Given that human labour is less necessary with every year that passes, why not tax automation rather than raise the age of retirement?


1 Answer 1

  1. Automation does not necessarily mean labor is less necessary. In 1900s there was less automation than there is today, yet today, even after 100 years of constant innovation and automation there employment in US is larger than ever before according to the department of labor statistics.

    Historically, automation in one industry (e.g. agriculture) just frees labor to pursue more valuable economic activity (e.g. computer programming). This historic trend might not continue forever as innovation in AI and robotics might create environment where there is no area where humans can be productively employed. However, it is not certain at all if it will ever happen (e.g. people might have taste for expensive hand crafter luxuries, people might start augmenting themselves - transhumanism), and it will almost certainly not occur in next few decades.

  2. Retirement ages are rising because of demographical decline. For example, projections show that 23% of total population in USA will be over age of 65 in 2060. EU or Japan will have much worse as they have healthier population. This rapid increase in number of old people will lead to drastic increase in retirement spending.

    Governments could raise taxes, but the tax increase would have to be massive, and it certainly could not be just limited to automation.

  • First just taxing 'automation' whatever that even means would be bad idea because that would lead to firms substitute away from automation, which would be bad idea, given that technological progress is improving standards of living. Moreover, depending on what you mean by automation even 100% tax might not be enough to cover for future pension expenses (e.g. do you only mean robots? That sector is not that big). You would need broader income tax.

  • Instead of taxing labor you could tax capital, but empirically most studies show that somewhere between 50-75% (and sometimes even nearly 100% but occasionally also bit less) of tax burden from capital is shifted onto labor in forms of lower wages (Roy-Cesar & Vaillancourt 2010, Gruber, 1997). So either half of the burden or most of it will be born by labor.

  • As to how big this burden is we can do some back of the envelope calculation. According to projections there will be 95 million people of retirement age in the US by 2060. Assuming, US will manage to grow at 2% its GDP at that time will be approximately \$48.52 trillion. Today USA pays retirees on average \$18,456 per year (Centre on Budget and Policy Priorities), assuming this will grow together with economy (so the relative position of retirees stays the same) USA will have to pay \$38,401 per retiree. Using this we can calculate that US would have to devote nearly 8% of GDP just to the retirement payments.

    This might not sound a lot but current US tax to GDP ratio is just 26.6% of which about 19% is social security and 21% of that are pensions so currently US devotes just about 1% of GDP to pensions.

    Hence 8% is 8x increase just in next 37 years. As a consequence this would mean that the tax to GDP ratio in US would have to increase from 26.6% to 33.6%.

    And we are talking about US here which has quite low pensions compared to the rest of the western world. In countries like Denmark, Sweden, France or Germany the tax increases would have to be much more significant. In addition, while US could in principle increase its taxes to match this countries like Sweden already have taxes so high that they are very close top of the top of Laffer curve (point beyond which higher taxes create so much distortions that they lead to less revenue not more). According to estimates provided by Traband & Uhlig (2011) EU-14 has on average capacity to raise labor taxes 8% and capital taxes 1% further before further tax increases will become self-defeating.

    So while US could in principle accommodate aging via higher taxation (although, my guess would be such unprecedented tax hike there would be extremely unpopular and voter repellent), for some European countries that is not an option. Only other alternative would be to either reduce pensions (which would fix the problem as well but probably be even more unpopular than raising retirement age) or alternative reduce other social or healthcare spending (In fact as society ages healthcare spending is also projected to increase, but this is whole other can of worms that I am not even going to open here).

    Moreover, the above back of the envelope calculation only consider next 37 years, this problem will not stop getting worse after next 37 years since demographic projections show massive nearly 10% population collapse in the Europe (and similar numbers hold for US) in next century (note the already scary 8x increase will happen even before population starts collapsing). This problem will be increasing exponentially (from 1% GDP needed in US now to 8% in 2060 to probably around 30% 2100 and so on).

    Original pension schemes were not designed to withstand this. When pension was first introduced in US with 65 retirement age, the average life expectancy was about 58 (meaning on average person would not even live to the age of retirement). Now life expectancy in US is at 77.

  • $\begingroup$ Electric vehicles will save the average car owning family 2000 dollars per year in petrol and car repair costs. That's equivalent to 5 years less work for an average person if the money is well managed, and that energy saving cost will also translate to all goods transported by road. So, not only is automation increasing the total working population, it is also reducing the price of basic commodities, which then require ingenuous blatant corruption like "fast fashion of single use clothes" when old clothes lasted 20 times longer. So prices are raised artificially and untaxed. $\endgroup$ Mar 20, 2023 at 6:32
  • $\begingroup$ @bandybabboon Electric car is not example of automation. Electric car needs the same amount labor as regular gas car to operate. In addition, for transportation of goods you need electric long hauling trucks and as far as I know technology is not there yet. Also average person in 2021 in US earned about 97k per year at 77 expected life time that is 1.5 year of income saved assuming we count even years where the person is baby, if we adjust for that its probably less than 1. Even if we take median income that is according to US census about 71K it would at best be 2 years (also counting the $\endgroup$
    – 1muflon1
    Mar 20, 2023 at 11:40
  • $\begingroup$ baby stage of life). census.gov/library/publications/2022/demo/p60-276.html In addition, we do not talk here about private retirement. Privately anyone can retire whenever they like even at age of 18. We are talking about public retirement, government won't be able to levy 100% tax on such savings, even for richest individuals top marginal tax rates that make sense even under most redistributive preferences are somewhere around 60-80% in static tax optimization models and 40-60 in dynamic ones. For an average person these rates are much smaller so taxing incomes more if things $\endgroup$
    – 1muflon1
    Mar 20, 2023 at 11:44
  • $\begingroup$ get cheaper is not realistic fix for this issue especially given that most countries already have high taxes (e.g. in Netherlands where I live top marginal income tax rate is 45.9% which is already within range of optimal top marginal income tax with dynamic models with most redistributive (Rawlsian) preferences, and that does not even include all social contributions that are implicit labor taxes which would push it for top incomes somewhere around 55% $\endgroup$
    – 1muflon1
    Mar 20, 2023 at 11:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.