LONGITUDINAL METHOD APPROACH IN ASSESSING

THE MOLDOVAN MIGRANT STOCK

DOI: https://doi.org/10.36004/nier.es.2023.1-03

JEL classification: F22, J10, J11, O15

UDC: 314.15(478)

Vitalie ȘTÎRBA,

Scientific Researcher, National Institute for Economic Research, Academy of Economic Studies of Moldova

PhD student, Department of Demography and Geodemography, Charles University, Czechia

https://orcid.org/0000-0001-5948-6509

e-mail: [email protected]

Andrei CRIVENCO,

PhD in Geography, Associate Professor, Shevchenko Pridnestrovian State University, Moldova

https://orcid.org/0000-0002-3096-6979

e-mail: [email protected]

Received:  12 January 2023         

Accepted for publication: 18 May 2023

ACKNOWLEDGEMENTS

The study was carried out within the project 20.80009.0807.21 ”Migration, demographic changes and situation stabilisation policies”, 2020-2023.

SUMMARY

Estimating international migration is a challenging exercise despite the information technologies used, regular censuses conducted, and available advanced administrative systems in collecting data on vital events. Recent works estimated the stocks and flows of Moldovan migrants mainly by using national administrative sources on population counts and data from the population censuses in the destination countries. This study aims to assess the stock of Moldovan migrants of the 1980–1995 birth cohorts, for which a longitudinal method was applied. Thus, we compared the de facto population of the corresponding generations to their initial size, adjusting by survival rate. The results show the existing stock of Moldovan migrants of the 1980–1995 year of birth and the impact of outmigration on the changes in the size of the corresponding cohorts as of 2016 and its trend until 2022. During the analysed period was emphasised a plateau in the migrant stock of the generations born in the early 1980s with a moderate return migration in the 2020–2022 period. On the other hand, the cohorts of the late 1980s and 1990s registered a significant increase in the number of migrants, which corresponds with the period of the entrance in high migration mobility ages. The study covers the entire territory of Moldova, including the left bank of the Nistru River and Bender municipality. The methodological part provides a complete description of the method applied, which might be further considered for estimating the migrant flows and stocks by using vital statistics or population census data. 

Keywords: migrant stock, cohort study, outmigration, Moldova, population dynamics, net migration

Estimarea migrației internaționale rămâne a fi un exercițiu complex chiar și în contextul disponibilității unui spectru larg de instrumente, precum recensămintele populației, resursele administrative de evidență a populației și tehnologiile informaționale avansate. Studiile recente au estimat stocul și fluxurile migranților moldoveni folosind datele naționale și din țările de destinație al statisticii migratorii și al recensămintelor populației. Acest articol are ca scop evaluarea stocului de migranți moldoveni din cohortele 1980–1995, pentru care a fost utilizată metoda de analiză longitudinală. Astfel, în cadrul studiului a fost comparată populația de facto a generațiilor corespunzătoare cu mărimea inițială a cohortelor, ajustată la rata de supraviețuire. Rezultatele studiului prezintă stocul existent de migranți moldoveni născuți în 1980–1995 și impactul emigrației asupra schimbărilor în dimensiunea generațiilor corespunzătoare pentru perioada 2016–2022. În perioada analizată s-au evidențiat schimbări minore în stocul de migranți ale generațiilor născute la începutul anilor 1980, care înregistrează valori moderate ale migrației de reîntoarcere, în special pentru perioada 2020–2022. În același timp, cohortele născute la sfârșitul anilor 1980 și 1990 au evidențiat o creștere semnificativă a numărului de migranți în perioada studiată, ceea ce se datorează vârstelor de mobilitate migratorie ridicată al acestora. Studiul acoperă întreg teritoriul Moldovei, inclusiv malul stâng al râului Nistru și municipiul Bender. Partea metodologică a lucrării oferă o descriere completă a metodei aplicate, care poate fi utilizată în contextul estimării longitudinale a fluxurilor și stocurilor de migranți prin utilizarea statisticilor vitale sau a datelor recensământului populației.

Cuvinte cheie: stoc de migranți, studiu de cohortă, emigrare, Moldova, dinamica populației, migrație netă

Оценка международной миграции нередко бывает затруднительной, несмотря на доступных для этого современный информационных технологий, проводимых регулярных переписей населения и качественного учета населения. Административные источники учета пересечений государственной границы, а также переписи населения, в том числе и в странах проживания мигрантов. В данной статье дается оценка численности молдавских мигрантов из когорт рожденных в 1980–1995 гг., для которых сравнивалась фактическая численность населения поколений с их исходной численностью, корректируемая по показателю выживаемости. Результаты исследования показывают текущую численность молдавских мигрантов 1980–1995 годов рождения и влияние эмиграции на изменение численности соответствующих когорт по состоянию на 2016 г. и ее динамику до 2022 г. В течение анализируемого периода, поколения родившиеся в начале 1980-х гг. регистрировали незначительный рост и умеренную возвратную миграцию в 2020–2022 гг. В то же время, когорты конца 1980-х и 1990-х гг. показывали значительный прирост численности мигрантов, что является следствием вступления соответствующих поколений в возраста с высокой миграционной мобильностью. Исследование охватывает всю территорию Молдовы, включая территории Левобережья Днестра и муниципий Бендер. В методологической части статьи дается полное описание применяемого метода, который может быть в дальнейшем использован для оценки потоков и численности мигрантов с использованием данных статистики естественного движения населения или переписи населения.

Ключевые слова: контингент мигрантов, когортный анализ, эмиграция, Молдова, динамика численности населения, нетто-миграция

INTRODUCTION

During the last decades, migration was the main reason for the population decline in Moldova, ultimately affecting other demographic phenomena. Consequently, outmigration caused an exodus of the labour force and depopulation, especially in rural areas, contributing to a population structure deterioration and accelerating the ageing process (Gagauz et al., 2023; Crivenco & von Löwis, 2022).

From the end of the 1980s until the mid-1990s, migration flows from Moldova included mainly ethnic minorities of Jews, Russians, Ukrainians, etc., who were involved in the process of repatriation to the countries of their origin (Tabac & Gagauz, 2020; Dietz, 2000). However, a pronounced outmigration from Moldova was registered since the end of the 1990s, which was mainly driven by the economic crises of 1998, when, in most cases, the decision to migrate was a response to high poverty in Moldova and emerging opportunities in the receiving countries (CIVIS & IASCI, 2010). Initially, the main destinations of the Moldovan migrants were CIS and EU countries (Piracha & Saraogi, 2012), while since the mid of 2010s, migration flows from Moldova have been predominantly to European countries (Tabac, 2021). Moldovan migrants initially opted for a determined period of stay in receiving countries, during which many migrants had an incentive for financial resource accumulation to improve their living conditions and cover current household spending. During the last decades, Moldovan migrants have opted for both seasonal and long-term migration (Görlich & Trebesch, 2008), depending on the destination country. Contrary to the above, migration became a life strategy for the younger generations, who consider early migration and social integration by obtaining higher education and vocational skills in the receiving countries.

Generally, emigration from Moldova is highly age-selective, involving the young and working-age population. Considering the settlement of Moldovan families in the countries of destination, there is an increase in the number of children migrating (Tabac, 2021). On the other hand, a positive net migration is observed among the population above 50, which is primarily involved in return migration.

During the years of independence of Moldova, outmigration has become an integral part of the public and socio-economic spheres of the state, shaping the patterns of the population’s economic, social, institutional, and political behaviour (Barsbai et al., 2017), at the same time contributing to economic development through remittances, transfer of know-how and investment, and changes in the population’s social values (Tabac & Gagauz, 2020).

The presented research aims to evaluate Moldova’s long-term international net migration of the 1980–1995 generations. For this, the 2015–2022 1^st January population of the corresponding cohorts were compared to their initial size (number of births) adjusted by survival rate. The study results are based on a longitudinal approach and emphasise the long-term net migration of the analysed cohorts as of 2016 and its trend until 2022. The study covers the entire territory of Moldova, including the left bank of the Nistru River and Bender municipality.

The study uses a new approach to studying international migration in Moldova by comparing the de facto population with the size of the corresponding cohorts. The method used might be further applied to countries with distorted population statistics and disrupted series of annual migration flows.

LITERATURE OVERVIEW

Migration is a demographic event, the reliable record of which is challenging even in countries with accurate population statistics. Generally, the migrant stocks and flows are often measured using population censuses, which collect information on population ethnicity and place of birth (Abel, 2013). Additionally, data from population censuses could be combined with administrative sources (Raymer et al., 2007). The article by Beer and colleagues describes the method of migration harmonisation when the administrative data on migration flows is available in both destination and origin countries (Beer et al., 2010). Therefore, a set of methods for estimating migration flows are also comprehensively described in the articles by Abel & Cohen (2019) and Azose & Raftery (2018). In recent times, social media platforms have been utilised to estimate the number of migrants through monitoring groups of expats and tracking migrants based on their place of birth, professional experience, and place of residence (Zagheni et al., 2017; Vieira et al., 2022).

Generally, migration is a highly selective (Rogers, 1981) and repetitive (Aude, 2017) process, which shapes the population structure and reflects on other demographic events in both origin and destination countries (Alho, 2008). In this regard, a longitudinal approach in migration studies allows for assessing the cumulative effect of the migration on changes in the cohort sizes. An eloquent example is the study by Kashnitsky, which assessed changes in the cohorts due to internal migration in central Russia (Kashnitsky, 2020). Moreover, the article by Willson and colleagues measured the impact of migration on birth cohort sizes in European countries (Wilson et al., 2013). However, these studies rely on net migration as their basis and ignore the step migration.

An estimation of the Moldovan migrant stock was made by Tabac, where the studies were based on population censuses and administrative sources in Moldova and receiving countries. According to the results presented, during the 2001 and 2011 census rounds in receiving countries, the Moldovan migrant stock was estimated at 338 thousand and 657 thousand, respectively (Tabac & Gagauz, 2020). On the other hand, based on data from population censuses, population registers, and national representative surveys, the United Nations Population Division estimates the 2020 migrant stock originated from Moldova at 1.1 million (United Nations, 2020).

In conditions of distorted population statistics, to adjust the population structure, Penina and colleagues estimated the stock of Moldovan migrants by using administrative data on border crossings from the population register, where is concluded that about 18% of the total population in 2014 were emigrants who did not live in Moldova (Penina et al., 2015). A similar approach in using data on border crossings from the population register was used by Gagauz, where the estimated net migration for the 2005–2013 period was 363 thousand (Gagauz, 2023). Ultimately, this method is used by the NBS for vital statistics evidence.

DATA AND METHODS

Using data on birth counts and age- and sex population with usual residence, we estimated the Moldovan migrant stock as of 1^st January 2016 and its trend until 2022. In the method used, we subtract the number of usual resident population on 1st January from its initial cohort size. To eliminate the effect of mortality, the life table probabilities of survival were applied. Thus, the arithmetic difference between the expected and de facto 1^st January population corresponds with the estimated migrant stock. The research includes the entire territory of Moldova, including the left bank of the Nistru River and Bender municipality.

The data on birth counts of the studied cohorts were retrieved from the NBS. The age- and sex population distribution were obtained from NBS and, by request, from the Transnistrian Statistical Office. Considering that in the Transnistrian region, the vital statistics include the permanently absent population, we adjusted its age- and sex structure according to net migration rates calculated based on data from the Moldovan Border Police.

To adjust the cohort size by survival rate, we used life tables of the generations of the corresponding year of birth calculated by Penina and colleagues (Penina et al., 2015).

Figure 1 shows the Lexis diagram that reveals the applied method, where: (1) c is the hypothetical cohort born in the year t; and (2) ct₁,x₁ and ct₂,x₂ are the population stocks of cohort c on 1^st January of the years t₁ and t₂ at the age x₁ and x₂. To exclude the effect of mortality on the generations studied, we adjusted the birth cohort size by survival rate.

Figure 1. The Lexis chart shows the method of estimation of the net migration within the hypothetical cohort c born in year t on 1^st January of years t1, t2 at age x1, x2

The estimation of the net outmigration of the 1980–1995 birth cohorts is made according to the following formula:

;

where (1) c_t is the initial cohort size; (2) p_x1,t1 is the probability of surviving from birth to the age _x1,t1; and (3) P_t1,x1 is the population of the respective cohort c_t on the 1st of January of the year _t1.

Table 1 illustrates the applied method of Moldovan migrant stock estimation. There is revealed the initial cohort size of the studied generations, the probability of their survival to 1^st January 2016, and the population with usual residency of the corresponding cohorts.

Table 1. Illustration of the method of estimation of migrant stock by using cohort size, probability of survival, and population with usual residency

Year of birth (1)	Cohort size (2)		Probability of survival to 1st January 2016 (3)		Population with usual residency on 1st January 2016 (4)		Expected cohort size on 1st January 2016, adjusted by survival rate (5)	Estimated migrant stock (6)
	Males	Females	Males	Females	Males	Females
1980	40687	38893	0.899	0.939	23541	23918
1981	42210	40069	0.898	0.940	23967	23718
1982	42751	40507	0.907	0.944	24688	24345
1983	46594	44710	0.913	0.949	27008	26648
1984	46234	43403	0.914	0.949	27248	26325
1985	46366	44087	0.922	0.953	28119	27356
1986	48821	45905	0.931	0.960	29597	28611
1987	47190	44572	0.940	0.960	29024	28139	(2) x (3)	(5) - (4)
1988	45473	43095	0.938	0.964	28432	27922
1989	42212	40009	0.942	0.966	27743	26851
1990	39499	37586	0.949	0.969	26683	26120
1991	36946	35074	0.953	0.969	25162	24895
1992	35727	33927	0.949	0.971	25367	25234
1993	33975	32204	0.953	0.970	24352	24245
1994	32030	30055	0.953	0.968	23780	23593
1995	29044	27367	0.958	0.972	22216	21421

Source: NBS; Tiraspol Statistical Office; Penina et al., 2015

The applied method encounters a degree of error, given the use of survival rates of the period life tables. However, the studied generations are still in their youth and have been minimally exposed to the risk of death.

RESULTS

Figure 2 displays the estimated number of Moldovan migrants of the 1980–1995 cohorts during the last available period. Thus, in 2016, the estimated stock of migrants born in 1980–1995 constituted about 380 thousand, out of which 193.4 thousand were males, and 186.1 thousand were females. Gradually, the estimated stock of the corresponding cohorts in males and females increased to 406.3 thousand in 2017, 432.7 thousand in 2018, 452.0 thousand in 2019, and 464.3 thousand in 2020. The observed increase in migrant stock is merely due to the entry of young generations into the age of migration activity. In contrast, the older generations emphasise a stagnation in outmigration mobility. However, a slight decrease in the stock of Moldovan migrants was observed at the beginning of 2021 and 2022 years, corresponding with the period of migration restriction during the COVID-19 pandemic.

The estimated stock of migrants emphasises a higher degree of male involvement in international migration compared with females in the analysed years. Nevertheless, the most significant gap in male/female migrant stock was observed within the 2017–2020 period, while it narrowed in 2021 and 2022.

Figure 2. The estimated stock of Moldovan migrants born in 1980–1995 as of 1st January 2016–2022, by sex

Source: NBS and Tiraspol Statistical Office

Among the studied cohorts, males and females (with a slight sex gap) registered a similar trend in changes in the stock of migrants during the 2016–2022 period. Generally, by 2016, the generations born in the early 1980s, to some extent, saturated their degree of involvement in international migration. The estimated stock of migrants of the 1980–1985 cohorts only slightly increased between 2016 and 2022, registering a decline in the last years (Table 2). To some extent, cohorts of the early 1980s are engaged in the process of the return migration, and their proportion in the overall stock of Moldovan migrants is slightly decreasing and likely will decline in the following years.

A significant stock of migrants in 2016 is also emphasised by the generations born in 1986–1991. However, compared with the generations born in 1980–1985, the 1986–1991 cohorts have shown a visible increase in migrant stock between 2016 and 2022 – the trend that could continue in the following years due to the high migration mobility in the young ages.

A notable increase in the 2016–2022 period in migrant stock is observed among the generations born in 1992–1995 that entered the active migration mobility ages in the period observed. However, compared to the 1980–1991 cohorts, the succeeding generations are less numerous, and a high degree of involvement in the outmigration of the population born after 1991 has a smaller impact on the changes in the migrant stock of the analysed cohorts.

Table 3 presents the proportion of Moldovan migrants of the 1980–1995 year of birth within the cohort size. Among the analysed generations, a visible disparity in the change of the cohorts’ sizes was emphasised due to the net migration of the population of the corresponding year of birth. Up to 2016, the generations born in 1980–1988 lost between 31–33% of the initial cohort size due to international migration. In the 2016–2022 period, the 1980–1988 cohorts registered a varied decline in their sizes because of international migration (between -2.1 and -6.9 percentage points in both sexes), where a higher loss in generations’ size was observed in younger ages. In the subsequent generations of the 1989–1991 years of birth, due to international migration, the cohorts’ size declined as 2016 by 27.2–28.5% in males and by 25.9–29.5% in females, and additionally reduced by 7.7–9.0 percentage points in males and by 7.7–9.3 percentage points in females in the 2016–2022 period. In 2016, the 1992–1995 cohorts emphasised the lowest decline in their size because of the net outmigration (between -18.9% and -23.9% in males and females), and, considering the high level of migration mobility of these ages, by 2022 the overall decline in the size of the corresponding cohorts constituted 31.3–33.3% in both sexes.