**3. The relationship between economic and financial crime and economic development**

#### **3.1 Data and methodology**

Our sample covers the European Union 27 member states (EU-27) at present. The *Gross Domestic Product per capita (GDP)* is used by this paper as an economic development proxy, following various research works [9, 33]. These prosperity levels, corresponding to per capita GDPs of EU-27 countries are provided by World Bank Group [34] for the 2005–2019 time period. According to the latest classification of countries and lending groups provided by World Bank Group [34], all EU-27 countries are classified as developed countries (high-income countries), the latest added to this privileged category being Romania and Bulgaria. Further on, *the Human Development Index* (*HDI*) comprises three key dimensions of human development: a long and healthy life, knowledge and having a decent standard of living, aggregated within a composite index through their geometric mean, according to UNDP [35]. It has been previously used as a proxy for sustainable development by Murshed and Mredula [36]. The most recent data on the HDI comprised by our study cover the 2005–2018 time interval.

Our financial and economic crime proxies include *Corruption*, *Shadow Economy*, *Money Laundering* and *Cybercrime*. The perceived levels of *Corruption* in the public sector are taken from the latest report of the Corruption Perceptions Index (CPI) provided by Transparency International [19]. Our study particularly deals with countries' rankings, generally ranged from 1 (lowest level of corruption) to 180 (highest level of corruption), selecting the EU-27 member states only, for the 2005–2019 time period. Furthermore, *Shadow Economy* is considered from the data provided by Medina and Schneider [20] for the 2005–2017 time period, throughout which it is calculated as a percentage of the official GDP. *Money Laundering* statistics cover the 2012–2020 time period. *Cybercrime* data reflect its 2018 values, extrapolated to the entire time frame.

The summary statistics for our independent and dependent variables are presented within **Table 1**, for our entire sample of 27 European Union countries, for the 2005–2020 available data. The average GDP of the EU in current US dollars is 32355.59, with the largest values attained in Luxembourg and the lowest values attained in Bulgaria and Romania. Romania and Slovakia have the lowest HDIs in the last reported year, while Germany, Ireland and Sweden lead in sustainable development. From the point of view of the financial and economic crime analysed proxies, the countries with the least developed such undesirable phenomena are Denmark, Austria and Estonia, while the countries with the highest economic crime levels are Bulgaria, Greece and Lithuania.


**Table 1.**

*Summary statistics of dependent and independent variables.*

The underlying relationships between the development proxies on the one hand and the financial and economic crime proxies on the other hand may be depicted from the graphical representations of one against each other from **Figure 12** that also contain the linear fit of data.

**Table 2** projects the correlation matrix between our variables. In order to fulfil the basic assumptions of multivariate data analysis through regression modelling, most variables are used with their natural logarithmic transformation [37], except for the Money Laundering variable. One may easily notice the indirect relationship that exists between GDP and HDI on the one hand and Corruption and Shadow Economy respectively on the other (negative correlation coefficients) and then the direct relationship that exists between GDP and HDI on the one hand and Money Laundering and Cybercrime respectively on the other (positive correlation coefficients).

Our unbalanced panel data are modelled through simple regressions, using the Pooled OLS method, in order to estimate the impact of financial and economic crime proxies upon the economic and human development. The resulting log–log and log-linear models have the following baseline equation:
