4.2 Regression Results
Based on the baseline results in Table 2 below, population diversity represented by the HIEF is positive and statistically significant at 1% to population growth. In model A, the fertility rate and net migration are positively correlated to population growth while the mortality rate is negatively correlated to population growth. All three of the previously mentioned control variables are statistically significant at the 1% level. Conversely, population density and sex ratio show no significant relationship to population growth. On a similar note, population diversity also shows a statistical significance at 1% level but negative relationship to income growth. In model B, population density, mortality rate, and sex ratio are negatively correlated with income growth at the 5%, 1%, and 5% significance level respectively. Fertility rate and net migration in model B do not show any significant relationship to income growth. Additionally, female labour force participation has also been shown to be statistically significant at 5% and having an inverse relationship with population growth in model A while showing no relationship at all with income growth in model B.
Table 2
System GMM Baseline Estimation Results
| Dependent: Population Growth | Dependent: Income Growth |
Diversity (HIEF) | 2.0661*** (0.7959) | -2.0308*** (0.56511) |
Log Population Density | 0.1818 (0.1750) | -0.2188** (0.1035) |
Fertility Rate | 0.8115*** (0.1815) | -0.1747 (0.1504) |
Mortality Rate | -0.0259*** (0.0085) | -0..0301*** (0.0079) |
Log Net Migration | 3.0876*** (1.1253) | 0.1849 (0.3058) |
Sex Ratio | -7.1887 (9.0649) | -12.7315** (5.0046) |
Female Labor Participation Rate | -0.0232** (0.0097) | 0.0042 (0.0084) |
Constants | -38.1278** (17.9830) | 22.1870*** (7.1313) |
Observations | 3946 | 2564 |
Countries | 150 | 127 |
Notes: Standard errors are shown in parentheses. ***, **, and * show significant level of 1%, 5%, and 10% respectively. |
The main findings indicate that ethnic diversity is statistically significant on both population growth and income growth. Diversity has a positive correlation with population growth and a negative correlation with income growth. The positive correlation of diversity with population growth aligns with Gören (2014) which concludes that diversity leads to higher population growth through fertility rates. Papyrakis & Mo (2014) provides an explanation for this phenomenon that a more diverse population could foster strategic interaction among ethnic groups to compete for influence within the society. To gain more influence, ethnic group leaders may promote pronatalism policy to increase the number of population within the ethnic group, thus elevating their position and dominance in the society. As a result, this will enhance the fertility rate and end up increasing the population growth in the highly diverse society. Similar explanation is also pointed out by Janus (2013) who argues that ethnic groups may promote higher fertility rate to increase their voting power relative to other groups. Janus (2013) also further specifies that weak political institutions are the key factor in the emergence of this condition, using fertility rates as the key to increase ethnic political power.
The negative correlation found in income growth is supporting the view that diversity inhibits economic development. The results are in line with Easterly & Levine (1997) who also identified a negative association between ethnic fragmentation and economic growth due to the frequent ethnic conflict occurring in the highly diverse region. Diverse society is prone to exhibit a propensity for rent-seeking behaviour, which hinders the effective formulation and implementation of sound public policies. This can lead to adverse consequences, such as poor education attainment, high financial debt, and low infrastructure quality, due to heightened levels of societal segregation. Gören (2014) also elucidated the adverse direct impact of ethnic diversity on economic growth and the indirect repercussions of ethnic polarization on economic outcomes through four channels, namely human capital accumulation, investment levels, trade openness, and the proclivity towards civil conflict.
In Gören (2014), it was stated that education, as one form of human capital accumulation, may not be properly and fairly distributed in an ethnically diverse environment due to the government’s intention of using education as a tool to control and influence particular ethnic groups. This situation may lead to lower schooling levels which negatively affects human capital and income growth. Easterly & Levine (1997) also noted that diverse societies may be less satisfied with the quality of education due to the disagreements between ethnic groups on the characteristics and aspects of the education itself, such as the language used and the learning materials which could lead up to less investment in human capital. Ogbu & Simons (1998) further pointed out that minority groups are more likely to become suspicious of the public schools that are considered to favor the ruling ethnic group and discriminate against the minorities, thus discouraging them to take on schooling.
The next channel, level of investment, has been partially explained by Easterly & Levine (1997) in the previous discussion. Countries with highly diverse populations are prone to rent-seeking behaviour by each ethnic group. These rent-seeking activities may generate conflict of interest and lead to a problem in reaching an agreement on the kind of public goods that will be provided to the society due to the disharmony between ethnic groups. As a result, this will hinder the development of public infrastructure, education institutions, healthcare, and government policy. This disruption would reduce the level of investment in the productive sector, which will weaken the economic growth and income growth. Ethnic diversity could also affect the economy through free trade channels. Regions with higher ethnic diversity are found to have lower quality of exported goods than the lower ethnic diversity regions due to the difficulties in communicating and collaborating between ethnic groups to produce high quality differentiated products (Luong, 2021). Lower quality of exported goods will weaken the export sector and lead to lower income from export activities. Another channel, which is civil war, has also been identified to hamper economic development and tends to happen in ethnically diverse countries. According to Collier (2001), civil war could rapidly escalate and cause human and physical capital destruction and also mass exodus of the educated population. The consequences of this would severely inhibit the country’s human development and the economy as the country is unable to function properly.
Furthermore, an increased level of ethnic diversity can lead to fierce competition for jobs among different ethnic groups. In an ideal scenario, this heightened competition should act as a filter for employers to recruit the most competent applicants. As a result, this competitive environment can be beneficial for highly skilled workers. However, those who are unsuccessful in this competition and have already relocated to these regions may find themselves settling for lower-tier jobs, which subsequently leads to lower earnings according to the statistical discrimination theories (Horvath & Huber, 2019). Additionally, DiRienzo et al. (2007) found that countries with higher ethnic diversity tend to be less competitive. Consequently, lower competitiveness means having lower productivity and this causes a country’s likelihood of sustained growth to be lower than more competitive countries, also causing the quality of labour to also be less competitive.
Based on the estimation result, another interesting point should be highlighted. Previous discussion has stated that ethnic diversity can have a positive or negative effect on the economy depending on the variable that is used as a measure (Rodríguez-Pose & von Berlepsch, 2019). Empirically, ethnic fractionalization measure is generally related to higher economic growth, while ethnic polarization measure is frequently related to lower economic growth (Ananta et al., 2023). The results of our estimation that shows a negative correlation between ethnic diversity, as measured by ethnic fractionalization index, and economic growth do not support the general conclusions that are found in the literature.
The baseline estimation results above do not describe the full picture as it only shows how population growth and income per capita growth (GNI per capita) is explained by population diversity and other control variables at a global scale. To obtain a more accurate description of the aforementioned condition, we show how population diversity relates with population growth and income per capita growth by categorizing it based on the World Bank historical classifications by income as of 2013, following the latest available data for this research.
Table 3
System GMM Estimation Results with Income Level Classification
| Dependent: Population Growth | Dependent: Income Growth |
Diversity (HIEF) | 2.4310*** (0.8265) | -0.8350** (0.3960) |
Log Population Density | 0.2108 (0.1620) | -0.0929** (0.0444) |
Fertility Rate | 0.8962*** (0.1712) | -0.0684 (0.0881) |
Mortality Rate | -0.0180** (0.0073) | -0.0081** (0.0041) |
Log Net Migration | 2.8950*** (1.0204) | 0.1345 (0.1253) |
Sex Ratio | -8.9130 (8.8632) | -12.5981*** (3.2192) |
Female Labour Participation Rate | -0.0183** (0.0083) | 0.0148** (0.0061) |
Interaction Variable: Middle-Income to Low-Income Countries | 0.8684* (0.4796) | 1.3814*** (0.3566) |
Interaction Variable: High-Income to Low-Income Countries | 1.6536*** (0.5497) | 3.0056*** (0.3543) |
Constants | -17.6425** (6.5861) | 14.0091** (5.3010) |
Observations | 778 | 471 |
Countries | 150 | 127 |
Notes: Standard errors are shown in parentheses. ***, **, and * show significant level of 1%, 5%, and 10% respectively. |
Table 3 above shows the relationship between population diversity and population growth as well as population diversity and income growth with the addition of a categorical variable based on countries income level, which also acts as a robustness measure of our baseline estimation. With the addition of a categorical variable, the results in model C illustrate that population diversity is statistically significant to both population growth and income growth which is consistent with the previous baseline estimation result. Other independent variables in model C consisting of fertility rate, net migration, as well as the high-income categorical variable are positive and statistically significant towards population growth while mortality rate and female labour force participation are negative and statistically significant towards population growth as the dependent variable. Conversely, population density, sex ratio, and middle-income categorical variable are shown to not have any significant relationship with population growth. On the other hand, the results in model D show that population density, mortality rate, and sex ratio show a negatively significant relationship with income growth as the dependent variable while female labour force participation, middle-income, and high-income categorical variable are positively significant towards income growth. Fertility rate and net migration are the independent variables that are not statistically significant towards income growth as the dependent variable.
Our estimation that involves the income class categorical variable highlights the importance of the significant relationship between diversity and population growth as this particular situation has been plaguing high-income economies, which may potentially cause social and economic problems if left alone (Peterson, 2017). Model C in Table 3 shows a positive significant relationship between diversity and population growth at the 1% level. At the same time, fertility rate, mortality rate, and net migration respectively are all showing statistical significance towards population growth. Fertility rate and net migration exhibit positive relationships towards population growth. It is also worth noting that female labour force participation causes a positive income growth and this relationship is statistically significant at the 5% level.
The estimation results also reveal a consensus across countries at all income levels that net migration is correlated with a positive relationship towards population growth. On the other hand, it has been proven that the historical decline in world population due to falling fertility rates occurred in response to higher levels of economic development (Murdoch, Chu, Stewart-Oaten, & Wilber, 2018). In general, the high population growth often found in low-income economies and the low population growth in high-income economies are likely to cause social and economic inequalities as well as inhibit development efforts (Peterson, 2017). While international migration may help in adjusting these imbalances, such policies are often opposed by many and quite unpopular (Peterson, 2017). The primary reason for the unpopular view of migration is, among others, due to the perceived low labour quality of migrants, political, and sociocultural differences. However, such views may not always reflect the actual migrants’ economic contributions. In fact, the perceived negative effects of immigrants are often baseless and create a dilemma as most migrants’ destination countries could benefit with more human capital and expertise brought by immigrants (OECD/ILO, 2018). Therefore, increased efforts to improve the overall well-being of less wealthy populations across the world are vital to solving the dilemma between the perceived low labour quality of migrants and plummeting population of a country (Murdoch et al., 2018; OECD/ILO, 2018). Additionally, asylum process claims which are commonly faced by refugee migrants, who are most consistently considered as low labour quality on arrival, should be kept short as this has a strong impact on their future labour outcomes and facilitating them to join the labour market at the earliest possible stage helps to minimize skill losses and increases the effectiveness of human capital investment (Brell, Dustmann, & Preston, 2020). This is particularly relevant in light of Gören’s (2014) findings as prolonged asylum processes may exacerbate disparities in human capital accumulation through education in ethnically diverse environments, leading to an inequitable distribution of resources due to government policies targeting specific ethnic groups.
We also found that female labour force participation is statistically significant and positive towards income growth implying that there has indeed been a shift in paradigm of how people in high-income countries think about posterity and labour market conditions. Initial fertility models in the past decades across many countries were developed to sufficiently explain two consistent trends: lower fertility rates mean higher income and higher women's workforce participation (Doepke et al., 2022). However, our finding on female labour force participation indicate that these facts seem to be changing in recent times as some observation by Doepke et al. (2022) has confirmed the positive relationship between fertility and income, implying that women are more active in the workforce while still maintaining the same or even achieve higher level of income. Furthermore, such trends are also supported by flexible labour market conditions with favorable family policies, especially towards women, and changing social norms with fathers caring for their children.
Policymakers should consider the effects of population diversity on population growth and income growth when designing and implementing policies that affect these outcomes. Adopting a dynamic and context-specific approach to those three aspects is also important as these factors may vary over time and across countries. For example, policies that aim to expand the population of advanced economies or reduce the population growth of developing economies may need to account for the different preferences, behaviours, and incentives of diverse ethnic groups in society. It is also worth pointing out that policies may work well only in certain countries or certain time periods which further reinforce the need for policymakers to assess and adjust their policies accordingly. Additionally, policymakers’ awareness should also be raised regarding the potential challenges and opportunities that population diversity brings to social cohesion, political stability, and economic development. For instance, policies that promote interethnic cooperation, integration, inclusion, and equitable distribution of income may help to mitigate the negative effects of ethnic conflict, rent-seeking, and policy inefficiencies in highly diverse settings. Meanwhile, policies that foster ethnic diversity, recognition, and representation may help to enhance the positive effects of cultural diversity, innovation, and creativity in society.