Migrant networks effects in migration decision in communities from - PDF document

Migrant networks effects in migration decision in communities from different regions of origin in Mexico: An Agent-Based Model approach Mauricio Rodríguez Abreu Introduction The purpose of this study is to analyze the effect of migrant networks in migration decision-making patterns in communities from different regions in Mexico. As social networks are extended and strengthened with every additional migrant, the effect of the network increases by facilitating and encouraging migration (Haug, 2008). The theory of planned behavior applied to migration decision-making states that migration intentions are the main determinants of migration behavior, but this effect is mediated by individual human capital, household and community characteristics (Jong, 2010). However, when migration in the community is well established, individual and household and community characteristics lose importance, and migration becomes generalized and hard to detain (Winters et al., 2001). A different approach to this process is provided by the use of a diffusion perspective, where the influence of migrants in the close network will determine the likelihood of engaging in the migration dynamics. The study will approach the study of a diffusion perspective in migration, which can be use in the development of new theoretical approaches in the study of the Mexico-US migration system. Entities, state variables and scales The model includes Mexican male and females individuals with the next characteristics. Agents: Age and sex. Network: number if network links with migration experience, previous migration experience.

Assumptions . Individual level characteristics: Age. Agents are only prone to migrate at working ages. Individuals in the model age, but exposure to migration occur only between ages 15 and 64. Gender . The likelihood of migration is different for men and women. a) Agents’ characteristics affect the likelihood of migration (rates). b) Migration of individuals in the network of agents affects migration preferences. c) As more agents migrate in the community, increases of migration prevalence directly affects migration decisions. The process of decision-making process follows a diffusion perspective where migration of other agents in their network and the total level of migration in the community affect the likelihood of becoming a migrant. Are agents heterogeneous? Agents are heterogeneous and their decision-making process depends on their sex, age and migration in their network. a) Age. Only population at working ages is prone to become a migrant. Working ages among this population will be assume to be 15 to 54 years; as a consequence, children under 15 and elderly population aged 65 and above, will have immunity to migration. b) Gender. Diffusive effects will be different for males and females and depending of the sex of the migrant; i.e., non-migrant females will be less likely to become migrants, however, female migration will result in increasing propensity to migrate for women. As a consequence, the diffusion effects will be stronger among same-sex agents. c) Type of migration. Circularity of migration will be reflected in the model. Return migrants will increase the likelihood of a non-Migrant to become migrant and return migrant as well.

Stochasticity : Initialization in the model assumes that the population is uniformly distributed throughout all possible ages and that male/female ratio is 1. Also, the links in the population are randomly created. 3. Data source and software The model is set to work with any data available. However, in order to test the reliability of the model, data from the Mexican Migration Project (MMP) 143 will be used. The MMP contains information regarding migration experience for individuals from 143 communities in Mexico and gathers information about migration of members of the household and close kin networks. NetLogo 5.0.5 (Wilensky, 1999) was used to operationalize the model and the networks relationships. 4. Model The model was based on the NetLogo Virus on a Network (Stonedahl & Wilensky, 2008). Individuals and networks are randomly created. However, once the network is set, no changes are allowed. Assume the in a small network with five agents, agent 1 becomes a migrant. The next step is to estimate the effect of this migrant in the likelihood of migration for all other 4 agents. After some iterations, agent 3 becomes a migrant as well. As a consequence, agent 4 will be now affected by agents A1 and A3; increasing the likelihood of becoming a migrant himself.

A2 A2 A3 A3 A1 A1 A4 A4 A5 A5 A2 A2 A3 A3 A1 A1 A4 A4 A5 A5 The model differentiates between migrants and return migrants. The underlying assumption is that these types of migration will result in different effects in increasing the likelihood of migrating for the members of the household. 5. Results. The findings presented in this section are preliminary results. In this situation the initial parameters in the model are: Population: 300 Average network size: 5 Initial migrants (as percentage of the population): 10% Migration spread change: 20% Return change: 25.3% Birth rate: 20 Death rate: 4 Life-expectancy (Maximum life span): 80 years.

i) Initial network distribution: ii) After 15 ticks. iii) After 180 ticks iv) After 715 ticks Plots

6. Final remarks. The present model provides a preliminary approach to the analysis of the diffusive mechanisms operating in the migration process. This approach represents a significant contribution since it will be possible to model different characteristics of the population by modulating the initial parameters and specific rates in which migration spreads among the community. Further work will improve the model by including specific rates from the data source (not included in this analysis) and by modeling the age distribution of specific populations. 7. References Haug, S. (2008). Migration networks and migration decision-making. Journal of Ethnic and Migration Studies , 34 (4), 585-605. De Jong, G. F. (2000). Expectations, gender, and norms in migration decision- making. Population studies , 54 (3), 307-319. Stonedahl, F. and Wilensky, U. (2008). NetLogo Virus on a Network model. http://ccl.northwestern.edu/netlogo/models/VirusonaNetwork. Center for Connected Learning and Computer-Based Modeling, Northwestern University, Evanston, IL. Winters, P., De Janvry, A., & Sadoulet, E. (2001). Family and community networks in Mexico- US migration. Journal of human resources , 159-184. Wilensky, U. (1999). NetLogo. http://ccl.northwestern.edu/netlogo/. Center for Connected Learning and Computer-Based Modeling, Northwestern University, Evanston, IL.