Work and unemployment

The essentials

To characterize underemployment, the unemployment rate is a questionable indicator.

Reducing the unemployment rate is one of the key objectives of public economic policy. However, as we shall see, this indicator is not without its faults: used on its own, it can be misleading.

Don’t confuse the unemployed with job seekers

The number of unemployed is estimated by national statistical institutes, such as the Bureau of Labor Statistics in the United States or Insee in France. They base their estimates on the International Labor Organization (ILO) definition of unemployment.

An unemployed person is a person of working age (15 or over) who :

• did not work at all during the reference week;
• is available for work within 15 days and ;
• actively looked for work in the previous month.

Data are estimated from questionnaire surveys(Enquête Emploi in France; in Europe, all national surveys are grouped together in the ” EU Labour Force Survey ” published by Eurostat). The aim is to provide a standardized, non-manipulable assessment, enabling comparisons over time and between different countries.

The number of jobseekers is also often reported in the press or by governments. They are produced by the government bodies that receive, compensate and guide jobseekers, such as Pôle Emploi in France. These figures correspond to the number of people registered with these organizations: they are therefore not comparable over time or between countries, as they depend on regulatory or institutional developments (who can register? how are people deregistered, etc.).

Logically, these data do not overlap with unemployment figures. For example, in France, category A ⁴ is close to the ILO definition of the unemployed, but does not include unregistered jobseekers (such as young graduates looking for their first job and not entitled to compensation). Conversely, some people registered with Pôle emploi may not meet all the ILO criteria (no active job search and/or unavailable within 15 days). At the end of 2017, the number of people registered with Pôle emploi in category A stood at 3.7 million, and the number of unemployed people in the ILO sense at 2.7 million. ⁵ .

The unemployment rate is too narrow an indicator

Unemployment has a strong conventional component: it’s a “status”, not a “physical” reality. It is possible to be unemployed but not unemployed (e.g. student, pensioner, housewife or househusband); someone who has only worked one hour in the reference week is not considered unemployed.

In its May 2017 Economic Bulletin, the ECB looked at “labor market slack” ⁶ . Noting that the fall in the unemployment rate in the eurozone has not translated into wage increases as it did in the past, the ECB highlights the fact that the unemployment rate is too narrow an indicator today to correctly assess the employment situation. It overlooks two categories of the population concerned by the phenomenon.

– The unemployment halo refers to people who are looking for work but who do not meet the other two criteria of the ILO definition: they are not available within 15 days (because they are on an internship, in training, ill, etc.) or they have not taken any steps to actively seek employment (these are the “discouraged” unemployed).

– Underemployment refers to part-timers who wish to work more (involuntary part-time).

Another example: the fall in the unemployment rate may in fact conceal the precariousness of the labor market, with the multiplication of short-term jobs or involuntary part-time work, and incomes that are far too low to live in dignity (notion of the working poor). ¹⁰ .

The COVID 19 crisis provided another particularly striking example. Despite the impact of the containment measures on the economy, the unemployment rate in France has continued to fall, reaching around 7% at the end of June 2020. Beyond the announcement effects, we need to look at the INSEE note to understand how such results are possible. It’s also well explained in Olivier Passet’s video on Xerfi Canal.

For this reason, some authors suggest measuring a “ full-time equivalent employment (or non-employment) rate”. “employment (or non-employment) rate in full-time equivalents”. See the work of economists Guillaume Duval and Gabriel Galand.

Productivity gains have been spectacular since the start of the industrial era

A study of long-term statistics shows the extent to which gains in labor productivity ¹¹ since the beginning of the industrial era.

The results are even more spectacular when working time is taken into account: between the beginning of the 19th century and the end of the 20th century, the average annual working time of employed persons was almost halved. ¹² as a result of government-led reductions in working hours (introduction and gradual increase in paid vacations, reduction in weekly working hours) and the development of part-time working at the end of the twentieth century. ¹³ .

As a result, GDP per hour worked increased by a factor of almost 25!

In short, over the period 1950-2010, output per hour worked increased almost 8-fold. What’s more, productivity gains over the long term, which are already impressive when considered at the macroeconomic level, are even more so when we look at specific sectors.

These productivity gains have enabled an extraordinary rise in living standards, the extent of which has been demonstrated by the work of French economist Jean Fourastié (1907-1990) and his successors. He set out to calculate changes in the “real price” of a good or service over the long term in France. To do this, he compares the nominal price of the product with the labor cost of the lowest-paid worker (today, this is the Smic in France). ¹⁴ . The “real price” of a product, at a given date and in a given country, thus represents in this work an estimate of the working time required by the least qualified worker to acquire this product.

Until the middle of the 19th century, the real price of a quintal of wheat fluctuated between 150 and 250 hourly wages. ¹⁵ . Today, it is 0.9: it takes less than an hour’s wage in France to buy 100kg of wheat. It’s easy to see why we’ve gone from a situation of endemic famine to one of near-abundance of food.

Gains in labor productivity are mainly due to the replacement of human labor by that of machines and the energy that powers them.

For productivity gains to occur, human labor must be replaced by “something else”. According to economic theory, this “something else” is technical progress. ¹⁷ i.e. all innovations that improve the efficiency of production: product innovations (e.g. plastics), process innovations (new manufacturing methods, such as mechanization, and new ways of using energy) or organizational innovations (e.g. division of labor with Taylorism, assembly line work with Ford, just-in-time with Toyotism, etc.).

These useful distinctions, however, mask the fact that over the long term, technical progress has mainly manifested itself in the replacement of man by machine.

Take agriculture, for example.

The French agricultural workforce (family and salaried) has fallen from 6.2 million people in 1955, or 31% of total employment, to less than 900,000 today, or around 3% of the working population. The number of farms in 2016 was 436,000, compared with 2.3 million in 1955. At the same time, farm yields have risen sharply. For wheat, they have risen from less than 20 quintals per hectare in the early 1950s, to an average of 70 in 2018 (and even over 100 on many farms). ¹⁸ . This trend cannot be explained without reference to the parallel development of machinery in the fields and on farms (tractors, combine harvesters, milking machines, irrigation equipment, etc.), as well as that used to manufacture and transport seeds, pesticides and fertilizers (it should be added that the latter two inputs are manufactured using fossil fuels), and to market production.

Let’s take a closer look at organizational innovations.

Admittedly, the division of labor introduced by Taylorism did increase the productivity of individual employees. But the increase in productivity enabled by this innovation was multiplied by the use of machines, whether within the factory itself (conveyor belts on assembly lines, progressive mechanization and robotization of tasks), for the supply of raw materials or for the flow of production (mechanization of mines, development of road and sea transport). Similarly, product innovations require machines for their manufacture (and very often use oil directly as a raw material: this is the case for plastics, fertilizers, pesticides, many medicines, etc.).

This evolution is linked to the fact that the work done by machines (and the energy that powers them) is out of all proportion to that done by humans. A miner with his shovel and pickaxe extracts 100 times less ore than a site machine operator in an industrialized mine.

Over the long term, unemployment in the broad sense of the term is mainly due to machinery and secondarily to competition from low-wage countries.

The question of the causes of unemployment is clouded by questions of terminology. We are interested here in the causes of the decline in total time worked in a “developed” country, i.e., the “shrinking of the cake” that constitutes the supply of jobs for a given working time. This shrinking of the cake has multiple consequences for work, depending on the social categories concerned, the skills of individuals, the functioning of the “labor market”, legal working hours, the institutional and economic conditions for supporting the unemployed, and so on. But this is the main issue at stake: if the supply of jobs falls, their distribution is bound to become more complicated to organize.

If you read the news, you’d think that the main cause of this drop in the job supply in Western countries is linked to relocation to low-wage countries. In this article, however, we aim to demonstrate that, at global level, the main cause of job destruction is linked to labor productivity, the replacement of man by machine. In the literature, this is referred to as the capital-labour trade-off.

Corporate accounting makes payroll ¹⁹ the employees’ share of production, as expenses to be reduced. This is a powerful incentive for business leaders to cut labor costs. At the microeconomic level ²⁰ it is clearly in their interest to reduce this share in order to increase their margins.

There are two ways to do this:

– increase labour productivity: produce the same output using less human labour;

– reduce salaries either directly in the company (which is difficult in countries with protective labor laws) or through the company itself. ²¹ ) or by subcontracting or outsourcing part of production (to companies with the same productivity but lower wages) in the same country or in countries with lower wages.

Clearly, these two methods do not have the same impact on employment. In the first case, there is a reduction in the volume of work required, and therefore a drop in employment. In the second case, “expensive” work is replaced by less “expensive” work: the impact is not on the overall volume of work, but on the quality of employment.

On a country-by-country basis, competition with low-wage countries can become a source of unemployment, as companies prefer (and sometimes have no choice) to produce where it costs them the least. Globalization and the lowering of customs tariffs, often accompanied by currency dumping, mean that wages in developed countries are competing with low wages in emerging countries. Low transport costs ²² did not change the equation, many companies have relocated their activities, while others (such as the electronics industry) have set up directly in these countries. This international competition has led to an uneven distribution of jobs, with certain countries doing better in one area or another. Nonetheless, the primary cause of unemployment remains the fact that the overall volume of employment is insufficient due to mechanization.

Finally, even in low-wage countries, mechanization is gaining ground ²³ because machine-generated work always ends up costing less than human labor, in those trades where substitution is possible.

This theme of replacing man by machine is not new. In fact, it’s resurfacing today with the development of robotization and artificial intelligence. ²⁴ with fears that it will replace not only physical work, but also intellectual activities (e.g. automatic translators, algorithms replacing traders, medical analysis robots, etc.).

Economists have regularly invalidated this thesis on the grounds that labor productivity is a source of economic growth, and therefore creates jobs either in new activities (e.g. those required to design and build the machines in question) or via the additional purchasing power (generated by the redistribution of productivity gains), which is a source of new demand. The jobs lost would therefore be more than offset by jobs in the same industry, thanks to market expansion, and by jobs in new industries. ²⁵ . However, while these mechanisms may have worked during the 30 Glorieuses, there’s nothing automatic about them (see Myth 2 for the redistribution of productivity gains).

The quest for productivity gains can have negative social and ecological consequences

Beyond its effect on employment volumes, the priority given to gains in labor productivity can have other negative effects on society, particularly in a context where the economy needs to adapt to the physical and ecological realities of our planet.

The quest for productivity gains can be particularly destructive in the service sector. The concept of productivity gains only makes sense if we can dissociate the quantity produced from the volume of work required to produce it. However, in many so-called “relational” services, there is no quantifiable product other than the time spent looking after beneficiaries who require care, advice, training, live shows, listening… It is therefore virtually impossible to measure physical productivity (i.e., working time in relation to the quantity produced). It is, however, possible to measure productivity in terms of value (i.e. the quantity of work in relation to the added value generated by the activity). Since we don’t (yet) know how to replace a doctor, teacher or lawyer with a machine, labor productivity in the service sector is limited to the ability to organize “better”. This is why productivity gains have been much lower in services than in industry or agriculture. The quest for labor productivity leads to a systematic undermining of working conditions and the quality of services rendered. ²⁶ . Productivity means fewer nursing staff to carry out a greater number of procedures, fewer supervisors for children in nurseries and schools, care for the elderly reduced to the strict minimum, less time devoted by social workers to the people they deal with… We apply a logic of industrial performance and cost-cutting that threatens the individual quality and social usefulness of the services provided, since these are largely based on human relations.

Furthermore, the quest for gains in labor productivity is a major obstacle to the necessary shift towards a more sustainable economy. Indeed, most production processes that consume fewer natural resources (energy, water, soil, biomass, etc.) and pollute less require more labor than the same polluting, resource-exploiting production processes. Many jobs in sustainable agriculture, for example, are undoubtedly less productive (in terms of manpower), even if they are better paid and more efficient when all the factors of production used are taken into account (including those now forgotten, i.e. natural resources and the planet’s capacity to absorb our waste and pollution).

Sharing work can create jobs, as was the case when the 35-hour week was introduced in France.

Over the long term, effective working time (as a percentage of the human lifespan) has steadily decreased, as the statistics for France show. So, in effect, we’ve been sharing the work. The 35-hour working week introduced under Lionel Jospin also helped to create jobs (an Insee study put the figure at 350,000). ²⁷ and a censured DARES report ²⁸ confirmed these figures). The extent of the debate at the time was due, in part, to the fact that the changeover was sometimes carried out too quickly (as in the case of hospitals, where there was a shortage of staff). And for company directors and employers’ representatives, the switchover was complicated to implement.

The condition of “workers” (in terms of employment rates and income levels) depends primarily on the overall macroeconomic situation.

It was the economist William Phillips (1914-1975) who first attempted to formalize a relationship that seems fairly obvious: at a given moment (and all other things being equal, in particular legal labor laws and working hours), it is the balance of power on the labor market that determines wages and the number of jobs. If the supply of jobs is higher than demand (i.e., if there is a shortage of manpower, or a “seller’s market”, as opposed to a “buyer’s market” situation where it is the buyers of labor – the employers – who impose their conditions), it is likely that wages will rise, and that we will approach full employment. Conversely, if companies (public or private) and government agencies offer few jobs to a labor force in search of work, wages are likely to fall and unemployment to stagnate or rise.

This tension can clearly be observed in sub-sections of the “labor market”; today, for example, demand for IT specialists is high, and so are their salaries.

On a global level, the statistical proof of these causal links is more difficult to demonstrate, and has been the subject of much debate. Given the poor quality of the unemployment rate to represent the situation of workers in precarious situations, it seems to us that the statistical difficulty remains unresolved. The ECB recognized this when it observed that the unemployment rate camouflaged a much worse employment situation ²⁹ . The labor market (in Europe) is therefore more of a “buyer” than the unemployment rate suggests.

William Phillips’ initial idea (to show the inverse correlation between unemployment and labor costs) was transformed into a discussion of the inflation/unemployment trade-off: a situation of full employment would result in inflation (with wage rises being transmitted into generalized price rises). Conversely, inflation could only be controlled at the cost of minimum unemployment.

Milton Friedman, the theorist of the Chicago School, introduced the concept of ” natural unemployment ” in 1968; New Keynesian economists introduced the concept of NAIRU (non-inflation accelerating unemployment rate) in 1968. ³⁰ ) in 1975. These concepts (natural unemployment and NAIRU) are tragic in the primary sense of the word: they suggest that the public authorities can do nothing about unemployment other than reduce it, albeit a little, at the cost of a resumption of inflation, clearly unwanted by the upper middle classes and pensioners…

However, it is quite clear that :

• a situation of strong growth, such as that of the Thirty Glorious, and as experienced by emerging countries, creates work and drives up the price of labor;
• a public stimulus can boost activity and thus help create jobs, at least in the sectors concerned.

But if we reason in a no-growth or very low-growth world, it’s just as clear that productivity gains, if not shared, create downward pressure on wages and upward pressure on unemployment.

Unemployment is seen as a burden on the community, whereas it is a potential source of wealth.

Unemployment is seen and managed as a burden on the community. The unemployed weigh on the cost prices of companies (which participate in the financing of unemployment insurance when it exists). ³¹ ) and on public budgets. The “global cost of unemployment” in France is estimated at 100 billion euros per year by economist Jean Gadrey. ³² . In international competition, this mechanism is unfavorable to companies in the most “generous” countries. The temptation is therefore great to reduce it, and then to justify these restrictions with moral or microeconomic arguments: paying the unemployed to do nothing encourages laziness, is unfair to those who earn their bread by the sweat of their brow, and so on.

If we stop looking at things from an accounting angle, it’s clear that the unemployed represent available energy, latent creativity, which would be very useful for developing many of the activities expected of the population, particularly with a view to the ecological transition, which could lead to less use of machines in certain cases, to limit the use of fossil fuels. There’s no shortage of money or useful projects. Numerous solutions and experiments exist: subsidized employment, public-sector jobs, increased public procurement in transitional sectors, experimentation with zero unemployment, support for transitional companies, etc.

Seen in this light, unemployment is a real saving, since it frees up forces from immediate production that are potentially available to…do or produce something else. Unemployment is therefore a sign of great wealth. ³³ of our societies. In France today, we have saved an amount of available time that can be estimated at 200 days x 3 million = 600 million man-days (considering only the unemployed in the strict sense). A considerable amount of savings which should be considered… but which we do nothing with.

Failing to mobilize these “idle” savings collectively is both a huge waste and a major political risk: too many people in an “inactivity trap” – possibly “from father to son” – are not inclined to trust governments and their promises.