Background and problem statement: For the purposes of its immigration policy, the Ministry of Asylum and Migration needs an overview of the development of the number of foreign nationals residing unlawfully in the Netherlands. To this end, estimates of this population are made periodically. Estimates have previously been made for the period 1997–2003, for 2005/06, for 2009, for 2012/13 and for the year from 1 July 2017 to 30 June 2018. There is now a need to update the figures again. To this end, estimates are made for five con-secutive years, each running from 1 July to 30 June of the following year, from July 2018 to June 2023. The estimates are made using a variant of the capture-recapture methods, known as the truncated Poisson model. This method offers a number of important advantages, such as relatively low costs, an adequate theoretical-statistical basis, explicit assumptions and clarity about the consequences of violating those assumptions. The specific method used to create the 20+-year series, the truncated Poisson regression model, has been identified in an international comparison as a method that meets scientific quality requirements for estimating population size. The problem statement for the new study is as follows: What are the estimated sizes of the population of foreign nationals residing unlawfully in the Nether-lands in the period from mid-2018 to mid-2019 up to and including mid-2022 to mid-2023 (five con-secutive years)? What is the trend in the estimated size of the number of foreign nationals residing unlawfully in the Netherlands when the new estimate is compared with previous estimates? How reli-able is the estimate? The main purpose of the estimates is to monitor developments in the size and composition of the population of foreign nationals without a right of residence. The idea is that the extent to which the estimation methods make realistic assumptions about the data used cannot be determined with any certainty, but that we assume that any violation of the assumptions is comparable over the years. Although the point estimates do not have to be exactly correct, the identified trend can be considered valid. Background and expectations Immigration policy The size and composition of the population of foreign nationals residing unlawfully in the Nether-lands are influenced by developments in immigration policy, immigration control, the expansion of the European Union, international migration and the labour market. These factors reinforce and limit each other. Since the 1960s, three phases can be distinguished in government policy on irregular migration. The first period, from 1960 to 1969, was characterised by organised and spontaneous labour migra-tion and a laissez-faire policy. In a period of rapid economic growth, foreign workers were actively recruited, but labour migrants also came to the Netherlands on their own initiative. Those who found work could usually obtain a residence permit. After the introduction of compulsory prior work and residence permits in 1968, residence without a permit was considered unlawful for the first time. A period of tolerance followed from 1969 to 1991. After the recession of 1973, a large number of mi-grants remained in the Netherlands, sometimes without the right of residence. In practice, this resi-dence was often tolerated. The Tax and Customs Administration also issued a social security number without a residence permit, which allowed people to work legally.2 There were no regular checks, and limited regularisations took place in 1974 and 1980. Since 1991, there has been a more restrictive and deterrent policy. This policy focuses on discourag-ing unauthorizedresidence. Important milestones were the linking of residence status and social se-curity number (1991), the Compulsory Identification Act (1994), the Linkage Act (1998) and the Aliens Act 2000. These laws restricted access to employment, social services and housing for persons with-out residence rights and strengthened supervision and deportation. EU enlargement and its impact on irregular migration: The accession of Central and Eastern European countries to the EU (particularly in 2004 and 2007) had a major impact on migration to the Netherlands. Citizens from these countries were granted freedom of movement and access to the labour market. Before 2004, a considerable number of East-ern Europeans were residing in the Netherlands irregularly (estimated at 65,000–107,000), but after EU enlargement, this number fell to around 6,000 in 2017–2018. Irregular residence of EU citizens mainly occurs when people stay for more than three months without (legal) work or sufficient means, especially in the case of temporary migrant workers. Immigration control and departure in the period 2018–2023 The expansion of investigation and supervision powers since the 1990s has increased the possibilities for enforcement. However, Covid policy caused a temporary decline in the number of checks and de-partures during the research period. The number of checks carried out by the Aliens Police, Identification and Human Trafficking Department (AVIM) fell from 5,150 in 2019 to 2,420 in 2020 and then rose gradually to 3,460 in 2023, still below the pre-Covid level. The number of departures across the chain – the total number of foreign nationals leaving the Netherlands via the Repatriation and Depar-ture Service (DTenV), the International Organisation for Migration (IOM) or after arrest – showed a similar trend: an increase to 25,600 in 2019, a decline to approximately 18,000 in 2020–2021 and a recovery to 23,000 in 2023. The proportion of demonstrable departures (forced or supervised) in-creased from 46% in 2019 to 67% in 2023. It is obvious that the limited departure capacity during the pandemic temporarily led to a larger irregular population and that the recovery after 2021 contributed to a decrease. International migration and diversity of origin: Over the past two decades, the Netherlands has increasingly become a migration society. The num-ber of immigrants rose from 154,000 in 2010 to around 270,000 in 2019, fell due to the pandemic in 2020 but peaked at 403,000 in 2022, mainly due to the arrival of Ukrainian displaced persons. In 2023, the number stood at 336,000. Net migration has remained positive since 2014, peaking at 224,000 in 2022. The migrant population is more diverse than ever, with origins in more than 220 countries. In addition to the traditional countries of origin, many new countries are represented, es-pecially from Asia, Africa and Latin America. Labour market and shortage sectors: Between 2006 and 2021, the number of labour migrants increased from approximately 200,000 to more than 800,000, two-thirds of whom came from Central and Eastern Europe. Migrant workers mainly fill shortages in healthcare, construction, distribution and technology. In addition, there are many posted workers, including tens of thousands of third-country nationals from former Soviet re-publics. When they remain in the Netherlands after their secondment has ended, they often have an irregular status. There are also indications of irregular work in the domestic sector, including by Fili-pino and Brazilian migrants. The Labour Inspectorate also points to abuse of the Non-Residents Reg-istration Scheme (RNI), which allows persons without a residence permit to obtain a Citizen Service Number (BSN) and work under the guise of legality. In short, structural labour shortages encourage labour migration, while supervisory shortcomings in specific sectors and administrative abuse perpetuate a reservoir of irregular labour. Expectations: The Dutch migration situation illustrates the so-called liberal paradox: a combination of restrictive policy and openness to humanitarian and economic migration. Due to ageing and labour shortages within the European Economic Area (EEA), the recruitment of workers from outside Europe is in-creasing. The population of unlawfully residing foreigners consists mainly of labour migrants from new recruitment countries and rejected asylum seekers. The total number is expected to remain sta-ble: labour demand stimulates inflow, while restrictive policies and legal migration routes slow down this development. European figures confirm that the number of irregular foreign nationals has not changed significantly in recent years. Methods and data: Capture-recapture methods are widely used to estimate the size of ‘hidden populations’. These are populations whose members are not easily observable and are not all registered. Standard estima-tion techniques based on sampling cannot be used in such cases because there is no sampling frame or it is very inadequate. One of the methods, Poisson regression, has been used to produce the en-tire series of estimates since 1997. This method uses police files containing data on the apprehension and detention of foreign nationals without legal residence. Since 2005, this has been the Police Suite Enforcement Foreign Nationals (PSHV) file (gradually replaced by HV, Enforcement, since 2019). The estimate is based on the number of times that persons registered in the file were apprehended or detained in a single year. As in the 2020 report, an estimate was also made for this study using the Dual Systems Estimation (DSE) method based on PSHV and IOM. This time, the method was applied to five consecutive years. Now that estimates have been made using these methods for several years, a large and difficult-to-explain variability in the estimates has become apparent. The data and results for this series therefore do not provide sufficient confidence to present this method in the report as reliable and usable. The DSE estimates are therefore only discussed in an appendix (see Appendix 2). Poisson size estimates: We will discuss the size estimate for the first year, 2018–2019, as an example to illustrate the method. The estimate of the total population of foreign nationals residing unlawfully in the Nether-lands is 34,885. The confidence interval indicates that, with a confidence level of 95%, the average lies between 21,288 and 48,482. Compared to the estimate for the period mid-2017 – mid-2018, there has been an increase of more than 12,000 persons (cf. Table 0.2), but the confidence intervals over-lap to such an extent that this difference is not significant. For the following years, too, due to the large confidence intervals, which overlap for all five years and also overlap with 2017/2018, it is not possible to determine with certainty whether there has been an increase or a decrease. Development of the estimates In this section, we compare the new, successive size estimates for the period from 1 July 2018 to 30 June 2023 with the estimates for previous years. In the second half of the 1990s, the first period for which we have estimates, the number of estimated unauthorized immigrants fell from over 194,000 in 1997 to 144,000 in 1999. From 2000 to 2002, the estimated numbers rose again: from almost 163,000 in 2000 to almost 212,000 in 2002. Population estimates after 2002 show a clear decline, which accelerated in the years after 2003 and especially after 2005-2006. When we look at the new series of estimates from 2018 onwards, the decline in the estimated size of the population appears to be continuing. However, this decline was temporarily interrupted in the years that largely coincide with the Covid period. The first year, 2018–2019, shows an increase in the estimated number of foreign nationals without legal residence status. In the following year, 2019/2020, the first Covid year, the number falls again. In the following year (2020/21), the estimate of the number increases slightly again. In the last year (2022/23), the estimate remains virtually un-changed from the previous year (2021/22). However, due to the large confidence intervals, which overlap for all five years and also overlap with 2017/2018, it is not possible to determine with cer-tainty whether there will be an increase or decrease for the period 2018–2023. Discussion The most important finding of the now 25-year series is that the number of foreign nationals residing unlawfully in the Netherlands has fallen sharply since 2003. Although exact causal links cannot be es-tablished, there are developments that make it plausible that the number of foreign nationals resid-ing unlawfully in the Netherlands has actually decreased over the past 10 to 15 years. Since the mid-1990s, policy measures have gradually made unauthorizedresidence less attractive. The Compulsory Identification Act (1994), the Benefit Entitlement Restriction Act (1998) and, in particular, the Aliens Act 2000 limited the possibilities of working or claiming benefits without a residence permit. Strengthened surveillance, detection and identification, including the introduction of biometrics, im-proved enforcement reduced the odds of long-term irregular residence. Insofar as the use of biomet-rics has enabled better identification of foreign nationals, the estimates for previous years are an overestimate and the current estimates are more accurate. In addition, the development of asylum inflows played a role: around 2000, high inflows and a low grant rate led to growth in the irregular population, while in 2015–2017, more residence permits than asylum applications were granted, which meant that there was no increase in unlawful resi-dence. However, the most significant structural effect comes from the EU enlargements of 2004 and 2007, which gave hundreds of thousands of Central and Eastern Europeans the right to reside legally in the EU. Although the point estimates are subject to some uncertainty and may be slightly underestimated, there are no indications that the quality or the extent to which the model’s assumptions are violated changed significantly over time. It can therefore be concluded that the observed downward trend is real, even though the exact numbers are only approximations.