Hesitant or determined? The influence of social and environmental factors on settlement decision-making of rural in-migrants: evidence from Dali, China

The paper addresses how rural in-migrants decide to settle, conceptualizing settlement as a dynamic process across stages—desire, expectation, and plan—rather than a single stated preference. Motivated by evidence that migrants’ motivations and constraints evolve over time, the study applies the Theory of Planned Behavior to measure generalized intentions and examines how social and environmental factors (economy, natural environment, public service facilities) shape settlement decisions in rural Dali, China. The purpose is to identify which factors matter at different decision stages and how effects vary by socio-demographics (age, gender, education), informing rural development strategies that enhance integration and vitality.

The authors distinguish rural in-migration from counter-urbanization, noting both benefits (well-being, lifestyle) and challenges (economic precarity, integration). They review measurement approaches: revealed preferences (actual moves) and stated preferences (intentions), highlighting that standard intention measures often miss process dynamics. Prior work links settlement intentions to socio-economic drivers (employment, income, development), policy contexts (notably China’s hukou system), and environmental amenities (public services, air quality), though findings on natural amenities are mixed. Socio-demographic factors (age, education, gender, marital status, household registration, housing tenure) significantly shape intentions; homeownership and local hukou presence increase settlement likelihood. The paper proposes TPB to capture graded intentions (desire, expectation, plan) and focuses on three external factor domains—economy, natural environment, public service facilities—that directly affect intentions.

Case selection: Four rural villages near Dali, Yunnan, China—Dali Ancient Town (commercially developed), Shuanglang (balanced across factors; used as reference group), Shaxi and Xizhou (strong natural landscapes and rural lifestyle). Data collection: Field surveys and questionnaires administered in May 2021 to in-migrants residing in the four villages. Of 372 distributed, 353 returned, 338 valid (95.75% effective response). Valid questionnaires by site: Dali Ancient Town 107, Shuanglang 80, Shaxi 78, Xizhou 73. Measures: Settlement decision-making operationalized via three TPB-aligned scenarios measured on 1–5 Likert scale: Scenario 1 (desire: “I hope to live here in the future, ignoring constraints”), Scenario 2 (expectation: “Considering my conditions, I want to live here”), Scenario 3 (plan: “I have planned to live here”). The overall settlement decision score is the mean of the three scenario items. Drivers: Social-environmental domains measured via multiple Likert items (1–5) across Economy (job accomplishment, work rhythm and income satisfaction, firm prospects, workplace relations, importance of income/opportunities, rural economic development, family financial pressure, competition, local policy support), Natural environment (satisfaction with environment/climate, importance of comfortable environment and local natural environment, frequency/time in public spaces), Public service facilities (satisfaction with medical, education, prices, public safety; importance of service convenience). Exploratory factor analysis (EFA) extracted first principal components as composite indices: Economy KMO 0.781 (p<0.001), Natural environment KMO 0.575 (p<0.001), Public service facilities KMO 0.602 (p<0.001). Key descriptive statistics: Overall settlement decision mean 3.81; Scenario 1 mean 3.96 (SE 0.061), Scenario 2 mean 3.82 (SE 0.062), Scenario 3 mean 3.66 (SE 0.068). Domain means: Economy 3.94 (SD 0.718), Natural environment 4.35 (SD 0.571), Public service facilities 2.81 (SD 0.775). Controls: Gender, age groups (Under 25; 25–49; 50+), marital status, household registration (rural/urban; local/non-local), education, presence of local family members, length of residence (<3 years vs ≥3), prior travel/sojourn in current village, pre-migration residence (first-tier cities/other cities/county and below), current dwelling type (rented vs purchased), and village fixed effects (reference: Shuanglang). Analysis: Multiple linear regressions. Model 1 uses overall settlement decision (mean of three scenarios). Models 2–4 use Scenario-specific scores (desire, expectation, plan). Subgroup regressions by gender (Models 5–6), education (Models 7–9), and age groups (Models 10–12). Software: IBM SPSS Statistics 25.

Overall effects (Model 1, Adjusted R2=0.311): All three domains significantly and positively predict settlement decision-making—Economy β=0.323***; Natural environment β=0.277***; Public service facilities β=0.254***. Stage-specific effects: • Scenario 1 (desire; Model 2, Adj. R2=0.271): Economy β=0.272***; Natural environment β=0.087*; Public services not significant (β=0.045). • Scenario 2 (expectation; Model 3, Adj. R2=0.254): Economy β=0.353***; Natural environment β=0.334***; Public services not significant (β=0.080). • Scenario 3 (plan; Model 4, Adj. R2=0.291): Economy β=0.269***; Natural environment β=0.178***; Public services significant (β=0.113**), indicating public services become salient at the planning stage. Notable controls (Model 4): Female shows lower planning tendency vs male (β=−0.080*). Age 25–49 more positive than under 25 (β=0.152**). Urban hukou associated with more negative overall and planning attitudes versus rural hukou (overall: β=−0.097*; plan: β=−0.148**). Prior travel/sojourn strongly positive across models (e.g., overall β=0.180***). Purchasing a dwelling positively associated with expectation (Model 3: rented β=−0.096*). Site effects: Shaxi higher overall vs Shuanglang (β=0.097*); Dali Ancient Town and Shaxi show stronger planning tendencies (Model 4: β=0.107* and β=0.100*). Gender subgroups (Table 7): Economy and natural environment positively associated for both genders; public services significant only for females (Female: Public services β=0.174**; Male: β=0.017 ns). Males 25–49 and 50+ higher than under 25; male urban hukou negative (β=−0.137*); owning (vs renting) positive for males (rented β=−0.163**). Females: local family members positive (β=0.208***); prior travel/sojourn positive (β=0.309***). Education subgroups (Table 8): • Junior high and below: Economy β=0.619***; natural environment/public services ns. • High school: Economy β=0.522***; Natural environment β=0.272**. • College and above: Economy β=0.229***; Natural environment β=0.276***; Public services β=0.113*. Additional controls: For high school group, urban hukou negative (β=−0.268**); in college+ group, females less positive than males (β=−0.114**); age 25–49 more positive (β=0.188**); prior travel/sojourn positive (β=0.234***). Age subgroups (Table 9): • Under 25: Economy β=0.462*** drives settlement; being unmarried negative (β=−0.329**); renting negative (β=−0.422***); prior travel/sojourn positive (β=0.250*). • Ages 25–49: Economy β=0.334***; Natural environment β=0.241***; Public services β=0.102***; female negative (β=−0.106*); urban hukou negative (β=−0.137**); prior travel/sojourn positive (β=0.154***); Shaxi positive (β=0.151**). • 50 and above: Natural environment β=0.534** is the sole significant driver. Descriptives: Settlement intention declines from desire (mean 3.96) to plan (3.66), indicating increasing caution at the planning stage. Public service facilities average satisfaction is lowest among domains (mean 2.81), suggesting an area for improvement.

Findings confirm that settlement decision-making is stage-dependent: economic prospects and natural amenities shape desire and expectation, while public service facilities become crucial when migrants move from intention to concrete planning. This aligns with TPB’s graded intentions and indicates that practical considerations (healthcare, education, safety, service convenience) become decisive in translating plans into action. The study underscores heterogeneity across groups: women weigh public service quality more, possibly due to caregiving roles; with age, decision criteria diversify but natural environment dominates among older migrants; higher education broadens the range of considerations (all three domains matter for college-educated). Prior familiarity with the destination (travel/sojourn) facilitates positive settlement decisions, echoing links between visitors and subsequent migrants. Policy relevance: rural development strategies should balance economic vitality and environmental protection with sustained upgrading of public service facilities to support actual settlement and integration, particularly for family- and career-stage migrants.

The study demonstrates that while rural economic conditions and the natural environment promote desire and expectation to settle, the quality of public service facilities becomes pivotal at the planning stage. Influences vary by gender, age, and education: females place greater emphasis on services; older migrants prioritize natural environments; higher education corresponds to a wider set of factors. Contributions include: (1) a process-oriented, TPB-informed assessment of settlement intentions across desire–expectation–plan stages; (2) spatial planning implications highlighting the need to improve rural public service facilities alongside economic and environmental strategies. Future research should adopt more comprehensive frameworks to examine additional dimensions and finer-grained aspects within the three domains to better capture the complexity of settlement decision-making.

The analysis focuses on three broad domains—economy, natural environment, and public service facilities—without deeply exploring sub-dimensions or other potential influences (e.g., social networks, cultural integration, policy specifics beyond hukou). The cross-sectional design and single-region case limit generalizability; more comprehensive and comparative frameworks could better capture complexity and dynamics.