September 21, 2018 Rohit 2 comments
Macro effects of automation, inequality and demographics

Our population is getting older and older.
New automation technology will put workers out of their jobs.
Wealth inequality is rising, and the middle class is shrinking.

These three trends are discussed abundantly in various journals and newspapers. However, very few research reports realize that these three trends should not be looked at individually. Rather than being isolated streams evolving on their own path, they are more intertwined and influence each other significantly.

In this article, I aim to show some of the effects that these trends will have on our economy and society and the cause-effect relations they have on each other.

The demographic shift towards an older society is marking the end of an era that was characterized by an abundance of labour. Since the 1950s three forces combined to create the fastest labour force growth in history: The integration of India and China into the global economy, the entry of women into the workforce and the entry of the baby boomers into the workforce. The contributions of these three forces that have fuelled economic growth for the second half of the 20th century is now fading since their contributions have largely been absorbed by the economies and there is not much room for incremental improvement. For example, in 1950 only 35% of the working age women were employed in the US. By 1970 the women participation rate rose to 50% and stands at around 75% today.

The Baby Boomer generation, which includes people born between 1946 to 1964, was born when the full effects of vaccinations and antibiotics led to sinking childhood mortality rates but before the mass introduction of birth control. The baby boomers are now entering the retirement age, which has massive effects on our demographics. Research states, that the population of those above 65 will grow faster than the working age population in the OECD countries. This would lead to a significant decline in our labour force growth rate.

One mitigating effect is that longer and healthier life spans have convinced baby boomers to defer retirement. Life expectancy in the OECD nations in the 1980s was around 73.5 years. By 2015, it had risen to 80.3 years. This has resulted in a rising effective retirement age which would reduce the negative effects on labour force growth.

However, we are not only experiencing a delay in the retirement age but also a delay in the age of entering the workforce. Cohorts in their teens and 20s tend to study longer and enter the workforce later.  The combination of these trends: older workers delaying retirement, young students delaying entry into workforce and baby boomers moving into retirement, does not alleviate much of the global deceleration in workforce growth. Bain & Company projects that the deceleration in labour force growth could result in a $5.4 trillion GDP shortfall by 2030 in the OECD countries.

What other implications could the increase in the number of old households have?

Well, old households, in absolute terms, are spending more due to their longer and healthier life spans. Between 1984 and 2015 US households that were headed by a person aged 75+ increased spending by more than 50%. Old households finance their spending largely through financial savings. As baby boomers begin using assets and pensions, the large-scale liquidation of financial assets could lead to a decline in their value. This could pose a risk to government pensions that relied on financial asset inflation to meet future obligations.

For corporates, talent acquisition will be a major theme in the coming decade. How can they incentivize millennials to work for their companies? Higher wages, flexible work arrangements and attractive corporate cultures (see Silicon Valley Tech companies) will certainly play a part in the incentive programme. Yet, what companies must realize is that the younger generation is looking to have a higher purpose of work, a sense of mission – providing that could be a valuable differentiator for companies in attracting and retaining talent.

Having considered the aspect of labour force growth, we will now look at the other factor that determines overall economic output: labour productivity growth.

Growth in labour productivity, as measured in output per worker, was relatively stable between 1980 to the mid-2000s in developed economies. Since the financial crisis in 2008, it has declined to nearly zero. There are two main forces that will determine the future growth rates of labour productivity.

First, labour productivity declines when workers shift to sectors in which less capital good options are available to improver productivity. The shift to a service focused economy moves workers from high productivity growth industries to low productivity growth industries. For example, between 1993 and 2014, the auto manufacturing industry in the US increased their output per worker by 128% while output per worker in hospitals only increased by 16% in the same period.

The automotive industry had much more options for capital good investments that made the production, assembly etc. much more efficient. The healthcare industry, represented by the hospitals, still requires a high degree of human action with less options of capital good investments making their services more efficient.

Also, the number of workers in the US automotive industry declined by 28% during 1993 and 2014, while the number of workers in US hospitals increased by the same number in the same time frame. This shift to lower productivity growth industries will likely continue for the foreseeable future.

Second, common consensus is that the new era of technological innovation will result in a strong increase in productivity as automation replaces or frees up human labour. Research states, that automation could put 20 -25% of the workers in the US out of a job in the US. The gap between the majority of the workers who will suffer from the negative impacts of automation and the few highly skilled who will benefit from it is likely to increase income inequality significantly.

Income and wealth inequality have been steadily increasing. The GINI Index for most of the OECD economies has been rising in the past 30 years. The GINI index measures the extent to which the distribution of income within an economy deviates from the perfect equal distribution. Consequently, a GINI index of zero represents perfect equality and 100 represents perfect inequality.

As mentioned in the beginning, the three forces – shifting demographics, automation and rising inequality are strongly linked. So how does the demographic shift effect income inequality?

First, an aging population typically increases inequality as older households tend to have higher levels of accumulated wealth compared to younger households at a similar socioeconomical level.

Furthermore, high income workers are more likely to be in occupations that do not require heavy physical work. Therefore, age related physical decline, is less of a restriction. Those in the bottom three income quintiles have a 64% chance of suffering age-related decline in the ability to work compared to 25% of those in the top quintile. As a result, high income workers work 77% more hours in 2006 compared to 1979 while those in the bottom 20% of income earners work 39% less hours in the same time frame. Adding to that, is the fact that people with higher incomes enjoy longer and healthier lives which enables them to sustain their period of wealth accumulation for a longer time.

Finally, we will look at how the increase in automation will affect inequality.

Experts analysis states that workers earning between 30 and 60k per year in the US will experience the greatest disruption from automation (disruption could mean displacement and depressed wage growth). The deployment of automation technologies is likely going to exceed the pace at which economies can re-educate and redeploy the millions of workers who will lose their jobs to automation. A study by the Federal Reserve Bank showed that a longer period of unemployment reduced earnings by 67% immediately and 32% 10 years down the road. Also, as automation spreads throughout the economy, occupations that are not easily automated, i.e. jobs requiring high social skills will experience strong wage growth. And since these are the jobs that are already well paid today, income inequality will be adversely affected.

Another aspect of how automation may increase income inequality is by increasing the share of income going to profits rather than wages. With increased automation, the share of income being produced by labour will fall. This means that labour will effectively also be entitled to a smaller share of the cash flows and the increased profitability, achieved by automation, would largely flow to capital providers (equity and debt investors). Generally speaking, capital providers tend to be in the higher income brackets, therefore the income gap will widen further.

As a final thought, the macroeconomic consequence of an increasingly unequal society seems to be a bit disturbing. If the higher income brackets of our society earn more income, they are most likely not going to spend that extra income. They are more likely to save the additional money than consume it. This would result in declining demand. Assuming automation will continue to increase the supply of goods and services, we would very likely be living in a demand constrained economy. This, in turn, will have big implications on what kind of monetary and fiscal policies and measures will work.

 

Source: Bain & Company’s Research Report:Labor 2030: The Collision of Demographics, Automation and Inequality

2 Comments on “

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.