In February, the . About . Or those of the , carried out in 2017 for the year 2060, . There are enough to create confusion.

Different methodologies

Why do the figures differ? The Idea Foundation explains that it carried out its calculations “with the hypotheses that [it] considered most reasonable in terms of mortality, birth rate and net immigration”. The difference with others is explained by its productivity estimates. “In the PDAT, they assumed a very high productivity. Whereas we calculated it by looking at the evolution of value added over the last 30 years by branch, which we projected with refinement. We deduce a number of jobs and a need for new population.”

“Our methodology makes it possible to determine the productivity gains necessary to achieve the various long-term GDP growth rates,” explains Tom Haas, head of the modelling and forecasting unit at Statec. “This is the concept of apparent labour productivity, which includes technological progress but also gains in capital intensity.”

In the document explaining its projections, the institute recalls that “there is no consensus on the very long-term growth prospects. Some consider that the productivity gains linked to digitalisation will remain rather limited. The ageing of the population and the necessary deleveraging of households and companies should even contribute to a continued decline in growth. Others anticipate that productivity gains linked to automation, artificial intelligence or renewable energies are only in their infancy. Hence four very different scenarios.”

Pascale Junker, director at the ministry of the economy, underlines that Luxembourg Strategy's approach is different. “Our method is that of participatory strategic foresight and not of macroeconomic projections. It is based on the collective imagination of what the future could be.” It does provide figures, but “purely for information purposes, based on known sources.”

To reach 1.2 million inhabitants in 2050, which the Statec only reaches in 2060, “we transposed [the Statec data] to push the extremes and reflect past trends to underestimate the migratory balance and the increase in life expectancy”. This is confirmed by Idea: “The ‘positive dynamism’ scenario of 2004-22 predicted that by 2020 Luxembourg would have 395,000 jobs (compared to 475,000 in reality), a number of cross-border commuters of between 136,000 and 168,000 (compared to 200,000) and a population of between 511,000 and 561,000 inhabitants (compared to 626,000).”

When asked about the origin of the 770,000 population scenario, Junker repeated that “these figures are fictitious.”

Caution on an increase in productivity

There are different methods for different results... But which one is closer to reality? Two economic specialists from the University of Luxembourg share their thoughts.

“In 10 years, the population has grown from about 500,000 to 600,000 inhabitants, i.e. a growth of 2.1% per year. If we extrapolate this, we get 1.1 million inhabitants in 2050,” says François Koulischer.

“I understand that there are several scenarios. That's what statisticians are asked to do,” explains Arnaud Bourgain, also a member of the National Productivity Council. “Academics are more cautious”. Nevertheless, he added “when you look at the literature, economists tend to go for the explanation of the slowdown in productivity. So I would be a bit wary of taking the most optimistic scenarios”. Even if “it's a simple deduction from the current literature. Can we extrapolate it over the next 30 years? Nobody can say. If you look at the last 15 years and the crises we have experienced…”

A basis for reflection

So why make these projections? “The IGSS (General Inspectorate of Social Security) makes calculations on the financing of pensions,” replies Bourgain. “This can also open up the debate on the necessary infrastructure.”

This is a more difficult exercise for a country like Luxembourg. “We have seen a sharp increase in the labour force, migration and cross-border workers. There is more volatility than in an average country.”

The Organisation for Economic Co-operation and Development (OECD) also makes its projections. “In the long term, we prefer to talk about scenarios,” says Yvan Guillemette, an economist at the organisation. This is to better “reflect the fact that we are not really in a position to determine productivity growth in 30 years.”

For short-term projections, the method consists of an analysis of “economic cycles”. However, “we don't project cycles” further ahead. In this case, “we look at the situation over the last 15 years” to project a central scenario, the economist explains. “In the 70s to 90s, we had higher rates. It is not impossible that we will experience new technological advances. Hence the addition of an optimistic scenario.”

Finally, a pessimistic scenario is added to ‘balance’. The expert admits that “it is normal to consider the baseline scenario as the most reasonable.” For the OECD, it serves as a “point of comparison to consider the impact of a reform”, on the labour market for example.

Unsurprisingly, no economist can predict the future. But it is at least a little easier to understand how and why they try to do so.

This story was first published in French on . It has been translated and edited for Delano.