The bill aimed at regulating immigration and strengthening integration revives the persistent debate on the contribution of immigration to the economy and more generally to French society. In this context, several media have paid specific attention to a recent assessment which suggests that immigration could cost the French economy an estimated amount of 54 billion euros, that of Jean-Paul Gourévitch (The cost of immigration in 2023, Studies by Associated Taxpayers). To guarantee a rigorous and dispassionate debate, it is essential to give room to reason, which can only be exercised on proven facts. Academic work and other expertise seeking to enrich this debate with reliable data are indeed essential to leave room for the factual rather than the emotional. However, a figure is not a fact in itself, it is its level of validity and robustness that gives it its degree of veracity.
Assessing the economic contribution of immigration is a particularly complex issue. Indeed, our public statistics system is not built with the objective of disaggregating its different components by population subcategories. Such an evaluation therefore deserves a lot of attention in the data used, the scope of revenue and expenditure taken into account in the calculation, the time horizon, etc. In short, any attempt at costing is entirely dependent on the methodology implemented and the assumptions made in the absence of specific data. The result should not be understood as a truth set in stone, but rather as an evaluation which depends entirely on the method used and the postulates retained. Which also explains why most studies propose a range of values, varying the likely hypotheses, rather than a precise figure. Jean-Paul Gourévitch’s study has the great merit of being transparent about the (very) numerous hypotheses put forward; transparency which nevertheless leads us to be doubtful, for two essential reasons, about the validity of this cost of 54 billion euros.
Several aspects of the methodology give rise to reservations
A large part of this evaluation, like many academic works, attempts to assess the impact of the immigrant population on the scope of French public finances alone. However, several aspects of the methodology employed give rise to substantial reservations. First of all, the outline of the population concerned on which the study is based (8.2 million regular migrants for 2023) deviates significantly (+17%) from the latest official statistics available for 2022 (7 million immigrants), based on civil registration data. Of course, it is not the same year, but their proximity leads one to question the reliability of the figures retained and used in each of the calculations carried out subsequently. The problem is quickly brushed aside by the author who criticizes INSEE for its “manipulations” (p.17), probably preferring to rely on his own intuition.
This scope of immigration (known as first generation) is also increased by the integration of direct descendants of immigrants (half when they come from a mixed couple). The inclusion of descendants of immigrants is obviously questionable to the extent that even if they are the result of past immigration, they nonetheless remain considered as indigenous within the meaning of the definition retained by the High Council at I ‘Integration and recovery by INSEE: an immigrant is a person born as a foreigner abroad and residing in France. This ultimately amounts to confusing the message about the object of what is being measured (what immigration are we talking about? Today’s, yesterday’s?) and to asking the question of where to trace the demarcation between the population linked to immigration and that which is no longer. The immigration used in this study is therefore approximately 14.5 million individuals (including those in an irregular situation), or nearly 21% of the population residing in France.
Just as problematic, the author applies the same characteristics to these descendants of immigrants as to new arrivals, brushing aside all work on the integration of migrants and their descendants. In terms of income taxes, for example, immigrant and immigrant households are attributed in this study an income (and therefore pay a level of taxes) 28% lower than non-immigrant households when latest official statistics from INSEE (Immigrants and descendants of immigrants – 2023) measure a gap of 22% for immigrants and less than 12% for descendants of immigrants compared to the rest of the population.
Once the perimeter of the immigrant population (re) defined, the author applies this proportion (share of immigrants, within the meaning of the study, in the total population) to all aggregates of taxes and public expenditure in order to deduce the contribution (and cost) of immigrants to public finances. The application of this rule of three (often refined by a few corrective elements based on intuition) however raises some major problems insofar as it amounts to not taking into account the legislation regarding taxation or payment of benefits. social Security benefits. For example, income tax is not proportional, but progressive with several tax brackets depending on the level of income (including a first bracket at zero rate); the calculation of VAT depends on the structure of individuals’ consumption (which is very different among the immigrant population); etc. Worse, the author ignores taking into account the specific age structure of the immigrant population (characterized by a proportional over-representation among the active population) and applies the average proportion of immigrants initially defined to the different aggregates of revenue and public spending (which amounts, for example, to considering that health spending is the same, whatever the age of the individual). And this is the whole limitation of this study, which is not based on microeconomic surveys making it possible to correctly consider the specific situation of immigrants in terms of public finances.
Social costs without reliable data
In our latest study published in 2022 (The budgetary impact of 30 years of immigration in France – An accounting approach, Revue Economique), we analyzed the budget of public administrations over around thirty years (1979 to 2011) by calculating the net contribution of immigrants to public finances as the difference between taxes, contributions and various levies paid and all the benefits obtained of the public sector. To do this, we used, among other things, the Family Budget surveys, carried out every 5 years by INSEE (the last one carried out covers the year 2017 and resulted in publication at the end of 2020, which explains why we we were unable to use it for our study), which make it possible to reconstruct household accounts, that is to say all of their income and expenses according to their origin, age and level of qualification. The estimates made show that over the entire period studied, the contribution of immigrants as a percentage of GDP is relatively low, between -0.5% and +0.05% of GDP. At the same time, an OECD report published in 2021 (Budgetary impact of immigration in OECD countries since the mid-2000s) compared the budgetary impact of immigration in 25 countries (including France) between 2006 and 2018 to reach the conclusion that it was between -1% and +1% of GDP in most countries ( France being situated according to the hypotheses retained on an average contribution of between + 1% and -0.85% of GDP). The negative balance of 54 billion euros put forward by Jean-Paul Gourévitch (i.e. around 2% of GDP if we relate this figure to the GDP of 2022) cannot however be compared to the results of the studies mentioned above. In fact, it includes societal costs which do not directly relate to public finances. These costs are absent from other studies, not because they would refute them, but quite simply because there is no reliable data to evaluate them. In this matter, the hypotheses can only be heroic.
This other part of Jean-Paul Gourévitch’s study leads the author to make numerous arbitrary choices in the absence of robust statistical elements: “For lack of precise statistics, we retained the rate of one third” (p .46) or even “the provisional figure that we have retained, i.e. 1 billion, unfortunately has no scientific value even if it is a minimum working basis.” (p.52), etc. To give just one more specific example, the author estimates the cost of drug trafficking attributable to immigrants in the following way. It takes the turnover of this traffic evaluated by INSEE to which it allocates the share attributable to the immigrant population, defined as follows: “One might think, reading the news items, that this traffic is entirely controlled by immigrants and their descendants. We limited ourselves to 90%.” (p.51). It is then expected that this same ratio, based on simple perception, will also be applied to the social costs of drug-related diseases and epidemics. This is not the case, since “due to the explosion in traffic over the last four years, we have adopted a ratio of 100% here”. The reasoning becomes even more complicated in the lines that follow, which focus on the cost linked to tobacco smuggling. This time, the author explains to us “that it is difficult to estimate the share of immigration” and decides to use a ratio of 12%. We could multiply examples involving this type of biased and unfounded hypothesis in the evaluation proposed to us of the costs, attributable to immigrants, linked to prostitution, counterfeiting, animal or arms trafficking. Their evaluation is far from trivial since it amounts to more than 41 billion according to this approach. In this part of his study, Jean-Paul Gourévitch extends to economic evaluation what Patrick Buisson confined to politics: “In politics, perceptions are facts” (France Culture, October 12, 2016).
In the field of economic evaluation, it is essential to comply with the recommendation that Bertrand Russell provided in his Message to future generations : “When you study any question […], ask yourself only what are the facts and what are the observations that confirm them? Never let yourself be distracted by what you want to believe.”
*Xavier Chojnicki is professor of economics at the University of Lille and Lionel Ragot professor of economics at the University of Paris Nanterre.