How a Basic Income puts a nation back to work (yes, WORK)

Gavin R. Putland, BE PhD – Author, researcher

The crucial difference between a Universal Basic Income (#UBI) and conventional “welfare” is that a UBI is not withdrawn or reduced if the recipient also works for a living. This is not the only difference. It is not even the defining difference. But it is the crucial difference, the redeeming difference, and — if the Basic Income is made less than universal — the one feature that absolutely must be retained.

Why? Because if the Basic Income (BI) is payable to workers, wages only need to supplement the BI. So wages can be lower, without leaving workers worse off. So it becomes more attractive to hire new workers, because the cost is reduced, and less attractive to fire current workers, because the saving is reduced.

But how do we pay for a BI? I just answered that question! If workers get $X more from the Treasury in the form of a BI, and the rest of the private sector pays $X less to workers in the form of after-tax wages, then the rest of the private sector can afford to pay $X more in tax to cover the BI. As long as that $X of tax is levied on anything but labour, it doesn’t reduce the newly enhanced profitability of hiring workers. And that analysis is pessimistic because it doesn’t allow for the growth dividend: as a BI encourages hiring, it creates a bigger economy, which therefore automatically pays more tax even if the tax scales stay the same.

But wouldn’t the extra hiring spend the growth dividend because more people would get the BI? No, because there would also be fewer unemployed people receiving benefits! (In practice, the unemployed would also receive the BI, perhaps with a supplement to assist with their peculiar expenses. All other welfare payments would likewise be replaced by the BI or, at worst, the BI plus supplements for special cases, leading to a simplification of the system.)

But wouldn’t the extra tax raise prices? No, not when you allow for the reduction in labour costs, which feed into prices just as effectively as any tax — even the GST! Moreover, not all taxes feed into prices. If a price is inflated due to protection from competition (as with monopolies, cartels, locational advantage, etc.), a tax on the value of the protection reduces the vendor’s unnecessary gains (economic rent) rather than raising the price. And don’t forget the growth dividend.

But wouldn’t the extra jobs feed inflation by tightening the labour market and raising wages? No. If we’re silly enough to pay for the BI using taxes that feed into prices as the cost of labour does, then the only downward pressure on prices will come from the growth dividend, and inflation will restart if wages rise high enough to overcome that downward pressure. But that still leaves some margin for reducing unemployment. Using taxes that don’t feed into prices will obviously allow a greater reduction in unemployment before we hit what economists call “non-accelerating inflation”. Until then, a BI is deflationary.

But wouldn’t a BI take away the incentive to work?

I must remark at the outset that the question is either appallingly naive or appallingly serpentine, because unemployment, by definition, is an excess of willing workers relative to the number of available jobs. In those circumstances, defending the incentive to work is like the captain of the Titanic calling the fire brigade. “Mutual Obligation”, as we call it, is the doctrine that the unemployed should try harder to take jobs from you and your kids. The popularity of this position exemplifies Henry George’s brutal observation that “the majority of men do not think.” When there are several unemployed jobseekers competing for each available job, the policy of making them compete harder can serve no purpose except to force down wages. That is indeed its official purpose — which can never be officially stated, because the voters might not approve.

But, to answer the appalling question, no, a BI would not take away the incentive to work, because:

(1) Nobody is suggesting that the BI by itself should be a comfortable living. Although conservatives and progressives would disagree on exactly how uncomfortable a BI should be, the common presumption is that almost all recipients would want something more.

(2) As the BI would be payable to workers, the reward for working would be in addition to, not instead of, the reward for not working. A standard unemployment benefit (“JobSeeker Payment”, as we now call it in Australia) is withdrawn in response to earned income. This withdrawal combines with income tax to produce Effective Marginal Tax Rates (EMTRs) that are much higher than those faced by the highest income earners. If more than one welfare payment is withdrawn in this manner, a family can easily face an EMTR in excess of 100%, with the result that working loses money instead of earning it. (For examples, follow @DPlunky on Twitter.) What sort of incentive is that? In Australia, the problem is compounded because the Commonwealth has adopted buggy data-matching algorithms that double-count earned income and assign income to the wrong accounting periods, generating bogus claims of overpayment (#RoboDebts). A BI is immune to such disincentives because eligibility is not affected by earned income.

(3) In assessing the incentive to work, we must consider not only the incentive to get a job, but also the incentive to create one’s own job. If the BI is payable to self-employed people, it creates a safety net that makes people more willing to take the risk of starting their own businesses, and gives them more time to build it up. Even an uncomfortable safety net is better than none. New businesses, of course, add to the growth dividend and help to pay for the BI.

(4) If points (1) to (3) are not good enough, the “crucial” feature of a BI does not preclude an activity test, provided that work passes the test.

But, concerning point (2), shouldn’t the BI be withdrawn from high income earners in order to save taxpayers’ money?

No, because:

(a) If we accept the premise that the BI should be income-tested, then successive governments, in pursuit of “savings”, will keep lowering the threshold until we’re right back where we started — with the lowest-paid workers facing the worst disincentives. If there is to be a means-test on the BI, it should be an assets test, not an income test. (And ideally it should target the kinds of assets that can’t be produced, so that it doesn’t deter their production.) Assets tests don’t hit you until you have built up some sort of buffer. Income tests can stop you from building up that buffer in the first place. Australia’s welfare system, with its severe income tests combined with generous and leaky assets tests, is not designed to give a hand-up to those who need it; it’s designed to pull up the ladder from those who need it, and to give a hand-out to those who don’t. Besides, the ability to build up financial buffers is an essential part of preparedness for the next pandemic, so that the Minister for Industrial Relations won’t sound so let-them-eat-cake if he suggests that casual workers with no paid leave should have “already made provisions” for self-isolation!

(b) Saving taxpayers’ money is a meaningful goal if the government simply takes money and spends it. But if the government also hands out money in the form of welfare, the level of taxation becomes a rubbery figure, because arbitrary conventions decide whether the clawing-back of welfare payments is taxation or means-testing, and whether the payments themselves are expenditure or tax rebates. If all means-tested benefits were reclassified as non-means-tested benefits plus taxes that claw back the same money from the same people as the means-tests, then the official levels of taxation and expenditure would rise enormously, although in fact, apart from the labels, nothing would change. Similarly, if all welfare payments were reclassified as tax rebates or prebates, the official levels of taxation and expenditure would fall enormously, although again, apart from the labels, nothing would change. Labels matter, however, because they influence what happens next. The “income test” label is particularly damaging because it has led us to tolerate a system in which a precariously employed waitress faces a higher effective marginal tax rate than the CEO of Qantas.

(c) In view of (b), if you want to turn a Basic Income into a paragon of fiscal rectitude, all you have to do is classify it as a refundable tax rebate/prebate, to be deducted from the revenue side of the budget rather than added to the expenditure side. Instant small government!

(d) Welfare payments can be withheld not only through means-tests, but also through activity tests (with exemptions for those who can’t be expected to pursue the activity). As an income-tested benefit is equivalent to a non-income-tested benefit plus an income tax, so an activity-tested benefit is equivalent to a non-activity-tested benefit plus a tax on the lack of the prescribed activity. The latter test is clearly preferable — provided of course that employment and self-employment count as “activity” — not only because it is better to punish inactivity than to punish working for a living, but also because we don’t have a separate inactivity-tax system that can double up with activity tests to produce absurdly high effective marginal inactivity tax rates!

In summary, we can afford a Basic Income, especially if it comes with a reasonable assets test and a reasonable activity test. But what we definitely can’t afford is a so-called “basic income” that is clawed back in response to earned income, because that means employers have to compensate workers for the loss of benefits, which employers can’t afford, causing a shortage of jobs. In other words, what we definitely can’t afford is the status quo.

______________________