve the timing of revenue that they desire. After all, lawmakers have expressed concern both about medium-term revenue (because of the optics of the 10-year budget window) and long-term revenue (because of the constraints of the reconciliation process and the Byrd Rule). For instance, one possible combination might allow for 35 percent immediate expensing of structures, decrease asset lives to 20 years and 25 years, and allow for depreciation indexing – all at once.

In any case, lawmakers looking to create a better tax climate for building factories, plants, office buildings, housing, and other structures in the U.S. have a number of policy tools to work with.


[1] A challenge with this proposal is that, at recent congressional hearings, members of the real estate industry have expressed skepticism about making expensing more available for structures. But, of course, the policy preferences of incumbent firms are not necessarily synonymous with the interests of the economy as a whole. For instance, all else being equal, existing owners of real estate might not want it to be easier to build new structures in the United States, because that might lead to additional competition.

[2] For this, and other proposals, we only considered changes to the tax treatment of structures that are currently subject to the 27.5- and 39-year depreciation schedules, which we refer to as “long-life structures.” Some structures are subject to shorter depreciation schedules, such as certain agricultural, utility, and improvement property; we did not model changes to the tax treatment of these assets.

[3] All revenue estimates here are static, in part because the official congressional tax scorekeeper, the Joint Committee on Taxation, does not tend to show large dynamic revenue effects.

[4] That said, this analysis is somewhat complicated by the existence of §1031 of the tax code, which allows companies to defer gains on the sale of certain assets, so long as the gains are invested in similar assets. §1031 is used widely by real estate companies, and has a similar effect to providing full expensing for structures. As a result, structures that are acquired under §1031 are partially or fully removed from the system of tax depreciation and receive considerably better treatment.


Related Articles