My first question here was - how did they estimate the size of the shadow economy?

In the underlying article (https://www.bcg.com/publications/2023/unveiling-the-shadow-economy), this is how they explain it:

Unlike conventional transactions that are documented and accounted for, the shadow economy comprises a range of informal, unreported, and often illegal activities that evade traditional methods of measurement. Therefore, to gauge its impact, researchers and economists must look past official records and established business channels to other data sources. There are three main strategies for doing this, each with its own merits, drawbacks, and constraints.

This approach uses representative household micro-surveys to investigate public opinion of the shadow economy, along with actual participation. The findings are then extrapolated to a larger scale.

Pros. Surveys provide detailed insights into the shadow economy’s structure. They are particularly useful when other data is not available. Cons. Surveys draw information from only a fraction of the population. They are susceptible to design flaws and response bias—for instance, people tend to avoid reporting their own shadow activities. The surveys may thus underestimate the informal economy’s size. Language and social disparities may make it difficult to compare results across surveys.

The indirect approach deploys macroeconomic indicators and models, using various economic indicators such as discrepancies between national expenditure and income statistics, disparities between electricity consumption and reported economic activity, and the demand for currency in cash to estimate the size of informal economic activities.

Pros. This approach typically leverages readily available data, making it relatively cost-effective. Cons. Each analysis measures a single aspect of the shadow economy instead of encompassing its complexity.

This approach uses a model called statistical multiple indicators and multiple causes (MIMIC). By analyzing the rise and fall of key variables over time, MIMIC can explore the evolving relationship between the shadow economy and its enablers. Variables include tax burdens, self-employment levels, and unemployment data. This model yields relative estimates of shadow economy activity, often applied to groups of countries.

Pros. MIMIC offers a more sophisticated analysis than other methods by considering multiple causes and indicators concurrently. Cons. The model generates relative estimated coefficients that require additional calibration. It cannot be exclusively applied to individual countries.

Given the complexity of measurement, only a limited number of studies have been published that compare the size of shadow economy across multiple countries using consistent methodologies. One ongoing database, released by the World Bank in 2018, covers shadow economy sizes and trends for 158 countries from 1991 through 2015. Remarkably, it shows a distinctive decline in the global average share of shadow activities by 6.7 percentage points—from 34.5% of total nominal GDP to 27.8%—over a span of 24 years. This decline can be attributed to factors such as economic growth and development, improved government regulation and enforcement, increased technology adoption and digitalization, and enhanced access to formal financial services. Nonetheless, there is still a long way to go for some countries that need to improve their government effectiveness in constraining shadow economy. Moreover, new technologies such as blockchain-based digital currencies and generative AI may also have profound impact on the shadow economy.