This project examines how major U.S. tariff episodes aligned with sector-level employment shifts using BLS, FRED, and USITC data.
I built an end-to-end labor-market analysis workflow covering 2015–2025, with forecasts through 2030, to test whether tariff-related protection supported durable job growth or instead contributed to more uneven sector outcomes beneath strong aggregate payroll numbers.
Headline Insight: Tariff-related protection appears to support select upstream industries, but broader labor-market effects remain mixed, unevenly distributed, and often delayed rather than immediate.
This project was built around a practical question:
Do tariff-related policy shifts create durable labor-market strength, or do they redistribute job growth in ways that make the economy look stronger in aggregate than it is at the sector level?
That matters for:
- Policy Analysis
- Labor-Market Interpretation
- Supply Chain Strategy
- Business and Investment Decision-Making
- Manufacturing remained structurally weak despite protection-focused policy support
- Construction appeared to benefit more than most sectors from reshoring and domestic investment dynamics
- Retail Trade showed weak net job growth despite broader recovery periods
- Labor-Market Effects Often Appeared With a Lag, rather than immediately after major tariff episodes
- Forecasts Validated Well Against Realized BLS Data, with forecast error below 4% in 4 of 5 sectors
Headline job growth can make the labor market look stronger than it really is.
This analysis suggests that tariff-related protection may support select upstream industries, but broader labor-market effects are more mixed. Employment gains appear unevenly distributed, and the impact on jobs often shows up with a lag rather than immediately after policy changes.
That means aggregate job numbers alone may not fully reflect what is happening underneath the surface.
Bottom Line: The labor market may appear stable in aggregate while sector-level conditions tell a more fragile and uneven story.
2015–2025
2026–2030
- Manufacturing
- Construction
- Retail Trade
- Transportation & Warehousing
- Total Nonfarm Payrolls
- 2018–2019 U.S.–China Trade War
- 2020–2021 COVID and Supply-Chain Disruption Period
- 2025 Tariff Escalation as a Forward-Looking Scenario Context
I built a multi-step workflow using public economic data:
- Collected labor-market and macroeconomic data from BLS, FRED, and USITC
- Cleaned and standardized time-series data in Python
- Stored structured datasets in PostgreSQL
- Used SQL for trend, sector, and lag analysis
- Generated five-year forecasts in Prophet
- Validated forecast outputs against realized BLS data where available
- Designed a Tableau dashboard to make findings easier to interpret
Forecast performance was checked against realized BLS data.
| Sector | Forecast | BLS Actual | Error |
|---|---|---|---|
| Construction | 8,286K | 8,309K | 0.3% |
| Manufacturing | 13,194K | 12,573K | 5.0%* |
| Retail Trade | 15,459K | 15,427K | 0.2% |
| Total Nonfarm | 157,616K | 158,466K | 0.5% |
| Transportation & Warehousing | 6,294K | 6,532K | 3.6% |
Validation Result: Forecast error remained below 4% in 4 of 5 sectors, supporting the model’s usefulness for directional sector-level forecasting.
*Manufacturing deviation was influenced by benchmark revision effects.
The labor market is not simply “strong” or “weak.”
What this project suggests is that job growth may be becoming more uneven, concentrated, and policy-sensitive.
In practical terms:
- Some sectors benefit
- Some sectors stagnate
- Some sectors absorb delayed downstream pressure
- Headline job growth can mask divergence underneath
The dashboard is designed to show:
- Sector-level employment trends
- Forecast scenarios
- Policy-event context
- Labor-market comparison across sectors
Live Dashboard: View on Tableau Public
- Python
- PostgreSQL
- SQL
- Prophet
- Tableau
- BLS / FRED / USITC APIs
This analysis does not claim that tariffs alone caused all labor-market changes.
Important limitations include:
- Tariff episodes overlapped with other macroeconomic shocks, especially COVID and broader business-cycle effects
- Sector-level employment data supports timing and association, not strict causal attribution
- Forecasts assume a degree of structural continuity and may weaken under future shocks or major policy changes