The Impact of Climate Change on Migration Patterns in Coastal Communities

Abstract

Climate change is a major global challenge affecting migration patterns, particularly in coastal communities vulnerable to sea-level rise, flooding, and extreme weather. Pakistan, with its extensive coastline and diverse environmental conditions, faces significant climate-induced migration issues, especially in Karachi, Thatta, Gwadar, Badin, and Muzaffargarh. This study aims to investigate the impact of climate change on migration patterns in these five selected regions of Pakistan. By analyzing climate variables and socio-economic factors, the research seeks to provide a localized understanding of how climate change drives population movements. A cross-sectional survey design was employed to gather data from 350 participants across these regions. Stratified random sampling ensured representation from each area, and data were collected using a structured questionnaire administered online. Statistical analyses included multiple linear regression, logistic regression, and structural equation modeling (SEM). This study found a strong positive relationship between climate change variables (sea level rise, temperature increases, and flooding) and migration patterns. Both direct impacts of climate change and indirect socio-economic factors influenced the likelihood of migration. The SEM analysis revealed that climate awareness partially mediates the relationship between climate change and migration. In conclusion, climate change significantly drives migration in Pakistan’s coastal communities, with both direct environmental impacts and socio-economic conditions playing crucial roles. Enhanced climate awareness and comprehensive adaptation strategies are essential. Policies should focus on climate resilience through infrastructure improvements, early warning systems, and socio-economic support programs. Strengthening education and economic opportunities is vital to build community resilience and effectively manage climate-induced migration.

1. Introduction

Climate change is one of the most pressing global challenges of our time, profoundly influencing various aspects of human life, including migration patterns. The impact of climate change on migration is evident worldwide as rising temperatures, shifting weather patterns, and increasing frequency of extreme events such as hurricanes, floods, and droughts drive people to relocate [1]. Coastal communities, in particular, are highly vulnerable due to their exposure to sea-level rise, storm surges, and coastal erosion, leading to significant displacement and migration.

Globally, coastal regions are facing severe threats from climate change. Rising sea levels threaten to inundate low-lying areas, while increased storm intensity exacerbates flooding and infrastructure damage [2]. These changes force many individuals and communities to migrate either temporarily or permanently, seeking safer and more stable living conditions. The situation is compounded in developing countries, where adaptation and resilience capacities are often limited, making migration a crucial adaptation strategy [3,4].

In Pakistan, the impact of climate change on migration patterns is increasingly becoming a critical issue. The country’s extensive coastline, stretching over 1000 km along the Arabian Sea, is particularly susceptible to the adverse effects of climate change [5]. Karachi, as Pakistan’s largest coastal city, faces significant challenges from rising sea levels and frequent heatwaves. Similarly, districts like Thatta and Badin experience severe flooding and saline intrusion, disrupting local livelihoods and prompting migration [6].

Moreover, areas like Gwadar are experiencing rapid development, adding another layer of complexity as the city grapples with climate-induced migration while undergoing significant economic changes. Muzaffargarh, although not a coastal area, is affected by changes in river dynamics and flooding patterns that can influence migration patterns indirectly. These regions collectively represent the diverse ways in which climate change impacts migration in Pakistan, from direct coastal impacts to indirect effects through altered environmental conditions.

2. Literature Review

Climate change has emerged as a significant driver of migration, with substantial empirical evidence highlighting its impacts globally [7]. According to reports from the Intergovernmental Panel on Climate Change (IPCC), rising global temperatures, sea levels, and the increasing frequency of extreme weather events are causing large-scale population movements (IPCC, 2021) [8,9,10]. For instance, Magnan and Oppenheimer’s research [11] emphasizes that low-lying coastal areas are highly vulnerable to sea-level rise, compelling communities to migrate inland. Similarly, Kakinuma and Puma [12] found that severe flooding in Southeast Asia has driven higher migration rates, showcasing the global pattern of climate-induced displacement [13].

In sub-Saharan Africa, Szaboova’s study [14] demonstrates that drought and desertification are major factors pushing people to migrate, as they erode agricultural productivity and force communities to seek alternative livelihoods [15]. Koubi and Stoll’s research [16] explored responses to extreme weather events, noting that both sudden and gradual climate impacts lead to different migration outcomes, ranging from temporary displacement to permanent relocation [17,18].

In Pakistan, climate change is increasingly influencing migration, particularly in coastal and flood-prone areas. Rahman and Rahman [19] document how rising sea levels and more frequent heatwaves in Karachi are prompting internal migration, with residents moving to cooler and safer regions [20]. Similarly, Salik and Jahangir’s study [21] illustrates that communities in Thatta and Badin are frequently displaced due to flooding and saline intrusion, which devastate agricultural livelihoods.

Ahmed and Padda’s research [22] highlights the direct link between climate variability and migration patterns in Karachi, with residents relocating from areas most affected by sea-level rise and heat stress. In Gwadar, Wu and Liu [23] disclosed that rapid urban development, compounded by climate impacts, is creating complex migration patterns as people seek better economic opportunities and more stable living conditions.

Focusing on five key areas—Karachi, Thatta, Gwadar, Badin, and Muzaffargarh—provides a localized view of how climate change affects migration in Pakistan. Studies such as Sulehria’s [24] reveal that Karachi faces significant migration due to its vulnerability to heatwaves and flooding. Memon and Aamir’s research [25] shows that frequent flooding in Thatta has led to substantial internal migration, disrupting community structures and local economies. Badin faces similar issues, with ABBAS [26] documenting increased displacement due to saline intrusion and water scarcity.

In Gwadar, Hussain and Shuai [27] examine how rapid economic development, alongside climate impacts, shapes migration patterns as people move in search of better opportunities and safer environments. Although Muzaffargarh is not a coastal region, it experiences considerable migration due to river flooding and shifting environmental conditions, as reported by Ahmad and Afzal [28].

While there is extensive research on climate-induced migration globally and in Pakistan, several gaps remain. Many studies focus broadly on national trends, often overlooking specific regional dynamics and how local socio-economic factors intersect with climate impacts. There is also a scarcity of granular research linking specific climate variables to migration patterns in lesser-studied areas like Muzaffargarh. Additionally, more longitudinal studies are needed to track migration over time and assess the long-term impacts of climate change at macro and micro levels.

3. Statement of the Problem

This study aims to examine the impact of climate change on migration patterns in five selected areas of Pakistan: Karachi, Thatta, Gwadar, Badin, and Muzaffargarh. By analyzing specific climate variables and socio-economic factors driving migration in these regions, the research seeks to provide a localized understanding of climate-induced population migration within Pakistan.

Climate change presents significant risks to human settlements, especially in vulnerable coastal and flood-prone areas. With its extensive coastline and diverse environmental conditions, Pakistan faces multiple challenges from climate change. However, there is limited research on the specific migration dynamics across different provinces/areas of the country. This study addresses this gap by focusing on five distinct areas, offering crucial insights for effective policy-making and adaptive strategies to mitigate the impact of climate change on migration.

The motivation for this research stems from the Sustainable Development Goals-13 (climate action) in collaboration with growing global mitigation practices to tackle the duress of climate change, which further perpetuates the migration of inhabitants in particular. The socio-economic disruptions caused by climate-related events, combined with the scarcity of localized studies, drive the need for this investigation. The findings aim to contribute to the existing stock of knowledge that informs resilient and adaptive responses to climate change at both local and national levels.

This study is vital in addressing a critical gap in the literature by providing empirical evidence on how climate change affects migration while mitigating vivid results-based policy interventions to ensure sustainable resilience and adaptation measures for community living in coastal regions. The novelty of this exploration presents a unique picture by investigating diverse areas of coastal and inland regions in association with socio-economic status and environmental dynamics. Lastly, by integrating quantitative data with qualitative assessments, this study offers a comprehensive perspective on the multifaceted impacts of climate change on migration, providing practical measures for mitigating the effects of climate change on the lives of inhabitants at macro and micro levels.

Conceptual Framework of the Study

The conceptual framework of this study examines the intricate relationships between climate change variables and migration patterns within coastal communities in Pakistan. The framework identifies several independent variables that directly impact migration, including sea level rise, temperature increases, flood frequency, extreme weather events, saline intrusion, and economic development. These climate change factors are hypothesized to exert a significant influence on the movement of populations as individuals and communities seek to escape the adverse effects of these environmental changes. The framework also incorporates mediating variables, namely socio-economic impact and climate awareness, which are expected to play a crucial role in shaping migration patterns. Climate change impacts the socio-economic conditions of affected areas, which in turn drives migration. Additionally, increased climate awareness can influence migration decisions as individuals become more informed about the risks and opportunities associated with changing environmental conditions. Furthermore, the framework acknowledges the moderating role of socio-demographic variables such as age, gender, education level, and socio-economic status. These factors are crucial in understanding the variability in migration responses among different population groups. For instance, younger individuals or those with higher educational levels may have different migration patterns compared to older individuals or those with less education. Similarly, economic status can significantly impact the ability and decision to migrate. To analyze these complex relationships, this study employs multiple linear regression to assess the direct effects of climate change variables on migration patterns. Logistic regression is used to examine the influence of both climate change and socio-demographic variables on the likelihood of migration. Structural equation modeling (SEM) is applied to evaluate both direct and indirect effects, considering the mediating roles of socio-economic impact and climate awareness.

The graphical representation in Figure 1 of the conceptual framework illustrates the impact of climate change on migration patterns in coastal communities of Pakistan. The diagram showcases the direct and indirect relationships between climate change variables and migration patterns, as well as the moderating effects of socio-demographic variables.

Figure 1 representing the relationships between climate change variables, migration patterns, socio-economic impacts, climate awareness, and socio-demographic factors. Each arrow illustrates the direct, indirect, or moderating effects, as per the framework.

4. Materials and Methods

4.1. Research Design

A cross-sectional survey design was used for this study, as it allows for the collection of data at a single point in time, providing a snapshot of the current impact of climate change on migration patterns [29].

4.2. Study Setting and Target Population

The study’s surveys were conducted in Karachi, Thatta, Gwadar, Badin, and Muzaffargarh, representing diverse environmental and socio-economic contexts in Pakistan. Karachi faces significant climate risks as a major coastal metropolis, while Thatta and Badin are vulnerable to flooding and saline intrusion. Gwadar, undergoing rapid development, is also affected by climate impacts, and Muzaffargarh, though inland, experiences river flooding.

The potential respondents consisted of households who had either experienced or were at risk of climate-induced migration. This selection ensured that participants directly reflected the study’s focus, capturing valuable insights into how climate change drives migration across varied settings.

4.3. Demographic Characteristics of Participants

Table 1. Socio-economic and demographic characteristics of participants.

Variable	Categories	Sample	Sample Size
Age	20–33 34–46 47–60 61 and Above	73 105 103 69	350
Gender	Male Female	203 147	350
Educational Level	No Formal Education Primary Education Secondary Education Higher Education	64 102 108 76	350
Socio-Economic Status	Low Income Middle Income Higher Income	113 136 101	350

Source: compiled by authors.

presents the socio-economic and demographic characteristics of the respondents while mitigating the yardstick of the study objectives. Understanding age demographics is crucial because climate change impacts and migration patterns may vary across different age groups. For instance, younger individuals might migrate for economic opportunities, while older individuals might be more affected by immediate climate impacts.

The age distribution of participants ranges from 20 to above 60 years, with the largest group being those aged 34–46 (105 participants). Moreover, gender disparities in migration patterns can reveal how climate change affects men and women differently. Women might face different challenges compared to men, such as access to resources or decision-making power, which can influence migration decisions. Thus, a sample includes 203 males and 147 females. In addition, education levels can impact individuals’ resilience to climate change and their ability to migrate. Higher education might correlate with better job opportunities and greater flexibility in migration decisions. The respondents education levels range from no formal education to higher education, with the majority having secondary education (108 participants). Moreover, those in lower income brackets might face greater challenges in relocating or adapting to climate impacts [30]. The socio-economic status of participants is divided into low-, middle-, and higher-income groups. With 113 participants in the low-income bracket, 136 in the middle-, and 101 in the higher-income group, this variable is essential for understanding how financial resources influence migration patterns.

Overall, the socio-economic and demographic characteristics highlighted in the Table 1 provide a foundation for analyzing how various factors, such as age, gender, education, and socio-economic status, influence migration patterns in response to climate change in coastal communities of Pakistan.

4.4. Sampling and Sample Size

For this study, the researchers used a stratified random sampling technique based on Sekaran’s method, incorporating multiple stages to determine the sample size [31]. The details of the stages are provided below.

Stage 1: Determine the Target Population: Define the target population and its characteristics. For this study on the impact of climate change on migration patterns in coastal communities of Pakistan, the target population includes households and individuals in five coastal communities—Karachi, Thatta, Gwadar, Badin, and Muzaffargarh—who have experienced or are at risk of climate-induced migration. The total estimated population in these areas is approximately 2,000,000 individuals.

Stage 2: Calculate the Sample Size: To calculate the sample size of 350 from a target population of approximately 2,000,000, statistical methods involve several key steps. First, parameters such as the confidence level, margin of error, and proportion of the population exhibiting the characteristic of interest are defined. For a 95% confidence level, the Z-score is 1.96, and a margin of error of 5% (0.05) is commonly used. Assuming a maximum variability proportion of 0.5, the initial sample size is calculated using the formula n₀ = Z²⋅P⋅(1 − P)/E². Substituting the values, n₀ = 1.96²⋅0.5⋅0.5/0.05² results in approximately 385. Since the population is finite, this initial sample size is adjusted using the finite population correction. With n₀ being 385 and N being 2,000,000, the adjusted sample size remains close to 385 after correction. Finally, a sample size of 350 is chosen to balance statistical power with practical considerations, ensuring that the sample is both representative and feasible for the study on climate change impacts on migration patterns.

Stage 3: Perform Stratified Sampling: The population is divided into strata based on the five coastal communities of Pakistan: Karachi, Thatta, Gwadar, Badin, and Muzaffargarh. This approach ensures that each region is adequately represented in the study, capturing regional variations in climate change impacts and migration patterns. By stratifying the population, this study can account for the diverse environmental and socio-economic conditions across these regions, leading to a more comprehensive and accurate analysis of how climate change affects migration in different areas.

Stage 4: Allocate Samples Proportionally: The total sample size is allocated proportionally based on the population size of each region, ensuring that the sample reflects the distribution of the population across different areas (

Table 2. Sample frame distribution by region.

Region	Population	Sample Size
Karachi	1,500,000	263
Thatta	150,000	26
Gwadar	100,000	18
Badin	150,000	26
Muzaffargarh	100,000	17
Total	2,000,000	350

Source: Compiled by authors in relation to the Sekaran and Bougie method.

). For example, Karachi, with the largest population of 1,500,000, is assigned 263 samples to accurately represent its significant share of the total population. This approach maintains a balanced and representative sample by giving larger sample sizes to regions with larger populations. Similarly, Thatta, with a population of 150,000, receives 26 samples, while Gwadar, with 100,000 people, is allocated 18 samples. Badin, with a similar population size to Thatta, is also assigned 26 samples. Muzaffargarh, having a population of 100,000, receives 17 samples. This proportional allocation ensures that each region is represented according to its population size, enhancing the accuracy and generalizability of the study’s findings.

Stage 5: Implement Random Sampling: Within each stratum, participants are randomly selected to ensure unbiased representation. This method minimizes selection bias by giving every individual an equal chance of being included in the sample. Random sampling within each region helps to eliminate selection bias and ensures that the sample accurately reflects the population of each region

4.5. Data Collection Tool

For this study, a structured questionnaire was used to collect quantitative data systematically and efficiently. This tool was appropriate for measuring specific variables and facilitating data analysis. The questionnaire included sections on demographic information such as age, gender, education, and socio-economic status; climate change perceptions, including awareness and experiences of climate change impacts; migration patterns, covering migration history, reasons for migration, and destination preferences; and socio-economic impacts, focusing on effects on livelihoods, housing, and community structures. Administered online to ensure wider reach and convenience for participants, the variables were measured using Likert scales, binary choices, and multiple-choice questions. Indices were created for climate change impacts, migration drivers, and socio-economic effects to facilitate analysis. Data collection was conducted over a period of 35 days in March 2024. Ethical considerations include informed consent, ensuring confidentiality and anonymity of participants, and providing the option to withdraw from the study at any time.

4.6. Reliability and Validity

Table 3. Reliability and validity statistics.

Variable	Measurement Tool	Reliability (Cronbach’s Alpha)	Validity (Expert Review)
Climate Change Perceptions	Likert scale items	0.85	Validated by experts
Migration Patterns	Multiple-choice and binary questions	0.80	Validated by experts
Socio-Economic Impacts	Likert scale items	0.83	Validated by experts

Source: compiled by authors in relation to field (2013).

presents the reliability and validity statistics for the measurement tools used in the study. The table includes three variables: climate change perceptions, migration patterns, and socio-economic impacts. Climate change perceptions refer to individuals’ understanding, beliefs, and awareness about the causes, impacts, and severity of climate change. Migration patterns refer to the movement trends, motivations, and decisions of individuals or families migrating due to climate-related factors such as sea-level rise, extreme weather events, and flooding. Socio-economic impacts refer to the effects of climate change on the economic and social well-being of individuals and communities, including changes in income, employment, education, and health conditions.

Each variable’s measurement tool is specified, with Likert scale items used for climate change perceptions and socio-economic impacts, and multiple-choice and binary questions for migration patterns. The reliability of these tools is measured using Cronbach’s alpha, with values of 0.85 for climate change perceptions, 0.80 for migration patterns, and 0.83 for socio-economic impacts, indicating strong internal consistency. The validity of the measurement tools was confirmed through expert review for all variables [32].

4.7. Data Analysis

Data were analyzed using SPSS version 21 statistical software. Statistical analyses, including multiple linear regression, logistic regression, and structural equation modeling (SEM), were used to examine the relationships between climate change variables and migration patterns. The variables are measured using a structured questionnaire with various types of questions tailored to capture quantitative data.

4.8. Models of the Study

4.8.1. Multiple Linear Regression Model

Model Specification: The multiple linear regression model investigates the relationship between migration patterns (DV) and climate change with four sub independent variables: sea level rise, temperature increases, flood frequency, and development.

Model Equation:

Y =β0 + β1X1 + β2X2 + β3X3 + β4X4 + ϵ

where:

Y = migration patterns: Measured through Likert-scale multiple-choice items and binary questions about migration history (whether migration occurred), reasons for migration (e.g., climate events like floods, economic reasons), and future intentions (whether they plan to move due to climate change).

X1 = sea level rise: Measured through Likert-scale items that assessed participants’ perceptions of the frequency and severity of sea level rise over the past decade. Respondents rated the extent to which they believed sea level rise had impacted their livelihoods, housing, or community on a scale ranging from 1 (no impact) to 5 (severe impact). This approach is appropriate for understanding subjective perceptions of environmental changes, particularly in coastal areas.

X2 = temperature increases: Measured using Likert scales, respondents indicated their experiences with increasing temperatures, particularly during peak seasons, and its effects on health, agriculture, and daily life. The questions ranged from 1 (no noticeable increase) to 5 (significant increase). This variable is crucial for identifying how changes in climate influence migration, as rising temperatures can push populations to move due to agricultural disruptions or heat stress.

X3 = flood frequency: It is measured using binary (Yes/No) questions and Likert scale questions assessing how often the participants’ regions had experienced flooding events and how severe the impact was. Respondents could report whether floods had caused migration, loss of housing, or employment, with scales ranging from 1 (no flood events) to 5 (frequent, destructive floods). The unit of measure reflects both the occurrence of floods and their intensity, making it useful for linking climate extremes to migration decisions.

X4 = economic development: Participants were asked multiple-choice questions to assess their perceptions of local economic development, with choices ranging from poor (no development) to strong (high development and infrastructure growth). This variable captures economic opportunities and infrastructural growth, which are essential factors in whether people stay in their communities or migrate in search of better living conditions.

β0 = constant

β1, β2, β3 and β4 = regression coefficients

ϵ = Error Term

4.8.2. Logistic Regression Model

The logistic regression model expands the analysis of migration patterns by assessing the likelihood of migration as a binary outcome (migrated vs. not migrated) based on climate change factors and socio-demographic variables. Unlike the multiple linear regression model, which explores the magnitude of migration patterns as a continuous variable, logistic regression focuses on predicting the probability of migration, which is crucial when the outcome is categorical.

The addition of socio-demographic variables—such as age, gender, education level, and socio-economic status—offers a deeper understanding of how personal characteristics influence migration decisions. These variables were included in the logistic regression model because migration decisions are often influenced by both external environmental factors (e.g., sea level rise, flooding intensity) and individual socio-demographic characteristics. For instance, younger individuals or those from lower-income groups may be more likely to migrate in response to climate stressors due to fewer resources or different risk perceptions.

Moreover, logistic regression allows the computation of odds ratios (Exp(β)), which quantify how much each independent variable increases the odds of migration. This helps capture both the impact of climate change and personal factors on migration decisions in a comprehensive way. The inclusion of these additional explanatory variables enhances the model’s ability to identify groups most at risk, making it more suitable for policy implications, as demographic factors often shape vulnerability and migration behavior.

Model specification: the logistic regression model assesses the likelihood of migration based on climate change variables and socio-demographic factors.

Model equation:

log(P(Y = 1)/1 − P(Y = 1)) = β0 + β1X1 + β2X2 + β3X3 + β4X4 + β5X5 + β6X6 + β7X7 + β8X8

(1)

where:

Y = Migration (1 = Migrated, 0 = Not Migrated)

X1 = sea level rise: as sea levels rise, saltwater is pushed further inland, leading to the contamination of freshwater resources.

X2 = temperature increase

X3 = flooding intensity: this refers to the severity of the flood in terms of factors such as water depth, speed of floodwaters, and the damage potential. High-intensity floods are more destructive, with larger volumes of water, higher flow rates, or greater areas affected, regardless of how frequently they occur.

X4 = saline intrusion: also known as saltwater intrusion, this refers to the process where saltwater from the ocean moves into freshwater aquifers or coastal water systems, contaminating drinking water, agricultural lands, and ecosystems. Key causes of saline intrusion include the following: sea level rise (X1), groundwater overuse, and climate change.

X5 = age (35–45)

X6 = gender (female = 1)

X7 = education level

X8 = socio-economic status (low income)

4.8.3. Structural Equation Modeling (SEM) Model

Model specification: the SEM model evaluates the direct and indirect effects of climate change on migration patterns, considering socio-economic impact and climate awareness as mediating variables.

Model pathways and equations:

Direct effect:

Migration patterns = β1 (climate change) + ϵ

Indirect effects:

Socio-economic impact = β2 (climate change) + ϵ

Climate awareness = β3 (climate change) + ϵ

Migration patterns = β4 (socio-economic impact) + ϵ

Migration patterns = β5 (climate awareness) + ϵ

Moderating effect:

Socio-economic impact = β6 (climate awareness) + ϵ

4.9. Methodology Limitations and Their Mitigation

This study’s limitations include reliance on self-reported data, which may introduce bias, and a cross-sectional design that captures a single time point. To address these issues, the research employed stratified random sampling to enhance representativeness and utilized validated instruments for data collection, improving reliability. Additionally, the combination of multiple regression and structural equation modeling provided a more nuanced understanding of relationships, allowing for a comprehensive analysis of the impacts of climate change on migration patterns.

5. Results

5.1. Multiple Linear Regression Analysis

The multiple linear regression analysis presented in

Table 4. Multiple linear regression analysis of migration patterns (DV) and climate change (IVs).

Variable	Coefficient B	SE	t-Value		p-Value	95% CI	VIF
Intercept	1.50	0.25	6.00		<0.001	[1.00, 2.00]	-
Sea Level Rise	0.85	0.10	8.50		<0.001	[0.65, 1.05]	1.20
Temperature Increases	0.75	0.12	6.25		<0.001	[0.51, 0.99]	1.30
Flood Frequency	0.60	0.15	4.00		<0.001	[0.30, 0.90]	1.10
Economic Development	0.50	0.20	2.50		0.013	[0.11, 0.89]	1.05
Territory Fixed Effects
Karachi	0.30	0.09	3.33		<0.001	[0.12, 0.48]	-
Thatta	0.42	0.11	3.82		<0.001	[0.20, 0.64]	-
Gwadar	0.55	0.12	4.58		<0.001	[0.31, 0.79]	-
Badin	0.25	0.10	2.50		0.013	[0.06, 0.44]	-
Muzaffargarh	−0.05	0.13	−0.38		0.705	[−0.31, 0.21]	-
Model Summary
Statistics				Value
R-squared				0.78
Adjusted R-squared				0.77
F-statistic				125.45
p-value				<0.001
Sample Size (N)				350

Source: calculated by authors.

provides crucial insights into the factors influencing migration patterns in coastal communities of Pakistan. This study underscores the significance of climate change variables—sea level rise, temperature increases, and flood frequency—alongside economic development, in shaping migration behaviors. The details of the results are given below:

The coefficient for sea level rise signifies a robust positive relationship with migration patterns. This implies that for every unit increase in sea level, migration patterns increase by 0.85 units. The small standard error (0.10) and high t-value (8.50) indicate that this estimate is precise and reliable. The statistical significance (p < 0.001) confirms the strong impact of sea level rise on migration. The findings of the fixed effects show that coastal regions like Karachi and Gwadar, which face rising sea levels, are at a heightened risk of displacement. The low VIF (1.20) suggests minimal multicollinearity, ensuring that this variable’s effect is not inflated by correlations with other predictors. This finding aligns with global trends where rising sea levels lead to the displacement of coastal populations. In Pakistan, many coastal communities are already vulnerable due to limited infrastructure and resources. As sea levels continue to rise, these challenges are exacerbated, leading to increased displacement. This underscores the need for proactive measures, such as enhancing coastal defenses and implementing sustainable land use planning to mitigate risks and manage migration effectively.

Temperature increases show a significant positive impact on migration patterns. For each unit increase in temperature, migration patterns increase by 0.75 units. The low standard error (0.12) and high t-value (6.25) indicate the reliability of this estimate. The statistical significance (p < 0.001) confirms this positive relationship, and the low VIF (1.30) indicates low multicollinearity. Rising temperatures can lead to adverse living conditions, especially in regions already experiencing high temperatures. With a p-value < 0.001, temperature increases significantly drive migration, particularly in regions where rising temperatures threaten livelihoods, such as in Thatta and Badin. Increased heat stress, reduced agricultural productivity, and higher energy demands are direct consequences. For Pakistan’s coastal communities, which are heavily dependent on agriculture and fishing, these changes threaten livelihoods and prompt migration as an adaptation strategy. This finding highlights the importance of investing in climate-resilient agricultural practices and infrastructure to mitigate adverse effects and reduce migration necessity.

The coefficient for flood frequency indicates a significant relationship with migration patterns. For each unit increase in flood frequency, migration patterns increase by 0.60 units. The moderate standard error (0.15) and high t-value (4.00) suggest that this estimate is reliable. The p-value (<0.001) confirms its statistical significance, and the low VIF (1.10) indicates minimal multicollinearity. Frequent flooding disrupts lives, destroys property, and damages infrastructure, leading to temporary and permanent displacement particularly in Badin and Thatta. In Pakistan, flooding is a recurring issue, and increased frequency and intensity due to climate change can severely impact vulnerable communities. This strong association highlights the need for improved flood management systems, including better forecasting, early warning systems, and resilient infrastructure, to protect communities and reduce forced migration.

The coefficient for economic development also has a positive relationship with migration patterns. For each unit increase in economic development, migration patterns increase by 0.50 units. The higher standard error (0.20) indicates more variability in this estimate. The t-value (2.50) and p-value (0.013) show that this relationship is statistically significant, albeit less robust than the climate variables. The low VIF (1.05) indicates no significant multicollinearity issues. While economic development generally improves living standards, it also facilitates migration by providing resources and networks. Economic development, particularly in Gwadar, facilitates migration by providing the resources necessary for relocation. However, it plays a dual role as both an enabler of migration and a mitigator of its necessity. This dual role highlights the need for balanced development strategies that enhance local resilience while providing safe and organized migration pathways.

The fixed effects analysis reveals that coastal regions such as Karachi (0.30), Thatta (0.42), Gwadar (0.55), and Badin (0.25) experience distinct migration dynamics, with significant positive coefficients for each. These coefficients suggest that migration patterns in these areas are particularly influenced by the local environmental and socio-economic conditions. The t-values for these regions—Karachi (3.33), Thatta (3.82), Gwadar (4.58), and Badin (2.50)—demonstrate strong statistical significance, with p-values of less than 0.001 for Karachi, Thatta, and Gwadar and 0.013 for Badin. These findings emphasize that coastal areas, facing direct climate impacts such as sea level rise and flooding, are more vulnerable to displacement. In contrast, the negative coefficient for Muzaffargarh (−0.05), which is not statistically significant (t-value = −0.38, p = 0.705), indicates that inland regions like Muzaffargarh are less affected by the same migration drivers, likely due to their distance from immediate coastal climate risks. This highlights the distinct challenges faced by coastal versus inland communities in the context of climate-induced migration.

The model’s R-squared (0.78) indicates that 78% of the variability in migration patterns can be explained by the independent variables. This high value demonstrates the strong explanatory power of the model. The adjusted R-squared (0.77) slightly adjusts for the number of predictors, confirming the model’s robustness and fit. The F-statistic (125.45, p < 0.001) further validates the model, indicating that the independent variables collectively provide a meaningful explanation for migration patterns. With a substantial sample size (N = 350), the findings are based on a robust dataset, enhancing their reliability and generalizability.

5.2. Logistic Regression

The logistic regression analysis presented in

Table 5. Logistic regression of migration patterns (DV) and climate change (IV).

Variable	β (Coefficient)	SE (Standard Error)	Wald Statistic	p-Value	Exp(β) (Odds Ratio)
Intercept	−2.125	0.550	14.2	< 0.001	0.119
Sea Level Rise	0.45	0.12	14.25	<0.001	1.57
Temperature Increase	0.38	0.10	15.36	<0.001	1.46
Flooding Intensity	0.52	0.15	18.52	<0.001	1.68
Saline Intrusion	0.33	0.11	10.67	0.001	1.39
Socio-Demographic Variables	β (Coefficient)	SE (Standard Error)	Wald Statistic	p-value	Exp(β) (Odds Ratio)
Age (35–45)	0.25	0.09	7.56	0.006	1.28
Gender (Female)	0.29	0.12	5.88	0.015	1.34
Education Level	0.18	0.07	6.01	0.014	1.20
Socio-Economic Status (Low Income)	0.48	0.13	12.85	<0.001	1.62
Model Fit Statistics				Value
−2 Log Likelihood				550.35
Chi-Square				120.45
Pseudo R2 (Nagelkerke)				0.31
p-Value				<0.001

Source: calculated by authors.

provides crucial insights into the impact of climate change variables and socio-demographic factors on migration patterns in coastal communities of Pakistan. The results are given below.

The coefficient for sea level rise (0.45) indicates a significant positive relationship with migration patterns. This means that as sea levels rise, the likelihood of migration increases. The odds ratio (1.57) suggests that for every unit increase in sea level rise, the odds of migration increase by 57%. The low standard error (0.12) and high Wald statistic (14.25) reinforce the reliability and significance (p < 0.001) of this predictor. This finding is consistent with global trends where rising sea levels force coastal populations to relocate. In Pakistan, rising sea levels threaten densely populated coastal areas, making migration a necessary adaptation strategy. Policies should focus on enhancing coastal defenses and developing sustainable land use plans to mitigate the risks associated with sea level rise.

Temperature increase has a coefficient of 0.38, indicating a significant positive impact on migration. The odds ratio (1.46) implies that a one-unit increase in temperature raises the odds of migration by 46%. The low standard error (0.10) and high Wald statistic (15.36) suggest a reliable and significant relationship (p < 0.001). Higher temperatures can lead to adverse living conditions, particularly in regions already experiencing high heat. In Pakistan’s coastal communities, rising temperatures can reduce agricultural productivity and increase heat stress, driving people to migrate. This underscores the need for climate-resilient agricultural practices and infrastructure to reduce the adverse effects of temperature increases.

The coefficient for flooding intensity (0.52) signifies a strong positive relationship with migration. The odds ratio (1.68) indicates that each unit increase in flooding intensity increases the odds of migration by 68%. The standard error (0.15) and high Wald statistic (18.52) confirm the reliability and significance (p < 0.001) of this predictor. Frequent and intense flooding disrupts lives, destroys property, and damages infrastructure, leading to displacement. Improved flood management systems, including better forecasting and resilient infrastructure, are crucial to protect vulnerable communities and reduce forced migration.

Saline intrusion’s coefficient (0.33) indicates a positive impact on migration, with an odds ratio of 1.39. This suggests that a one-unit increase in saline intrusion raises the odds of migration by 39%. The standard error (0.11) and Wald statistic (10.67) confirm the reliability and significance (p = 0.001). Saline intrusion affects water quality and agricultural productivity, pushing communities to migrate. Addressing saline intrusion through improved water management and resilient agricultural practices can help mitigate its impact and reduce migration.

The coefficient for age (35–45) (0.25) indicates that individuals in this age group are more likely to migrate. The odds ratio (1.28) implies a 28% increase in the odds of migration for this age group. The standard error (0.09) and Wald statistic (7.56) suggest a reliable and significant relationship (p = 0.006). Individuals in this age group are typically more involved in economic activities and family responsibilities, making them more likely to migrate for better opportunities. Targeted support and resources for this demographic can help manage migration effectively.

The coefficient for gender (Female) (0.29) indicates that females are more likely to migrate, with an odds ratio of 1.34. This suggests a 34% increase in the odds of migration for females. The standard error (0.12) and Wald statistic (5.88) indicate reliability and significance (p = 0.015). This finding highlights the importance of gender-sensitive policies and programs that address the specific needs and challenges faced by women in migration contexts.

The coefficient for education level (0.18) indicates a positive impact on migration, with an odds ratio of 1.20. This implies that higher education levels increase the odds of migration by 20%. The standard error (0.07) and Wald statistic (6.01) suggest a reliable and significant relationship (p = 0.014). Higher education provides individuals with better opportunities and resources for migration. Enhancing educational opportunities and support can help manage migration flows and improve outcomes for migrants.

The coefficient for socio-economic status (low income) (0.48) indicates a significant positive impact on migration, with an odds ratio of 1.62. This suggests that low-income individuals are 62% more likely to migrate. The standard error (0.13) and Wald statistic (12.85) confirm the reliability and significance (p < 0.001) of this predictor. Low-income individuals often migrate in search of better economic opportunities. This finding highlights the need for economic support and livelihood opportunities to reduce forced migration among vulnerable populations.

The −2 Log Likelihood value (550.35) indicates the goodness of fit for the logistic regression model, with lower values suggesting a better fit. The Chi-square statistic (120.45, p < 0.001) shows that the model significantly improves the fit compared to a null model with no predictors. The Pseudo R² value (0.31) indicates that the model explains 31% of the variance in migration patterns.

5.3. Structural Equation Modeling and Path Coefficients

Table 6. Structural equation modeling (SEM) and path coefficients.

Path	Coefficient	Standard Error	t-Value	p-Value	Effect Type
Climate Change → Migration Patterns	0.62	0.08	7.75	<0.001	Direct
Climate Change → Socio-Economic Impact → Migration Patterns	0.35	0.09	3.89	<0.001	Indirect
Climate Change → Climate Awareness → Migration Patterns	0.40	0.07	5.71	<0.001	Indirect
Socio-Economic Impact → Migration Patterns	0.50	0.10	5.00	<0.001	Direct
Climate Awareness → Migration Patterns	0.30	0.11	2.73	0.007	Direct
Climate Change → Socio-Economic Impact	0.55	0.06	9.17	<0.001	Direct
Climate Awareness → Socio-Economic Impact	0.25	0.08	3.13	0.002	Direct
Socio-Economic Impact → Climate Awareness	0.20	0.09	2.22	0.027	Moderating

Source: calculated by authors.

presents the results of the structural equation modeling (SEM) analysis, illustrating the complex relationships between climate change, socio-economic impact, climate awareness, and migration patterns in coastal communities of Pakistan. The SEM analysis helps us understand both direct and indirect effects, highlighting mediating and moderating variables.

Direct effects: climate change → migration patterns: The path coefficient of 0.62 indicates a strong direct effect of climate change on migration patterns. This implies that an increase in climate-change-related factors significantly increases migration. The t-value (7.75) and p-value (<0.001) confirm the significance of this relationship. This result aligns with the observed trends where climate change impacts, such as sea level rise, temperature increases, and flooding, directly force coastal populations to migrate. It emphasizes the urgent need for climate adaptation strategies to mitigate these direct effects.

Direct effects: socio-economic impact → migration patterns: A path coefficient of 0.50 indicates a significant direct effect of socio-economic impacts on migration patterns. The t-value (5.00) and p-value (<0.001) confirm this relationship’s significance. Socio-economic factors, such as income levels, employment opportunities, and living conditions, play a crucial role in migration decisions. Improving socio-economic conditions can potentially reduce the need for migration.

Direct effects: climate awareness → migration patterns: The path coefficient of 0.30 suggests a moderate direct effect of climate awareness on migration patterns. The t-value (2.73) and p-value (0.007) indicate that this effect is significant. Higher climate awareness among communities can influence their migration decisions, potentially by enabling them to better understand and respond to climate risks. Enhancing climate education and awareness programs can therefore play a role in managing migration patterns.

Direct effects: climate change → socio-economic impact: A path coefficient of 0.55 indicates a strong direct effect of climate change on socio-economic impacts. The t-value (9.17) and p-value (<0.001) confirm this relationship’s significance. Climate change can exacerbate socio-economic challenges by affecting livelihoods, economic stability, and infrastructure. Addressing these socio-economic impacts is critical for mitigating migration pressures.

Direct effects: climate awareness → socio-economic impact: The path coefficient of 0.25 indicates a moderate direct effect of climate awareness on socio-economic impacts. The t-value (3.13) and p-value (0.002) confirm the significance of this relationship. Climate awareness can influence how communities perceive and respond to socio-economic challenges posed by climate change. Enhancing awareness can help communities adopt more resilient socio-economic practices.

Indirect effects: climate change → socio-economic impact → migration patterns: The indirect path coefficient of 0.35 indicates that socio-economic impacts partially mediate the relationship between climate change and migration patterns. The t-value (3.89) and p-value (<0.001) confirm this mediation effect’s significance. This suggests that climate change influences migration patterns not only directly but also indirectly by affecting socio-economic conditions. Interventions aimed at improving socio-economic resilience can mitigate the indirect effects of climate change on migration.

Indirect effects: climate change → climate awareness → migration patterns: The indirect path coefficient of 0.40 suggests that climate awareness partially mediates the relationship between climate change and migration patterns. The t-value (5.71) and p-value (<0.001) confirm the significance of this mediation effect. Increasing climate awareness can help communities better understand and prepare for climate-related risks, potentially reducing the necessity for migration. This highlights the importance of climate education and information dissemination.

Moderating effects: socio-economic impact → climate awareness: The path coefficient of 0.20 indicates that socio-economic impacts moderate the relationship between climate awareness and migration patterns. The t-value (2.22) and p-value (0.027) confirm this moderation effect’s significance. This suggests that the effect of climate awareness on migration patterns is influenced by socio-economic conditions. Enhancing socio-economic resilience can therefore strengthen the positive impact of climate awareness on reducing migration pressures.

The diagram in Figure 2 illustrates the relationships between climate change, socio-economic impact, climate awareness, and migration patterns, including direct and indirect effects. The edge styles indicate the nature of the effects (solid for direct, dashed for indirect, and dotted for moderating effects), while the labels show the path coefficients along with their significance levels.

The diagram presents a Structural Equation Modeling (SEM) framework that illustrates the relationships between Climate Change, Climate Awareness, Socio-Economic Impact, and Migration Pattern, along with the associated path coefficients.

Each arrow between the variables represents a hypothesized causal relationship, with coefficients indicating the strength and significance of these relationships. For example, Climate Change has a strong, significant positive effect on Climate Awareness (coefficient of 0.40, marked with ***) and on Socio-Economic Impact (coefficient of 0.35, also marked with ***), indicating that as climate change intensifies, it raises awareness and adversely affects socio-economic conditions.

Socio-Economic Impact significantly influences Migration Pattern with a high path coefficient of 0.50 (***), suggesting that deteriorating socio-economic conditions due to climate change are a strong driver of migration. Climate Awareness also positively influences Migration Pattern with a coefficient of 0.30 (**), indicating that heightened awareness of climate risks contributes to migration decisions, though to a lesser extent than socio-economic impacts.

The coefficient of 0.20 (**) from Climate Change to Migration Pattern suggests a direct but relatively weaker influence on migration decisions, compared to the indirect effects mediated through socio-economic impacts and climate awareness. This framework highlights the primary paths through which climate change impacts migration: directly, and indirectly by altering socio-economic conditions and increasing climate awareness.

6. Discussion and Conclusions

This study’s findings offer comprehensive insights into the factors influencing migration patterns in coastal communities of Pakistan, integrating multiple linear regression, logistic regression, and structural equation modeling (SEM). Each method contributes uniquely to understanding how climate change impacts migration, with empirical evidence providing context and validation for the results.

The multiple linear regression analysis highlights the critical role of climate change variables—sea level rise, temperature increases, and flood frequency—in shaping migration patterns. Sea level rise exhibits a robust positive relationship with migration, indicating that higher sea levels significantly drive migration patterns. This finding is consistent with global observations where rising sea levels lead to coastal displacement [33,34,35]. Similarly, temperature increases significantly affect migration, aligning with research showing that higher temperatures adversely impact living conditions and agricultural productivity, prompting migration [36]. Flood frequency also demonstrates a significant relationship with migration, corroborating studies that link increased flooding with displacement [37]. Economic development’s positive association with migration suggests that economic improvements may facilitate migration by providing resources and networks, echoing findings that economic factors influence migration decisions [38,39].

The logistic regression analysis further confirms the impact of climate change on migration. Sea level rise and temperature increases significantly increase the odds of migration, supporting the findings of previous research linking these factors to relocation [40,41]. Flooding intensity and saline intrusion also significantly affect migration, aligning with studies that report increased migration due to frequent flooding and deteriorating water quality [42,43,44]. Socio-demographic factors, such as age, gender, education level, and socio-economic status, also significantly influence migration, highlighting the complex interplay between individual characteristics and migration patterns [45,46].

The SEM analysis reveals intricate relationships between climate change, socio-economic impacts, climate awareness, and migration patterns. The direct effects of climate change and socio-economic impacts on migration underscore the direct influence of environmental and economic factors on migration decisions. The moderate direct effect of climate awareness on migration highlights the role of climate education in shaping migration behaviors. The indirect effects, where socio-economic impacts and climate awareness mediate the relationship between climate change and migration, suggest that climate change influences migration not only directly but also through its impact on socio-economic conditions and awareness. This is consistent with research demonstrating that socio-economic and awareness factors mediate the effects of environmental changes on migration [47,48,49].

In conclusion, the multiple linear regression analysis shows a strong positive relationship between climate change variables—specifically sea-level rise, temperature increases, and flood frequency—and migration rates. These environmental factors significantly drive displacement, while economic development plays a lesser role. The model indicates that these variables collectively explain a substantial portion of migration variability, underscoring the need for targeted interventions addressing environmental changes.

Supporting these findings, the logistic regression analysis reveals that sea-level rise, temperature increases, flooding intensity, and saline intrusion significantly increase the likelihood of migration. Socio-demographic factors, such as age, gender, education level, and socio-economic status, also influence the likelihood of migration, emphasizing the need for policies that address both environmental and socio-demographic factors.

The SEM analysis demonstrates that climate change affects migration patterns both directly, by worsening environmental conditions, and indirectly, through socio-economic impacts. Climate awareness is identified as a partial mediator in this relationship, suggesting that enhancing community understanding of climate risks can mitigate some migration pressures.

6.1. Policy Implications

The findings of this study highlight the urgent need for integrated and multifaceted policy approaches to address the challenges posed by climate change and its impact on migration in coastal communities in Pakistan. Policies should prioritize enhancing climate resilience by investing in infrastructure improvements to mitigate the effects of sea level rise, flooding, and temperature increases. There should be a focus on developing early warning systems and adaptive measures to reduce the vulnerability of coastal communities to extreme weather events and environmental degradation. Additionally, socio-economic support programs are essential to help affected populations adapt to changing conditions and reduce the pressures leading to migration. This includes improving education and economic opportunities to build community resilience and capacity for adaptation. Enhancing climate awareness and education can also empower communities to better prepare for and respond to climate risks. Finally, comprehensive migration management strategies must be developed, integrating environmental, socio-economic, and demographic factors to effectively support displaced populations and reduce the impacts of climate-induced migration.

6.2. Limitations and Future Directions

This study’s reliance on cross-sectional data restricts insights to only a single point in time, along with smaller representation of sample size, thus limiting the validity and reliability while mitigating the effects of climate change. Thus, a longitudinal study design with corroboration of larger sample size (from all provinces) were the order of the day through a multistage sampling technique. In addition, a holistic team encompassing a research coordinator and investigators will be taken into account while limiting the biasness in the data collection process. To mitigate this, future studies should employ mixed methods, including qualitative interviews and focus groups, to validate self-reported data and enrich the overall analysis. Moreover, the focus on coastal communities in Pakistan may limit the applicability of the findings to other regions with different environmental and socio-economic contexts. Expanding research to include a diverse range of geographic areas will enhance the generalizability of the results and offer a broader perspective on climate-induced migration. This study acknowledges direct and indirect effects but does not fully explore the complex relationship between various factors influencing migration. Future research should incorporate additional variables, e.g., climate change impacts on institutional and subjective quality of life, climate-induced brain drain, climate-induced militancy and terrorism, climate-induced hate crimes, climate-induced gender diversity, climate induced food and water scarcity, agricultural productivity, and infant mortality.