How Automation Will Affect Unfunded Pension Liabilities
In 2016, Deloitte Consulting and Oxford University did a study of public employees in the United Kingdom and estimated that up to 16% (almost 1 million jobs) could be eliminated by automation in the next twelve years. I think this is a wild underestimation.
In the U.S. military, we call it the tooth to tail ratio—how many people does it take in the rear (the “tail”), to keep the individual fighter locating, closing with, and destroying the enemy (the “tooth”). In World War II it was about 7 to 1. In Vietnam it was about 12:1. In the past ten years it has ballooned to over 20:1.
Automation holds out fantastic possibilities of rolling back that ratio, and thus beefing up the warriors in our armed forces as we redeploy people from tail to tooth. The same can also be said of our city, county, and state employees, except that there may not be the requirement for more employees interacting with the public. As the back office of government gets automated, those jobs may simply be eliminated.
This is great news for taxpayers and government efficiency, but really bad news for public pension systems. Fewer employees equals fewer people paying into pension funds, the vast majority of which across the nation, are grossly underfunded, with Connecticut, New Jersey and Illinois being the worst.
Another example of the coming automation wave, is that it dramatically affects certified public accountants. Our analysis shows that if it now takes 100 accountants to audit a 5% sample of all of your accounts, in the next ten years, all that will be needed is 15 accountants who will audit 100% of your accounts. Super human CPAs? Not quite. Machine learning and neural networks reduce the CPAs needed by 85%.
So back to the Oxford University study. If their 2016 estimate—that public employment can be reduced by 16% in the next 10 years – is accurate, it will devastate public pension plans. If Oxford has wildly underestimated the power of back office automation, and in reality, it approaches the vast change about to occur in the accounting world, there will be no feasible way to ensure the sustainability of public pension plans.
With politicians already playing wild games with our pensions—from politically driven investment decisions to closing general fund budget shortfalls by failing to appropriate the required annual pension fund contributions, as I have written frequently in this column, we are already in deep trouble. Add to the problem market-based rates of return and mortality tables, and it becomes deep, deep trouble. And when we look at the wonderful and exceptional changes to the quality of life coming to all nations with the advent of machine learning—in services, education, medicine, transportation, and others— but then figure that into what it means to already bankrupt public pension systems in the U.S., and deep, becomes deep, deep, deep trouble.