Situation Analysis

Potential Dangers of Unwanted Automation

Automation of processes has been on the rise for many decades now. Repetitive mechanical work has steadily been replaced by robotics and those processes are now more efficient, with a rise in productivity and output. Another aspect of automation, which has been in vogue for some time is Robotic Process Automation or RPA. This is used in repetitive human-specific processes such as automated chat-bots. This has reduced dependency on human power to perform such mundane and repetitive tasks.

On top of this, Artificial Intelligence (AI) has become the latest technology, which is expected to perform some level of thinking and analysis. As time goes on, this technology will become increasingly advanced, to replace human decision making in the process.

With remarkably high adoption rate of robotics by factories and business corporations, moving increasingly more work to the RPA and lately to AI is a sign of things to come. However, the frenzy to be ahead of the competition and to score favors with management and shareholders, a high percentage of executive management is indulging in implementing such technologies without doing a robust Return-on-Investment analysis. Such tools do not come cheap and the costs of replacements of systems and processes, in addition to the support and maintenance required to use such tools takes up lot of money saved by reducing the human power engaged in such activities. Implementation of AI tools to run routine business tasks just adds to the costs of running business. Many times, the productivity gain, if at all, falls short of the investments made and unnecessary changes, with possible complication of routine processes, makes the utilization of these tools ineffective. Wherever business growth is identified, it is because of the natural market conditions than use of any such tools. In summary, the utility of such tools for the purpose those are deployed is still highly marginal and thus questionable.

Even if we consider the sub-par benefits of such technologies, the research and innovations continue unabated. The end-goal of replacing humans in the entire process is both ambitious as well as scary.

At the same time, such innovations are not necessarily useless. Specifically, the use of AI can be harnessed to drive further innovation and discoveries in areas such as medicines, space, physics, and energy sources. As innovations in these areas are increasingly scarce and stagnated, AI can be used to advance human understanding and innovations.

Humans are always considered to be inefficient and prone to mistakes and fatigue. Thus, though the idea of replacing humans to increase productivity by using such automation is favorable, it has a heavy social downside. The reduction in manpower of many factories and corporations, by way of increased automation is steadily increasing, with the chances of it increasing manifold in the near future. Most of the jobs carried out by humans will entirely shift to automation tools, making these people redundant. With the frenzy of adoption of such automation tools across the world, it can render a large chunk of workforce redundant and will create an entirely different and unimaginable social problem. This large-scale deficit of jobs can also lead to systematic killing of individual productivity through skills and knowledge.

With ever increasing proliferation of such automation through any means, the real danger is in the employability of the current workforce which is liable to be replaced. This can lead to a net reduction in employment as very few people would be employed to maintain and support such tools.

With the advent of new AI tools every day, it has become extremely important to address such tools with their potential destructive nature. To counter this threat, a dedicated regulatory mechanism is needed, which would evaluate all such products from all angles and provide a certification only after ensuring that such products would be positive for the workforce as well as overall economy. If some vulnerability is found during evaluation, then the regulatory mechanism should ensure ways and means to neutralize misuse of such vulnerabilities. Only after due evaluation, such products can be certified for active use.

India would be one of the biggest victims of such automation, as a high percentage of its workforce is engaged in mundane support jobs, process outsourcing and call centers. In addition, with the growing industrial footprint, some of the existing manufacturing jobs also may be lost to automation.

Indian policies on the employment front now need to reflect these realities, given the current euphoria to adopt different technologies, without having a proper assessment. The employment policies need an overhaul, given the skewed employment structure and the growing threat to employability by automation.

On a broader level, the employment policies adopted by the Indian government need some fundamental improvements. The progressive shift of the working-age people to more productive jobs should anyway be on top priority for the Indian policymakers. Thus, a significant shift from agricultural workers to construction workers and then to jobs in the manufacturing and services sectors with increasing skills would be an obvious way to adopt, following the pyramid of sectoral output. To address this growing threat where almost half of the workforce employed in the least productive sector of agriculture and the skilled and productive workforce facing a threat of redundancy, the policymakers have a task at their hands.

When the need to sustain employability in the most productive sectors of manufacturing and services is paramount, excessive automation needs to be tackled with appropriate policy measures. One of the most important and obvious steps in this direction would be to identify those sectors which still need skilled people and are productivity positive. With continuous innovations and requirement of workforce in such sunrise sectors, it would be helpful to reorient policy measures to ensure adequate employment is generated through such sectors and skilled or even re-skilled workforce is available for them. In addition, encouraging school level curriculums to identify required skillsets which will help students acquire such skills earlier so that they can find employment in these sunrise sectors would be helpful. An across-the-board measure of promoting entrepreneurship and micro and nano businesses would also help encourage people to find gainful incomes.

At the same time, the non-essential or non-productive automation should be discouraged to the extent possible. Though as a thumb rule, private enterprises would conduct their due-diligence in implementing technology for it to remain productivity and cash-positive, such might not always be the case for some larger corporations. Corporate politics and one-upmanship make executive management in larger corporations indulge in expenditure which may not necessarily boost productivity but would end up making them stars with larger pay and incentives. This usually leads to wastage of hard-earned money over some vanity projects, cash leakages and productivity-neutral expenditures.

More transparency through regulatory guidelines on any such significant automation, which has potential to reduce the workforce, should become part of the monitoring process. Any automation, including AI and Robotics should be justified by commensurate output rise. Even publication of such results can lead to correction of stock prices of such corporations, where unproductive automation is done. This will instill some accountability into the corporate behavior. It is never a doubt that automation which shows rise in productivity and profitability should always be welcome. However excessive and unproductive automation should be discouraged at all costs.

In summary, the AI and robotics are only going to become more pervasive in future and ways to control their destructive proliferation as well as ensuring sustained employment of skilled workers should become most primary motives of the policymaking and its administration.

Economics is a social science by definition. This means that social behaviors and interactions determine how the economic system works. There is a considerable difference between social science and pure science. Pure science works on formulaic foundations and it is emotionless. It works on natural principles rather than behavioral principles. On the other hand, behavioral rules cannot be easily fitted in all-weather formulae. Thus, social science works on aspects which cannot be formulated but only assimilated to bring out rules which sustain in practice when the behavioral conditions are fulfilled. Economics works more on praxeology rather than mathematics.

Coming to economics, the demand and supply law has multiple behavioral aspects built into it. However, economics fails to affirm significant other formulaic rules where behavioral aspects are ignored. One such formula or relationship between economic aspects has led to high abuse of economic principles. This is the relationship between inflation and unemployment.

Many economists in the past and present have been attempting to establish a relationship between these variables through formulaic interpretations. This activity of building formulae for such parameters governed more by social behavior than any natural science is akin to hammering square pegs in round holes and for some other economists, hammering round pegs in square holes. Economists from John Maynard Keynes right till today’s Nobel Laureates have spent their life in such impractical interpretations and the tragedy is such that they have been revered and honored. This shows the lack of knowledge and inability to question ordinary citizens and the cabalistic nature of such economists. Many variations of these interpretations exist. These variations over a century miss the fact that in essence establishing a relationship between inflation and unemployment itself is flawed and inconsistent.

The key theme of these mathematical interpretations is that a rise in unemployment leads to lower inflation or a fall in unemployment leads to higher inflation. The justification for this goes on something like when more people get jobs, they spend more money and with the demand rising due to more people having disposable income, it to price inflation if supply is stable.

Going into detail of this preposterous relationship, let us look at the factor or assumptions:

·        The basket of commodities included in the Consumer Price Index (CPI) varies from country to country. Even in the US, instead of CPI, they use Personal Consumption Expenditure (PCE) which has a smaller set of commodities. Thus, the calculation of inflation as change in CPI would vary significantly from country to country and index to index.

·        CPI, by rule, consists of many components which are essential for survival. Any change in unemployment assumes that previously unemployed people were not consuming food and beverages, electricity, and some kind of shelter. The assumption that they suddenly start consuming these services, to have an impact on the CPI is bordering to naivete.

·        The proportion of change in unemployed people as against already employed is a fraction of 1%. This is highly insignificant to cause any meaningful change even if they start spending their entire salary in the CPI basket. For commodity and services prices to change, there needs to be a much larger and meaningful change in demand rather compared to historical changes in employment levels.

·        Another possible assumption of additional employment at median salary is also full of holes as many such employments happen at the lower end of the spectrum, for subsistence wages, and thus has an even lower, if at all, impacts on the demand. Such an insignificant rise in demand does not have any impact on the price rise.

·        In many cases, the unemployed get allowances, enabling them to spend on their necessities, keeping the CPI basket balanced and negating any impact of change in unemployment levels.

Beyond these, the development of such relationship ignores many more crucial factors, such as:

·        Rise in unemployment leading to lowering of inflation conveniently ignores which factors affect inflation. Demand and supply gaps cause natural inflation while government interventions cause artificial inflation, thus forming a feedback loop, where all wise central bankers look at the unemployment/inflation relationship to formulate their policies. Government releasing more money in the system through bond sales has a much higher impact on the demand and supply balance.

·        Inflation on food items and other important commodities in the CPI basket is more reliant on transportation costs, which in turn depend on fuel prices. The housing market is another significant factor impacting inflation numbers. Fuel-related inflation rise, or fall is entirely beyond the purview of any employment related actions. The ability to maintain adequate fuel supplies at steady prices is the responsibility of the governments and minor changes in consumption pattern by low-salaried people cannot dream to have any impact on fuel prices and thereby inflation.

·        Fiscal policy measures within the housing sector would more likely impact on the prices than any consumption change. Anyway, someone who has recently changed their role from being unemployed to being employed is usually not in a position to boost housing demand within the first few months with their newly found source of income.

·        Any interest rate increases by central banks to tame inflation also result in increasing the cost of borrowing. If a business has to shell more money to meet the interest outgo, such businesses would be reluctant to decrease prices of their goods and services, negating the action of central banks for reducing inflation.

·        Irrespective of interest rate reductions and/or increased demand, businesses are not keen to reduce prices, unless they face severe competition or they are unable to hold on to the goods, perishable or not, to keep the inventory costs down. When they reduce prices, it is based on their judgement, and they are least worried about the changes in unemployment.

This list can go on and on but suffice to say that inflation has many causes are situations to change but change in employment levels is certainly not one of them. By an extreme stretch of imagination, such relationship might feebly work, but lot of stars need to align for that.

In history, when such an equation worked, there have been multiple other crucial factors, which happened to support such equation.

The biggest worry is that all powerful central banks watch this equation closely and decide on interest rate changes and other tools to impact inflation. Such is the sad state of affairs that we live in.