SabahKu · District brief · Tawau Division
Tawau
AI-generated · unreviewed draftData release 2026.09.26 · 50/50 citations verified · azure:gpt-5-mini
Situation
Tawau is classified as a "Services-led towns & suburbs" district with large population and labour force relative to Sabah's districts .
Median household income was RM 4,852 in 2024 and has grown slowly since 2019 (median income growth 1.9% over 2019–2024) .
Real GDP published to 2020 was RM 6,535 (2015 prices), and model nowcasts project GDP rising to a 2026 central estimate of RM 7,958 (80% interval RM 7,384–RM 9,663) for the baseline scenario .
Labour-market participation is high while unemployment has risen; labour force participation was 73.3% and unemployment 7.1% in 2024 .
Strengths
- Labour force participation is above structural peers and improving, a rule-based strength for Tawau in 2024 .
- The driver model associates higher labour force participation positively with median household income in Tawau's structure (labour force participation top associated factor) .
- The district has shown improving absolute poverty outcomes since 2019, with the absolute poverty rate falling from 14.2% in 2019 to 13.2% in 2024, recorded as a mixed verdict in the scorecard (improving trend) .
Constraints
- Median household income is below the peer median and flagged as a concern in the scorecard for 2024 .
- Unemployment rate is higher than peers and marked as a concern in the scorecard for 2024 .
- Income is lower than the modelled structural expectation (actual median RM 4,852 vs expected RM 5,550), with unemployment and piped-water access among top associated negative factors in the drivers analysis .
- Income growth since 2019 is weak relative to peers and the scorecard flags income growth as not comparable but the trend places Tawau low on welfare momentum .
Options
- Targeted employment activation to convert high participation into formal employment (Evidence: moderate); precedent: the atlas positive-deviant peer Sabak Bernam improved welfare momentum, offering a structural peer learning case .
- Local programmes to reduce unemployment among working-age groups, combined with skills matching to services and manufacturing sectors (Evidence: moderate); supported by the drivers associating unemployment with higher poverty and lower income .
- Invest in rural and digital infrastructure to mobilise rural economic potential and remove access constraints (Evidence: limited); national plans recommend expanding infrastructure and digital access for regional balance but these are planning documents rather than evaluated interventions .
- Use peer exchange with services-led towns that improved incomes to pilot private–public collaborations for jobs and firm support (Evidence: limited); peers listed include Sandakan and Samarahan as comparable service-led examples .
Uncertainties
- Median income and GDP after 2020 are modelled: income projections for 2025–2027 have wide 80% intervals and GDP nowcasts/projections rely on modelled central estimates .
- Forecast backtest error for income_median is 2‑year absolute error 6.9%, adding uncertainty to near-term income projections .
- District statistics carry boundary-change flags (2019 values include Kalabakan), which affects comparability of trends across surveys .
- Some scorecard verdicts rely on peer medians which are structural matches, not causal proofs of policy effectiveness .
What data would change this view
- Official district-level GDP published beyond 2020 would replace nowcasts and narrow uncertainty about recent growth .
- A sub-district employment survey distinguishing formal vs informal jobs would clarify why high participation coexists with rising unemployment and would test driver associations between participation, unemployment and income .
- Reconciled series that apply consistent boundaries (separating Kalabakan where relevant) would change trend assessments for income and poverty .
- Impact evaluations of local job and infrastructure programmes would upgrade evidence for options from limited/moderate to strong .