I began studying economics at Hitotsubashi University in the 1990s, and so it was an honor to return to Kunitachi last fall with the support from the Center for Economic Institutions. During the visit, I constructed a dataset of Thai enterprises from 1940 to 2016, to document historical developments of the Thai military as a business group. This is an unconventional subject, given that I am a macroeconomist whose publications have been on capital flows, exchange rates and international price differentials. Why did I get into this subject?
Thailand, where I am from, has two unique characteristics related to the military. First, there is a military-owned commercial bank, and it was literally named the Thai Military Bank. Second, Thailand has had frequent military coups after becoming a constitutional monarchy in 1932. On average, there is one military coup in five years. The latest one occurred in May 2014, and the military junta has remained in power since then.
Shortly after the latest coup, the military junta moved the investment promotion agency out of the Ministry of Industry, and placed it directly under the Office of the Prime Minister. The military junta also appointed the governor of the central bank in the following year. New laws related to insurance, investment and industrial policy were enacted. These post-coup changes and the existence of the military bank raise the following questions. Is it possible to characterize the Thai military as a group of for-profit enterprises? If so, when did the group emerge?
My dataset includes three types of enterprises. First, military firms are defined as for-profit companies owned by the military. Second, military enterprises are defined as military units that produce non-security goods or services for sale. Finally, military-related firms are defined as companies owned or directed by a military officer, or awarded a production-for-sale concession by a military unit. I rely on the Business Registration and Records from the Ministry of Commerce, companies’ annual reports from the Stock Exchange of Thailand, and the Government Gazette.
My dataset indicates that the military-related firms emerged before other types of enterprises. There were military-related transportation companies from 1940, military-related commercial banks from 1942, and military-related trading companies from 1948. The military enterprise first emerged in 1948 as a commercial airport. In 1957, the military firm emerged as a commercial bank, and the military began broadcasting television. From the 1960s, all types of enterprises increased in number and industry coverage.
In sum, seven military firms (in commercial banking, non-bank financial services, and media) and more than twenty military enterprises (in media, real estate, recreation, and transportation) form the core businesses of the Thai military. There are more than one hundred military-related firms (in banking, media, power generation, real estate, recreation, trading, transportation, a wide range of manufacturing industries and primary industries). Most of them are related to the military via military officers on the board the directors.
(Adjunct Research Associate, IER, Hitotsubashi University)
|
|
"Historical data and Hitotsubashi University"
It is my pleasure to stay one more year here at IER, Hitotsubashi University. I’m working on macroeconomics and macroeconomic theory. One of my works uses historical data provided by IER-LTES database. So in this column, I would like to introduce a benefit of studying historical data for answering macroeconomic questions.
The word `data’ means a collection of past events. Macroeconomists try to understand these past events, construct a model which is consistent with these events, and use the model to reason (or guess!) what would happen in the future or a counterfactual policy scenario. This approach is often called a structural approach.
Obviously, this approach is not the only approach macroeconomists have. We can understand the economy by studying natural experiments. For example, Professor Nirei, a former colleague of Hitotsubashi University, studies the effects of production chain on the aggregate economy by examining exogenous variations caused by the Great East Japan Earthquake. While this approach is quite popular among applied microeconomists, as argued by
Nakamura Steinsson (18), we do not have a lot of “good” experiments in order to answer macroeconomic questions. Put differently, we seldom have natural experiments which test what we want to know. Therefore this approach is less systematic than the structural approach is, which is a part of the reasons why this approach is less familiar among macroeconomists.
However, if we dig into really old data (historical data), we might be able to find a lot of good natural experiments, which have not examined yet. For example,
Bernstein et al. (2019, JPE) examine a role of clearinghouse by studying the economy before and after the introduction of a clearinghouse to NYSE around 1890. Because of the recent financial crisis, policymakers and academics both have discussed roles of financial intermediaries and whether the government should regulate them. The introduction of clearinghouse by NYSE is a particularly good natural experiment for answering the policy question, and we do not have such a good experiment in recent data. Therefore this research highlights one reason why studying historical data can give us a better answer to a certain macroeconomic question.
The LTES dataset provided by IER covers an exciting period, too. For example, Japan had experienced high inflation and high deflation around WW1. During WW1, the inflation rates were around 18%, and from 1918 to 1924, the deflation rates were on average 10% annually. Given the fact that recent Japanese deflation was around 1% at most, it is clear how big the deflation was. Also, while developed countries have been experiencing low and stable inflation, the debate about inflation never stops. For example, recent debates about MMT are an example of why inflation is still a big policy question. By (re-)examining past inflation period, we could shed new light on recent policy questions. The LTES dataset is well-suited for this purpose since the LTES dataset covers a wide range of economic activities. Moreover, we can download the entire dataset quite easily.
I have been working on wage determination and rigidity around this period with supports by faculty members here at IER. Professor Kambayashi and Moriguchi were particularly encouraging, and I really appreciate the research environment of this institute. I hope that I can share my future findings soon in this letter, which is one of my goals of this year. The other goal is advertise this amazing institute with good historical data and great faculties to macroeconomists who are less familiar with historical data.