Tax reform is hard enough when it is based on a set of principles around which Congress can write a new law. But when lawmakers start with no strong foundation, it is easy for them to sink into a morass of inconsistency and special-interest dealing. That’s what’s happening with the Tax Cuts and Jobs Act.

The House and Senate bills are filled with provisions that simply make no sense as tax policy. For example, the House bill would tax a form of non-cash income, tuition waivers for graduate students, but continue to exclude much more lucrative forms, such as employer-sponsored health insurance. The Senate bill would disallow a deduction for some ordinary business expenses--such as a portion of interest costs--but only for some business and not for others. 

Congress could have done this differently. It could have built its bill around the concept of lowering tax rates and eliminating tax preferences, as Congress did in 1986 and as former House Ways & Means Committee Chair Dave Camp proposed in 2014. Or it could have framed a bill as a shift towards a consumption tax, an idea proposed by the House Republican leadership in its 2016 tax reform blueprint.

But this year President Trump and the Hill GOP leaders quickly abandoned those principles. The TCJA seems only to have three goals: Cut tax rates for corporations and non-corporate businesses, add no more than $1.5 trillion to the budget deficit over the next 10 years so a measure can be enacted on a partisan basis using special budget rules, and pass a bill—any bill.

Of course, even starting with a framework based on solid principles does not keep a tax bill pristine. The inevitable legislative give-and-take is always going to require compromise. The ’86 Tax Act tried—but eventually failed—to eliminate the state and local tax deduction. It created many “transition rules” to ease the burden of its changes on specific industries and firms. And it created a curious “bubble bracket”—effectively a hidden tax rate that allowed the bill’s sponsors to claim a lower top individual income tax rate than the bill really delivered.

That bad idea, sadly, has been emulated by the House in the TCJA. But that’s only one of the bill’s many head-scratchers. Let’s look at two others.

Tuition waivers. Workers in the US receive all kinds of non-cash compensation: employer-paid health and life insurance, employer contributions to retirement plans, and paid family leave are all forms of non-wage income. So is the deferred compensation that makes up a large share of executive pay. Nearly all of it is tax-free or tax-deferred. Indeed, the tax exclusion for employer-paid health insurance alone will be worth more than $3 trillion over the next decade.

The House version of the TCJA effectively touches none of this compensation. But it would tax other, much more modest forms. One example: tuition waivers that universities grant graduate students in return for teaching courses or doing research. It would not tax scholarships, including the free tuition that, say, college football players get in return for their labor on the gridiron. Yet it would tax cash-poor grad students on the compensation they get for work that often is a step above indentured servitude.

Why? It isn’t for the money. The revenue gain would be tiny. It isn’t for the principle since neither the House nor the Senate bill would take a bite out of any big-ticket forms of non-cash compensation. It just…is.

Interest expenses. As a cost of doing business, interest is normally fully deductible from the corporate income tax. However, both the House and Senate bills would limit the deduction to 30 percent of business income. Because the TCJA would allow firms to take faster deductions for their capital investments, there is a good policy reason for changing the tax treatment of interest costs. If companies take more generous deprecation and also deduct interest on the money they borrow to purchase new equipment, they could end up paying a negative effective tax rate on new investment.

But here is the problem: The Senate bill doesn’t eliminate the interest deduction for all businesses. It exempts (so far) real estate developers, utilities, and car dealers—all heavy users of debt. Why? The glib answer is that they have good lobbyists. And that may not be so far from the truth. There certainly is no policy reason for the different treatment.

With no guiding framework, it is easy for a big tax bill to become little more than a legislative free-for all. Case in point: The Tax Cuts and Jobs Act.