From Sentiment to System Metrics: Measuring Societal Readiness with NLP

Modern infrastructure projects—especially large energy systems—often succeed or fail not only because of technology or economics, but because of societal acceptance.

Public opinion influences:

• Regulation - Investment decisions - Policy timelines - Labor markets - Infrastructure deployment

Yet most economic models treat societal acceptance as qualitative or subjective.

In our research, we attempted to address this gap by developing a measurable variable called Societal Readiness Level (SRL).

SRL transforms large-scale public discourse into a quantitative indicator that can be used in econometric models.

This article explains how SRL can be constructed using Natural Language Processing (NLP).

Why Measuring Societal Readiness Matters

Consider large-scale technologies such as:

• Nuclear energy - Carbon capture - Hydrogen infrastructure - Artificial intelligence - Genetic engineering

In many cases, these technologies are technically feasible, yet deployment stalls due to:

• Public resistance - Regulatory uncertainty - Political pressure

Traditional engineering frameworks already measure Technology Readiness Level (TRL).

However, TRL measures technological maturity, not whether society is ready to accept the technology.

This motivates the concept of Societal Readiness Level (SRL).

Conceptual Framework

SRL attempts to measure how prepared society is to support a technology’s deployment.

We define SRL as:

A quantitative measure of how socially, politically, and institutionally prepared society is to support large-scale technology deployment.

To estimate SRL, we analyze large-scale public discourse using NLP.

The high-level pipeline looks like this.

SRL Construction Pipeline

This pipeline converts unstructured text into a structured time-series variable.

Step 1: Collecting Public Discourse

The first step is gathering large-scale textual data.

Examples include:

• National news articles - Policy debates - Institutional reports - Public commentary

These sources provide a continuous record of how society discusses a technology.

For example, after the Fukushima nuclear accident in 2011, media coverage of nuclear energy changed dramatically. Capturing these shifts in discourse is key to measuring societal readiness.

Step 2: NLP Sentiment Extraction

The next step is extracting sentiment from text. Each document is processed using NLP techniques such as:

• Tokenization - Stopword removal - Sentiment scoring - Contextual classification

Each document receives a sentiment score:

• Where: - `` indicates strongly negative sentiment - `` indicates neutral discourse - `` indicates strongly positive sentiment

These scores reflect the tone of public discussion about the technology.

Step 3: Temporal Aggregation

Individual sentiment scores must be aggregated into a time series.

For a given time period ``:

• Where: - `` = number of documents in period `` - `` = sentiment score for document ``

This produces a sentiment trajectory over time.

Step 4: Transforming Sentiment into SRL

Raw sentiment is not the same as societal readiness. SRL introduces normalization and threshold mapping.

Conceptually:

• Where: - `` is aggregated public sentiment - `` transforms sentiment into readiness levels

SRL Evolution Over Time

The sharp drop represents the shock in public discourse after Fukushima.

SRL converts this sentiment signal into readiness levels such as:

Step 5: Using SRL in Econometric Models

Once SRL is constructed, it becomes a quantitative explanatory variable.

Example regression specification:

• Where: - `` represents an outcome variable (e.g., market volatility or employment) - `` captures societal readiness

In our research, SRL was incorporated into panel regressions with time fixed effects to analyze how societal readiness influences energy market behavior, capital allocation, and workforce dynamics.