A philosophical economist's case against a government-guaranteed basic income.

• Author: Henderson, David R.
• Position: Report

To argue about a policy, it helps to know the specifics of the policy being considered. In an August 21, 2014, email, Matt Zwolinski told me that although he "can't stand too firmly behind any particular dollar value" for a basic-income guarantee (BIG), he had in mind $10,000 annually for every U.S. citizen age twenty-one or older. That is the proposal I discuss throughout this essay, unless I specify otherwise.

The Program's Budgetary Cost

The specifics matter. In 2013, about 230 million people in the United States were age twenty or older. (1) Of all U.S. residents, 87.1 percent were born in the United States (U.S. Census Bureau 2014). Of the foreign-born people, a substantial percentage would have, like me, become U.S. citizens. So a lower-bound estimate of

the percentage of residents who are citizens would be 90 percent. Applying this factor to the 229.8 million adults gives us 206.8 million adult U.S. citizens. The annual BIG expenditure for U.S. citizens, then, would be approximately $2,068 trillion. This expenditure estimate does not include any expenditure for administering the program or for monitoring for fraud. In other words, it is a minimum estimate.

Zwolinski has argued elsewhere that a BIG "would be much better than the current welfare state." He writes: "Current federal social welfare programs in the United States are an expensive, complicated mess. According to Michael Tanner, the federal government spent more than $668 billion on over one hundred and twenty-six anti-poverty programs in 2012. When you add in the $284 billion spent by state and local governments, that amounts to $20,610 for every poor person in America" (2013b, emphasis in original). He wonders, "Wouldn't it be better just to write the poor a check?"

It might be better to "write the poor a check," but notice what Zwolinski has done. He started by arguing for a check for every adult American citizen and then in his cost comparison shifted to having the government write a check only to poor people. In that same essay, he writes: "A Basic Income Guarantee involves something like an unconditional grant of income to every citizen. $o, on most proposals, everybody gets a check each month. 'Unconditional' here means mostly that the check is not conditional on one's wealth or poverty or willingness to work" (2013b). For that reason, the relevant expenditure does appear to be the $2,068 trillion figure given earlier.

Assume, as Zwolinski advocates, that such a program would displace all 126 federal antipoverty programs and all state and local government antipoverty programs. Later, I challenge that assumption, but for now imagine that it is true. Notice what would happen. A $2,068 trillion program would replace programs whose total expenditures in 2012 were $952 billion. Even rounding up the $952 billion to $1 trillion, the program that Zwolinski advocates is more than twice as costly in budgetary terms as current antipoverty programs.

How would such a program be funded? Let's grant for now Zwolinski's assumption that all other antipoverty programs would be eliminated. How would the federal government obtain the additional $1,068 trillion? (2)

Before considering that question, let's put the number in perspective. Federal government spending for fiscal year 2014 will come in at about $3,576 trillion (U.S. Joint Economic Committee and Council of Economic Advisers 2014, 32). So if the federal government were to implement a $10,000 BIG today, spending by the federal government would increase by 30 percent. This amount is actually an underestimate. I am assuming that Zwolinski would want the BIG to be a federal program. So the $284

billion that he estimates would be saved by state and local governments would not be automatically remitted to the federal government. Rounding up the $668 billion spent by the feds in 2012 to $700 billion, I conclude that the federal government would spend not an additional $1,068 trillion, but rather an additional $1,368 trillion. (3) This would be a whopping 38 percent increase in federal spending.

How would Zwolinski fund this major increase in federal spending? If his goal were to keep the already bloated half-trillion-dollar federal deficit constant rather than increasing it, he would need to have the federal government increase taxes from their estimated $2,993 trillion to $4,361 trillion, an increase of 45.7 percent.

One of the most well-established facts in the economics literature on government finance is that raising a tax rate by x percent raises the revenue from that tax by less than x percent. (4) The reason for this relationship is that the higher tax rate discourages the activity being taxed, so the tax base on which the tax is levied is smaller than otherwise. So if the federal government were to raise all tax rates by the same percentage to generate the revenue needed, it would have to raise all tax rates by more than 45.7 percent and probably substantially more.

Assume, for simplicity, a 50 percent increase in all tax rates, although the percentage would probably be more. Why more? Assume conservatively that a 45.7 percent increase in tax rates would reduce by only 5 percent the base on which the taxes are levied. (5) Then a 45.7 percent increase in tax rates would increase the federal government's tax revenues by only 38.4 percent. (6) Thus, tax rates would have to be increased by substantially more than 45.7 percent.

That conservatively estimated 50 percent increase in tax rates means that the current Social Security payroll tax (Federal Insurance Contributions Act or FICA tax), instead of its current 6.2 percent each on employer and employee, would be 9.3 percent. The bottom marginal tax rate on individual income, instead of being 10 percent, would be 15 percent. The top marginal tax rate on individual income, instead of being its current 39.6 percent, (7) would instead be 59.4 percent.

With the magnitude of these tax increases, one strong concern would be the size of what economists call "deadweight loss." The deadweight loss from taxes is the loss in wealth caused by taxes that goes to no one. So, for example, a 10 percent tax on gasoline might cause a loss in producer and consumer surplus of $10 billion while creating "only" $9 billion in tax revenue for the government. The $1 billion loss in consumer and producer surplus that goes to no one is called "deadweight loss."

Every tax, except a lump-sum tax, (8) causes a deadweight loss. Moreover, a theorem in economics shows that the size of the deadweight loss from a tax is proportional to the square of the tax rate. So, for example, raising a 39.6 percent tax rate by 50 percent up to 59.4 percent would not raise the deadweight loss by 50 percent; it would raise the deadweight loss by 125 percent. (9) A reasonable minimum estimate of the deadweight loss from the current federal tax system is 25 percent of revenues raised. So the tax increases required to fund the BIG would push deadweight loss from about 25 percent of revenue to 62.5 percent of revenue raised.

In short, three effects of a $10,000 BIG for all U.S. adults would be (1) a huge increase in the size of the federal government, (2) a huge increase in federal taxes, and (3) a huge increase in the deadweight loss from federal taxes. This analysis assumes, moreover, that the BIG would replace all existing federal welfare programs, including Medicaid, that are aimed at the poor and near-poor.

I have ignored until now one source of revenues other than taxes to fund a BIG: expenditures on programs other than those that are targeted to low-income people. Zwolinski cites somewhat favorably a study by economist Ed Dolan in which Dolan estimated the amount of a BIG that could be funded by limiting these other programs. Dolan proposed, for example, giving people on Social Security a choice "between the benefits to which they are presently entitled, or the UBI [universal basic income], but not both." This change in law, he estimates, "would add about $18 billion in funding and reduce the number of UBI claimants by about 57 million" (2014). It should be noted that, with the kinds of numbers being discussed, $18 billion is a drop in the bucket, less than 2 percent of the additional funding required. Dolan comes up with the big numbers by advocating higher taxes on people who pay mortgages, claim the personal exemption on their income tax forms, or take advantage of other distorting aspects of the income tax system. He points out that whereas...