Case study · 04

Taiwan Resistance Analysis

Analyzing who would resist a Chinese invasion — and why.

Researcher and Analyst · UC San Diego · 20-year longitudinal dataset

~2.5M

Fighting-age Taiwanese willing to resist

20 yrs

Of TNSS data analyzed

Predictor: military confidence

Rising

Resistance trend across the period

Context

The Problem

Taiwan contingency planning rests on a frequently cited assumption: Taiwan's population will resist. This assumption shapes assessments of conflict duration, required force levels, and the plausibility of various PRC strategies.

But relatively little work had been done to quantify how much resistance might exist, how stable that sentiment is, or what factors actually drive it. The assumption was treated as given rather than analyzed.

My question: Can 20 years of public opinion data provide operationally useful insight into this assumption?

Method

Analytical Approach

Focus on behavior, not opinion

Rather than measuring attitudes toward China or independence, I focused on what respondents said they would actually do during an invasion — fighting, supporting resistance, remaining neutral, fleeing, or surrendering. This moves the analysis closer to operationally relevant outcomes.

Treat resistance as a spectrum

Dividing respondents into "pro‑China vs. anti‑China" misses operational nuance. The difference between active fighters and passive supporters matters enormously for planning purposes.

Find predictors, not headlines

A single aggregate number is less valuable than understanding which variables drive resistance. Identifying the levers matters more than measuring the current setting.

Findings

Key Findings

Resistance is stronger — and growing — than most analyses assume

The data suggests the assumption is directionally correct, but the magnitude is underappreciated. Willingness to resist has increased meaningfully over the 20‑year period, tracking alongside Taiwan's democratic consolidation and sharpening cross‑strait tensions.

~2.5 million fighting-age Taiwanese could participate directly

Based on behavioral survey responses mapped against demographic data. This is a planning‑relevant number — its value is in anchoring scenario analysis, not predicting outcomes.

Military confidence is the strongest predictor

Respondents who believed Taiwan could defend itself were significantly more likely to indicate they would resist. This has direct policy implications: efforts to improve Taiwan's military credibility may have a multiplier effect on population resilience beyond their direct military value.

Predictor strength · ranked

01Confidence in Taiwan's military
02Taiwanese identity (vs. Chinese)
03Political affiliation
04Expectation of U.S. support
05Views on economic integration with China

Reflection

What I Learned

Good intelligence work is rarely about predicting a single outcome. It's about identifying the variables that matter most, understanding the uncertainty around them, and improving decision‑making under incomplete information. Planners shouldn't treat "Taiwan will resist" as a binary assumption — the conditions that make resistance likely are themselves conditions that policy can influence.

Back to home Talk about this work →