Regional Economic Impacts of the Just Energy Transition: Lessons for Coal Regions

When coal regions lose their main industry, they often enter a deceptive economic trap. On paper, wealth per person appears to be growing. However, this is often a statistical mirage caused by the departure of the workforce itself. This "hollowing-out" process leaves behind high unemployment and eroding local services.

A new study by Imke Rhoden and Jae-Hyuck Lee analyzes two decades of European Union data. They aim to quantify this phenomenon. By examining NUTS 2 regions (territorial units used for statistical purposes, similar to a province) from 2000 to 2022, the authors reveal a persistent economic penalty. Simple market corrections cannot fix this. The research is timely for South Korea, where regions like Chungnam face a scheduled coal phase-out.

The failure of reactive policy

The transition to renewable energy is a global necessity. However, its implementation is intensely local. Most decarbonization targets are set at the national level. Yet, the economic disruption occurs in specific, highly concentrated clusters. These coal regions suffer from "spatial lock-ins" (where capital and labor are tied to a specific location). This prevents workers from easily moving to new sectors.

Current policy approaches often struggle. They treat economic shifts as temporary shocks rather than permanent structural changes. Standard instruments often rely on passive support. Examples include unemployment benefits or early retirement schemes. These stabilize society in the short term. However, the authors argue they do not rebuild the underlying economic engine. Without active investment, these regions risk a downward spiral. As young workers leave, the local tax base erodes and public services decline.

Quantifying the hollowing-out effect

To move beyond anecdotal evidence, the authors employ a two-way fixed effects (FE) model. This statistical architecture compares coal and non-coal regions within the same country. This helps filter out broad national economic trends. It isolates the specific penalty associated with coal specialization. They also implement Conley standard errors (a method to correct for spatial dependence). This ensures that the proximity of one region to another does not bias the results.

The study identifies a specific mechanism called "hollowing-out." In these regions, GDP per capita actually grows faster than in non-coal regions. Specifically, it grows by 0.23 percentage points annually according to Model M3. While this sounds positive, it is a mathematical byproduct of population decline. As workers leave, the denominator in the "GDP per capita" equation shrinks. This creates an illusion of prosperity while labor conditions worsen. This divergence is visualized in .

Persistent labor market penalties

The empirical results suggest the damage is most visible in the labor market. The authors report a persistent within-country unemployment premium. This ranges from 0.87 to 1.06 percentage points [Table 3, Table 4]. This means workers in coal-heavy areas face significantly higher risks of joblessness. This holds true even when controlling for population density and national economic conditions.

The data shows this is not a fleeting issue. As seen in, coal regions consistently remain below their national average in GDP per capita.

This trend lasted throughout the entire 2000–2022 period. Similarly, shows that coal regions maintain a steady unemployment premium.

The authors emphasize that this penalty persists regardless of regional urbanization or initial income levels.

Limits of the empirical lens

Several limitations remain in the study. First, the "coal region" status is largely time-invariant (it does not change during the study). Therefore, the authors cannot use region-specific fixed effects. The findings represent a conditional association rather than definitive causal proof.

Second, the use of NUTS 2 data aggregates smaller municipalities. This granularity gap might mask more intense crises at the local level. For example, a specific town hosting a plant might suffer more than the larger region suggests. Finally, the exclusion of United Kingdom data after 2020 due to Brexit creates a composition shift. The authors acknowledge this as a limitation of the longitudinal analysis.

Verdict: Proactive or bust

The evidence suggests that transition success depends on timing. The authors conclude that support must be proactive. It should be initiated well before the first plant closes. Reactive policies arrive only after mass layoffs occur. These tend to be more expensive and less effective at preventing decay.

For policymakers in regions like Chungnam, the verdict is clear. Funding alone is insufficient. Success requires investing in "institutional capacity" (the administrative ability to manage complex programs). Local governments must be able to manage retraining and attract private investment. If the goal is to avoid a permanent hollowing-out, the window for building this infrastructure is closing.