Anyone who engages in econometrics pays attention to this account
Manuscript: [email protected]
All the code programs, macro and micro databases and various software of the econometric circle methodology are placed in the community. Welcome to the econometric circle community for exchanges and visits.
Today, we recommend the latest publication of "Weak Country: The Origin and Influence of the Sicilian Mafia" published by the Great God Mao Guru in RES (one of the Top 5) in 2020. Regarding Daron Acemoglu, you can take a look: ①Who is the god who has published the most articles on top5 publications in 40 years? ②The top 500 list published in Top5 journals is released. These Chinese people deserve their names. ③Some interesting instrumental variables in Mao Guru's paper! ④Ghost, "Daron Acemoglu Production Function" was born! ⑤Asimoglu won the MIT's highest faculty honor—"College Professor", ⑥The three most promising Nobel economists in 2020! br /> ** ** text
on the bottom of the text content, author: Lisong Ze, Economics and Management Institute of China Central University of Finance and Communications E-mail: [email protected]
Note: The original is more detailed suggestions after reading this review, to text Read the original text later. There is another point worth noting in this article: Although it is an ordered dependent variable, the author can still use OLS for regression analysis.
Reference: Daron Acemoglu, Giuseppe De Feo, Giacomo Davide De Luca, Weak States: Causes and Consequences of the Sicilian Mafia, The Review of Economic Studies, Volume 87, Issue 2, March 2020, Pages 537–581.
1. Introduction
Increasing evidence shows that weak state governance hinders the provision of public goods and economic development. Among them, various criminal organizations have had a very negative impact on regional development. In the past 150 years, the "Mafia" has played a key role in Sicily's economy and politics, and is often considered the primary reason why Sicily's economic and social development lags behind other parts of Italy. What the Sicilian Mafia has in common with other criminal groups is that these organizations partially fill the gaps in the rule of a weak state and contribute to the continued weakness of the state system and the underdevelopment of the economy. In this article, we try to explain the origin and influence of the Sicilian Mafia.
The first part of the article provides an explanation for the expansion of Mafia forces in Sicily in the 1890s. We believe that this is closely related to the large-scale socialist movement "Fasci" initiated by the peasant group in Italy. Especially after the severe drought in 1893 caused a sharp decline in agricultural production, Fasci's development momentum has become even more rapid. Due to the lack of strong national rule to respond to the socialist movement and the demands of the peasants, landlords and managers of large agricultural real estate can only seek assistance from the Mafia. Our empirical test confirms the above view.
The second part of the article uses the connection between the drought of 1893 and the Mafia to estimate the impact of criminal groups on local economic development and other social outcomes. We used the drought in 1893 as an instrumental variable to estimate the mid- to long-term effects of the Mafia under the 2SLS framework. Through estimation, it is found that the Mafia had a great influence on the economic conditions (such as the literacy rate) in the early 20th century, which at least partially reflects the low ability of the local government to provide public goods. Our research has also found that political dominance is also the cause of the mafia’s major impact on the local economy. In addition, our estimates also found similar long-term negative effects, although these effects are relatively small and unclear compared to the medium term.
There are two main differences between this article and previous studies. First, this article emphasizes that the decisive factor for the proliferation of the Mafia at the end of the 19th century was a series of events that occurred during the rise of peasant organizations, rather than long-term structural differences in different regions of Sicily. Secondly, this article focuses on the historical evolution of the Mafia.
2. Historical background
This part briefly introduces the historical background of the origin of the Mafia in Sicily, the agricultural situation in Sicily at the end of the 19th century, and the rise of the Fasci movement of farmers. Interested readers can view the original text (already attached to the text).
3. Data and descriptive statistics
Our database includes data for 333 cities provided by Sicily's 1853 cadastral table. These cities belong to 24 districts in 7 provinces. The data for subsequent periods (sometimes more scattered) will correspond to the initial 333 cities.
The main variables we studied are:
-
Mafia 1900: The prevalence of mafia in various cities in 1900, with a value between 0 (no mafia) and 3 (there is a large number of mafia).
-
Peasant Fasci: If there is a farmer Fasci organization, it is equal to 1, otherwise it is 0.
- Relative Rainfall 1893: The relative rainfall in the spring of 1983.
The following table shows the descriptive statistics of the variables used in the article. For the specific definition and source of each variable, please refer to the original text:
4. Reasons for
the expansion of the Mafia in the 19th century This section empirically examines the influence of the drought in 1893 and the Fasci movement of farmers on the expansion of the Mafia in Sicily.
We first verified that the drought in the spring of 1893 had a huge negative impact on local agricultural production. We use data from the Director-General of Agriculture (Direzione Generale dell'Agricoltura) between 1886 and 1896 to measure the production of 20 crops in 24 regions in 1893. In order to control the potential differences in agricultural productivity between regions, we divided the region's output in 1893 by the region's average output in 1885-1892 and 1894-1895. Combining these data with the relative rainfall in the spring of 1893, we estimate the following equation:
Here, the picture represents the relative rainfall in the area, which indirectly reflects the severity of drought; the picture measures the relative importance of spring rainfall to crop c. The following table shows the regression results:
Columns 1 and 2 did not include interactive items, focusing on the impact of the relative rainfall in the spring of 1893 on the overall crop production. It can be seen that crop yields are also lower in areas with more severe droughts (low relative rainfall). Next, we focus on the picture of the interaction coefficients between the relative rainfall in the spring of 1893 and the relative importance of rainfall to crops. The sixth column is our preferred recognition result, which includes the complete regional fixed effects. In this case, the estimated value of the picture is 1.45, which means that in areas affected by severe drought in the spring of 1893 (corresponding to the first quartile of rainfall distribution), the production of wine (picture) will fall by 6% , The output of olive oil (picture) will drop by 32%, and the yield of wheat (picture) will drop by 85%.
Overall, this evidence confirms that the drought in 1893 had a serious impact on agricultural production in parts of Sicily, especially on crops that depend heavily on spring rains, such as wheat.
Next, we estimate the cross-sectional relationship between farmer Fasci and the drought of 1893:
The picture represents whether there is a dummy variable of farmer Fasci in city i, and the picture is a series of control variables. The above equation is our first stage in estimating the relationship between farmer Fasci and the development of the Mafia. We use 245 cities in the sample to estimate the above equation, and the results are shown in the following table:
Among them, panel A is a rough estimate of the relationship between farmers' Fasci and relative rainfall, while panel B controls the fixed effect of the province. At the same time, the second to fourth columns of the above table gradually controlled 1) variables related to the appearance of Fasci, 2) variables related to the appearance of the Mafia, and 3) a series of control variables of geographic factors. Quantitatively speaking, the coefficient estimated value of -0.72 in the fourth column of Panel B indicates that in cities with severe drought, the probability of the emergence of farmer Fasci organizations is 0.35 higher than in cities with sufficient rainfall.
Exploring the influence of farmer Fasci on the expansion of the mafia is the core content of this part. In the previous article, we put forward the following hypothesis: because landlords and rural real estate managers need to seek help from the Mafia to fight against peasants, the Mafia expands to areas where Fasci is a powerful farmer in Sicily. We test this hypothesis by estimating the following equation: 
where the picture is an indicator of the strength of the mafia in the city i, taking a value between 0 and 3 (the larger the value, the stronger the power).
It should be noted that using OLS to estimate the above equations cannot obtain a consistent estimate of the picture. On the one hand, the birthplace of the peasant organization itself may not have strong resistance to the mafia, which will lead to overestimation of the picture; on the other hand, because the mafia will confront the peasant organization (as we assume), then it is suitable for the mafia. In areas where the party is expanding, it is difficult to establish peasant organizations themselves, which in turn will cause an underestimation of the picture. To overcome this problem, we use an instrumental variable identification strategy, in which the equation in section 4.2 is the first stage estimate. The premise that this strategy is effective is that the error term picture in the above equation has nothing to do with the instrumental variable picture we choose. This assumption is reasonable, because although many areas of Sicily experienced drought in 1893, the real cause of the serious consequences of this drought was the rise of the Fasci Farmer Organization during this time. According to our investigation, the droughts around 1893 will not have an impact on the Mafia or other socio-economic results. Therefore, our causal channel is: the drought has strengthened the power of the peasants, prompting the landowners to turn to the mafia to counteract the power of the peasants.
The following table is the result of 2SLS estimation of the above equation:
Comparing the estimated coefficients of instrumental variables in the first column of panels A and B, it can be found that comparing the development of the mafia among different provinces cannot provide effective information, so you should choose to compare within provinces. From a quantitative point of view, the estimated value of the coefficient in the fourth column of Panel B means that the power of the Mafia will increase by 1.5 in cities where Fasci farmers appear. This order of magnitude means that in 1900, up to 38% of the Sicilian Mafia forces were deployed to confront the farmer Fasci. Therefore, we are concerned that this period may have played a key role in the mafia's dominance on the island.
Panels C and D in the above table show that the OLS correlation between farmer Fasci and the Mafia is significantly weaker. We think this is because the Mafia caused difficulties for the establishment of Fasci, a farmer from the very beginning. We designed empirical strategies in the appendix to confirm this conjecture.
In the next study of the mid- to long-term effects of the Mafia, we will directly use rainfall as an independent variable, ignoring the causal channel of farmer Fasci. To prepare for the identification, we will directly estimate the relationship between the drought variable and the mafia:
The estimation results are shown in the following table, where panels A and B use OLS estimation combined with the bootstrap standard error that allows two-way clustering, and panels C and D use maximum likelihood estimation. It can be found that the results obtained by the two methods are very similar.
So far, we have explained the expansion of the Mafia since 1893 from the "demand side", that is, the forces that landlords need to contend with peasant organizations. So is there a "supply side" factor? Our findings in the appendix show that areas such as the Mafia’s pre-existence and the intensity of its development have no obvious impact on its subsequent expansion. We believe that once there is a demand for its services, the Mafia can always find a way to gain a foothold in a city, regardless of whether it has previously developed power locally or in the surrounding area. This is partly due to the general lack of rule of law in Sicily and the presence of a large number of local villains who are willing to join the Mafia.
5. The Mid-term Influence
of the Mafia This part of the article explores the influence of the Mafia on a number of economic outcomes in the early 20th century. We will estimate a cross-sectional model of the following form:
We will use the 2SLS method to estimate the above equation. The first stage equation is as follows, using the relative rainfall in the spring of 1893 as the instrumental variable:
In addition, in the follow-up identification, we will also control the provincial fixed effects, because some economic results have obvious inter-provincial differences.
First, we use the literacy rate as a proxy variable to examine the influence of the mafia on human capital. The data we use comes from the census data of 1911, 1921, and 1931 (for people over 6 years old). The results are shown in the following table:
Columns 1-3 of the above table report the influence of the Mafia on the literacy rate of the population in 1911, and it can be found that it is not significant. This is because the literacy rate is a stock concept, so even if the mafia can affect human capital, it is difficult to show it in the short term. Columns 4-6 and columns 7-10 respectively indicate that the power of the Mafia has a significant negative correlation with the population literacy rate in 1921 and 1931. From a quantitative point of view, if the power of the Mafia increased from 1 to 2 (from weak to obvious) in 1900, then this would cause the literacy rate in 1921 to drop by about 10 percentage points. It is worth noting that if the OLS method is used for identification, the estimated value obtained is almost zero. We believe that the reason for the difference between the results of 2SLS and OLS is the selection problem, that is, the Mafia tends to choose cities with better economic prospects, which will cause upward bias.
We believe that one of the main reasons why the development of the Mafia has a negative impact on the economy is that it reduces the ability and incentives of the local government to provide public goods and protect citizens. The following table provides evidence for this view, which shows that the development of the Mafia has a significant negative impact on the provision of public goods:
It can be seen that the expansion of the mafia power will significantly increase the infant mortality rate (columns 1-3) and significantly reduce the development expenditure of citizens (columns 4-6) and the proportion of judicial personnel (columns 7-9).
The mafia may also influence economic results by interfering in politics. In columns 10-12 of the above table, we examined the influence of the Mafia on the distribution of votes among candidates in the 1909 parliamentary election. We are concerned about the Herfindahl index of vote share in this election. It can be seen that the Mafia had a very obvious influence on the political competition in 1909. For example, when the power of the Mafia developed from weak to obvious, the Herfindahl index of vote share in 1909 increased by 30 percentage points. The huge connection between the Mafia and local politics is probably caused by the intimidation and fraud of voters by the former.
6. Long-term results
The expansion of the Mafia at the end of the 19th century may have long-term effects through two complementary channels. The first is the indirect channel, that is, the emergence and expansion of the mafia in different cities have caused these cities to enter different economic and political development tracks, which are of great significance to many results in the late 20th century. The second is the direct channel, that is, the local mafia may have existed in 1900 and the following decades, which has a direct impact on the economic and political results of the second half of the 20th century.
We first use human capital as a proxy variable to estimate the long-term impact of the mafia on economic outcomes. The dependent variables of specific concern are the population literacy rate between 1961, 1971, and 1981 and the proportion of the population who completed high school education. The estimated results are shown in the following table:
Panels A and B of the above table show that the mafia power in 1900 and the two proxy variables of human capital are negatively correlated, although the coefficient of literacy rate after 1971 and the coefficient of high school education in 1981 have lost statistical significance. Panels C to E examine the supply of public goods from three aspects, namely, infant mortality, development expenditure, and the availability of aqueducts. Although the point estimates for these variables are all negative, their significance is not high. The above results show that the negative influence brought by the Mafia is persistent, but its extent is significantly lower than in the first 50 years of the 20th century.
Next we examine the long-term influence of the Mafia on politics. Panel F in the table above reports the influence of the strength of the Mafia in 1900 on the Herfindahl Index in 1963, 1972, and 1983. The emergence of the Mafia has made the distribution of votes more concentrated. This impact is significant, but in terms of quantity, it is significantly lower than the mid-term impact.
In 2020, the number of citations has reached 4000.
Long press the above QR code to read the original PDF of Acemoglu
If you are eager to study the Top articles of the world's top management scholars such as Acemoglu and Card, you can leave a message in the message area of this article, and then follow the process of replying to you.
Extended reading:
① "13 Empirical Studies" Kung Fu Secrets, a must-have kit for young and middle-aged scholars! ①Summary of 120 classic empirical literatures on various causal identification methods", ②The newly revised classic masterpiece of causal inference from Harvard University is free to download! With data and code, ③Summary of statistical methods of causal inference, 177 documents, ④Policy evaluation A review of the measurement methods, including the latest causal inference methods, ⑤Do you use IV, RDD, DID, PSM more in the education field? Use specific literature, ⑥After reading the top journal articles, organize the endogenous treatment booklet, ⑤Innovative tool variables Explain, to ensure that you will never forget it forever, ⑦DID, synthesis control, matching, RDD four methods comparison, applicable scope and characteristics, ⑧32 selected Articles about double difference method DID!⑨About (fuzzy) breakpoint regression design 100 selected Articles! ⑩Matching method (matching) operation guide, 16 articles worth collecting, etc., ⑪ MIT's widely circulated policy "processing effect" reader, ⑫DID research trends and literature review applied in policy evaluation, ⑬Four methods of the latest policy effect evaluation, ⑭Basic problems of policy effect evaluation.
The following short-linked articles belong to a collection, you can collect them and read them, or you won't find them in the future.
In 2.5 years, nearly 1,000 non-weighted measurement articles in the econometric circle,
You can search for any measurement related issues directly in the official account menu bar,
Econometrics Circle