The impact of crime on foreign investment
The negative effect of homicides and robberies on flows of Direct Foreign Investment into Mexico

In the third quarter of the year, Direct Foreign Investment (DFI) in Mexico fell by 75.7% compared to the second quarter of the year. Prior to this, Mexico was following a positive trend in the reception of foreign investment. According to the 2017 United Nations World Investment Report, in 2014 the nation ranked 13th worldwide among the countries that received the most FDI, thanks to its good macroeconomic performance and the recent structural reforms in the energy, telecommunications and financial sectors. However, despite the country’s good economic performance, it is possible that FDI inflows are being affected by the high growth in violence in various geographical areas of the country.

Crime levels in recent years in various states of the country have created a climate of insecurity which is affecting the performance of various economic indicators, such as FDI, at the regional level. In the article "Direct Foreign Investment in Mexico, Crime and Economic Forces," published in the academic journal Contemporary Economic Policy, we examine the effect of different types of crime on the flows of Direct Foreign Investment (DFI) into the 32 states of Mexico using quarterly data for the period of 2005-2015. Our work considers alternative FDI specifications to identify its determinants.

Results suggest that homicides and robberies, specifically, have a statistically significant negative effect on FDI inflows, while property crimes and other crimes, including kidnappings, extortion, fraud, etc., have no effect.

Violence drives away DFI

The previous hypothesis is based on various bibliographical contributions, for example Daniele and Marani (2011 in the case of Italy, and Ramos and Ashby (2013) for the case of the Mexican economy, asserting that there is an inverse relationship between violence and FDI. Additionally, there are several studies analyzing the impact of violence on the Mexican economy where different economic variables are analyzed, showing it can be directly or indirectly linked to FDI.

For example, Pan, Widner and Enomoto (2012) in a state-level study for Mexico note that the per capita GDP growth rate is negatively affected by the crime growth rate in neighboring states. BenYishay and Pearlman (2013) found that increases in the homicide rate affect the number of hours worked in various segments of the labor market, with the self-employment segment being the most affected. The same authors in a study for 2014 found evidence that an increase in robberies reduces the likelihood that small businesses in Mexico will expand their operations.

In summary, several empirical studies exist that document the negative impact of crime on various economic variables, such as the gross domestic product (GDP) per capita, hours worked and labor productivity, to name but a few. But what about DFI?

Unequal impact on the Mexican states

This work involved conducting an analysis at the state level for the years of 2005 to 2015. It is important to note that FDI inflows received by the country are highly concentrated in a few specific entities. For example, the five states that received most FDI during the period analyzed in this study were Mexico City, Nuevo León, State of Mexico, Chihuahua and Jalisco, which together captured 51.4% of the total FDI that entered the country. In contrast, the five states that received the lowest FDI in that period were: Yucatan, Tlaxcala, Campeche, Colima and Chiapas. Together, these entities only attracted 2.4% of the total FDI received in the country during the analyzed period in this research.

Moreover, crime is highly concentrated in some entities in the country. For example, the states with the highest incidence of homicides, robberies, property crimes and other crimes in recent years are: Sinaloa, Baja California, Baja California Sur and Yucatan.

How we measure the impact of crime

In our study, we analyzed how different forms of crime have had an effect on DFI inflows to the country. We also included in our analysis other control variables, which have been recognized in the literature as determinant variables of FDI flows, such as the real salary, the interest rate, the real exchange rate and electricity sales, the latter as an approximation to the existing level of infrastructure. Additionally, we incorporated into the study a dummy variable that enabled an analysis of whether the economic crisis of 2008 and 2009 had any significant effect on the FDI flows into the country during that period.

Through an econometric model of panel-type data with fixed effects, we analyzed the 16 states with the least crime and the 16 states with the most crime. In both cases, we performed a regression without including crime and compared it to other regressions in which said variable, in its various modalities, appeared to be determinant. It is important to note that these estimates were made using a fixed-effects model, which enables the identifying of the number of quarters that the crime takes to have an impact on FDI, in addition to estimating the cumulative magnitude of that impact. Additionally, we created a second dynamic model using the generalized method of moments (GMM), which also included the different types of crime.

The collateral damage of robberies and homicides

The results of our estimates agree with the theory on the determinants of FDI. They indicate that the interest rate, the real exchange rate and wages are relevant FDI determinants, while the effects of infrastructure and the economic crisis of 2008 were not statistically significant.

In addition, results show that among all analyzed crimes, , only homicides and robberies have a significant impact on FDI at state level. . Basically, results indicate that, on average, a 1% increase in the number of homicides causes an FDI reduction of up to 0.28% as a total effect in the following five quarters after an increase in homicides. Additionally, a 1% increase in the number of robberies causes a fall in FDI of up to 0.33% as a total effect a year and a half later.

Moreover, the estimates made for the regressions separating the 16 states with more and less crime suggest that the effect is only important in the 16 states with the highest level of crime. For these states, the crime takes less time to affect the FDI. Finally, the dynamic specification shows that homicides and other types of crime negatively affect FDI flows arriving in the country in a statistically significant way.

This article contributes to an understanding of the effect that insecurity exerts on a fundamental variable in the country's economic growth and, therefore, highlights the importance of using public resources to reduce crime levels in Mexico. This work also suggests that the effects that each type of crime exerts on investment decisions are differentiated, an element that must be considered in the design of the country's security strategies.