Firms’ Trading Behavior in the Global Economy: Evidence from Japanese Transaction-level Customs Data*

　This article analyzes firms’ trading behavior using Japanese transaction-level customs data in 2017, focusing on trading partner countries, airport/seaport use, declaration frequency, and product variety. We find that 45% of exporters and 36% of importers used only one port (air or sea), while some firms used over 20. Additionally, 53% of exporters and 46% of importers made fewer than or equal to nine declarations during the year, whereas approximately 10% made more than 190 declarations per year. Regression results show positive correlations among most margins—the number of partner countries, ports, and product variety. However, we find that the number of export destinations is negatively correlated with export value per declaration, and there is no correlation between the number of import products and import value per declaration.

 

Keywords: international trade, firm behavior, Japanese transaction-level customs data

JEL Classification: F14，L25

* This study is an outcome of the research conducted jointly with the Policy Research Institute (PRI) after submitting a request for use of custom’s export and import declaration data to the Japanese Ministry of Finance (MOF) based on the “Guideline on the utilization of custom’s export and import declaration data in a joint research with Policy Research Institute,” and receiving approval in February 2022. The views expressed in this study are those of the authors’ personal responsibility and do not represent the official views of MOF or PRI.

Top of the page

---

What can we learn from customs declaration data? -Invoice currency selection and exchange rate pass-throughs*

　This study was the first phase of a publicly solicited research project using customs declaration data and involved the following two analyses. First, we analyzed the factors behind the share of Japanese imports and exports in the choice of the invoice currency, organized by partner country, industry, and a country/industry cross-table. Second, we constructed panel data at the firm-good-transaction level and analyzed the determinants of invoice currency choice and exchange rate pass-through (ERPT).
　The invoice currency shares compiled from the customs declaration data showed a change in the invoice currency share that was not evident in the Ministry of Finance's past publications, confirming the necessity of developing and analyzing data on the invoice currency share by country. At the transaction level, Japanese exporters tended to choose the same invoice currency as their competitors’ invoice currency and the invoice currency of their imports, and to choose the local currency, especially for intra-firm trade. In addition, unit prices were calculated using import/export value and volume data, and an empirical analysis of the ERPT for imports and exports was conducted using exchange rates considering invoice currency. For the export ERPT, we found that in terms of the invoice currency, the local currency denomination was the most responsive to exchange rate fluctuations. In contrast, at the local price in the importing country, the yen or the vehicle currency denomination (in US dollars) was the most responsive to the exchange rate. Regarding the import ERPT, we showed that differences in import ERPT occurred between the strong yen and weak yen phases.

 

Keywords: customs declaration data, invoice currency, exchange rate pass-through, intra-firm trade, currency matching, strategic complementarity

JEL Classification: F23, F31, F33

* This study is an outcome of the research conducted jointly with the Policy Research Institute (PRI) after submitting a request for use of custom’s export and import declaration data to the Japanese Ministry of Finance (MOF) based on the “Guideline on the utilization of custom’s export and import declaration data in a joint research with Policy Research Institute,” and receiving approval in February 2022. The views expressed in this study are those of the authors’ personal responsibility and do not represent the official views of MOF or PRI.

Top of the page

---

The ‘Flexible Declaration Scheme’ of Japan Customs: Prevalence, Determinants, and Consequences*

　This article examines the prevalence levels, determinants, and consequences of the Flexible Declaration Scheme (hereinafter referred to as "the Scheme"), which allows exporters and importers to declare transactions at a customs office where their shipments are not stored, introduced in October 2017. Regarding the prevalence levels, our results show that the utilization rate increased by 27 percentage points for exports and 18 percentage points for imports over around four years after the introduction of the Scheme. We also show that the utilization rate varies significantly across sectors, regions within Japan, destination regions, and years. Regarding the determinants, we find that customs brokers play the most significant role in explaining variations in the utilization rate relative to other factors such as trading partner countries and products. Regarding the consequences, we find that exporters and importers using this Scheme reduced the number of customs offices used for declarations and storage by more than 50%, suggesting that the Scheme resulted in a consolidation of customs clearance operations.

 

Keywords: import/export declaration data, international trade, customs procedures, trade facilitation, AEO

JEL Classification:F14, F23, F68

* This study presents the research findings of an internal research project conducted by researchers of the Policy Research Institute (PRI), Ministry of Finance (MOF) Japan, using export and import declaration data. When conducting this project, we strictly follow all prescribed procedures regarding data usage and publication of findings to ensure the confidentiality of the data. The views expressed in this paper are those of the authors and do not represent the official views of the MOF or the PRI.

Top of the page

---

Significance of analysis based on tax data in Japan*

　In this paper, we explain the significance of the academic use of tax data in Japan, focusing on the example of analyzing the income distribution of high-income earners. After describing the current academic use of tax data in Europe and the U.S., various empirical studies using tax data, and past research on the income distribution of high-income earners in Japan, we will outline the joint research using tax data initiated by the National Tax Agency and explain our estimation of the Pareto coefficient of the income distribution of high-income earners as an example of research results using income tax return data. The estimated Pareto coefficient for the income distribution of high earners in Japan in 2020 is around 1.45, which is much lower than the historical level, indicating that income concentration among very high earners has increased, especially in capital income. This point was made clear by the availability of income tax data in this joint study with the IRS. The academic use of tax data in Japan has important implications not only from an academic perspective, but also from a policy perspective.

 

Keywords: income distribution, income tax, tax data

JEL Classification: H21, H24

* This study is a part of "Joint Statistical Research Program" conducted in collaboration with the National Tax College (NTC), under the approval of the National Tax Agency (NTA) (in March 2022), in accordance with "Guideline on the Utilization of National Tax Data in the Joint Statistical Research Program." The views expressed herein are those of the authors and do not necessarily reflect the views of NTC or NTA.

Top of the page

---

The Burden Reduction and Redistribution Effects of Deductions in Income Tax: An Analysis Using Tax Data*

　The deduction system is one of the most important policy issues in Japan's income tax system. In capturing how tax deductions affect the progressive burden structure of the income tax, it would be beneficial to have a data-based understanding of the burden reduction effect of deductions and their impact on income redistribution. This paper uses tax data from income tax returns held by the National Tax Agency to quantitatively evaluate the burden reduction and redistribution effects of the tax deductions and credits. In income tax, the burden reduction effect (the ratio of the amount of reduction to income) is largely contributed by income tax deductions rather than tax credits, and the burden reduction effect is larger for higher income brackets. This is largely due to a system in which the amount of applicable deductions increases with income, such as the income-increasing part of the employment income deduction and the deduction for social insurance premiums. This paper also showed that tax deductions reduce the redistributive effect, and that income-increasing deductions contribute to this result.

 

Keywords:income tax, tax deductions, redistributive effects, tax data

JEL Classification:D31, H24

* This study is a part of "Joint Statistical Research Program" conducted in collaboration with the National Tax College (NTC), under the approval of the National Tax Agency (NTA) (in March 2022), in accordance with "Guideline on the Utilization of National Tax Data in the Joint Statistical Research Program." The views expressed herein are those of the authors and do not necessarily reflect the views of NTC or NTA.

Top of the page

---

Tax return data as income data*

　This paper evaluates tax return data on personal income tax as income data and examines its properties, usefulness, and limitations. Tax return data is based on information from tax returns submitted by taxpayers to report their income and tax payments, and its size and high accuracy are major advantages. On the other hand, it does not include information on individuals who are not required to file income tax returns or on non-taxable income, so there are certain limitations to understanding the income situation in Japan as a whole.
　There are approximately 23 million tax filers each year, and approximately 5 million are dependents of tax filers. The combined total of both is only about 30% of the total population aged 15 and older, and tax return data alone does not provide a sufficient picture of the income distribution in Japan. In particular, the majority of salaried workers and pensioners do not file tax returns, and another data supplement is needed to describe the income of these individuals.
　Many salaried workers and pensioners who file tax returns are eligible for Deduction for Medical Expenses and Special Credit for Loans, etc. Related to a Dwelling. Given the nature of these deduction and credit, it is assumed that filers are often in special economic circumstances, such as illness or immediately after purchasing a home, and that they are not a national-representative group.
　In contrast, comprehensive information is available for business income earners who are roughly equivalent to self-employed persons. The number of business income earners in the tax return data is close in level to the number of self-employed in an official statistics, and the amount of income reported is generally consistent with data from the National Accounts. Since more than half of all filers file tax returns almost every year, tax return data are useful as panel data for self-employed individuals.
　In addition, the study examined the use of tax return data as data on top income earners. The top 0.1% of the population, or very high earners, can be comprehensively captured in the tax return data. However, when expanded to the top 1%, there are a non-negligible number of people who do not file a tax return. In addition, due to the existence of withholding taxation, the coverage of interest and dividend income is inadequate and likely to underestimate the income of high-income earners.
　Future work includes the construction of comprehensive income data by integrating with other data on salaried and pension income earners. Efforts are also required to compare tax return data with existing statistical surveys to more precisely understand the actual distribution of income.

 

Keywords: tax returns, income data, top income earners

JEL Classification: D31, E01, H2

* This study is a part of "Joint Statistical Research Program" conducted in collaboration with the National Tax College (NTC), under the approval of the National Tax Agency (NTA) (in February 2023), in accordance with "Guideline on the Utilization of National Tax Data in the Joint Statistical Research Program." The views expressed herein are those of the authors and do not necessarily reflect the views of NTC or NTA.

Top of the page

---

Analysis of the actual status of loss-making corporations using individual data from corporate tax returns*

　The purpose of this report is to analyze the distribution of income, losses, and corporate tax amounts by capitalization, industry, and other attributes (e.g., non-family companies, specified family companies, and family-owned companies) using individual corporate tax return data, and to analyze the actual status of loss carryforwards and corporate tax amounts. First, we constructed panel data from available corporate tax return data for fiscal years 2014-2020, correcting for mergers, multiple returns within a fiscal year, missing measurements. During this period, more than 60% of all corporations reported tax losses.
　Next, we analyzed all corporations (excluding consolidated corporations, foreign corporations, and corporations on leave of absence or in liquidation) from corporate tax returns and found the following results. The distribution of corporations with loss amounts, which are not disclosed in Corporation Sample Survey, appears to be bunching at zero yen of tax loss (taxable income). However, the proportion of loss-making corporations with a net taxable income of zero yen was higher among corporations with capital of 100 million yen or less than among those with capital of more than 100 million yen. Corporations that were loss-making for seven consecutive years accounted for 36.5% of all corporations, and the smaller the capital size, the higher the percentage. Only 1.1% of all corporations applied the tax loss deduction for seven consecutive years, and 21.2% of all corporations did not apply the tax loss deduction for any of the seven years. The analysis of the transition between the years for corporations with capital of over 100 million yen and those with capital of less than 100 million yen showed that the probability of a corporation with capital of over 100 million yen reducing its capital to less than 100 million yen in the following year ranged from less than 4% to over 5% for each year, while the probability of a corporation with capital of less than 100 million yen increasing its capital to over 100 million yen in the following year was extremely small. The transition probability that a corporation with capital of more than 100 million yen would reduce its capital to less than 100 million yen in the following year was higher for loss-making corporations than for profit-making corporations. Furthermore, the number of corporations and the distribution of corporate tax amounts by corporate tax bracket were shown. Corporations paying more than 5 million yen in corporate tax accounted for about 5% of all corporations and paid about 95% of all corporate taxes. In addition, the larger the capitalization size, the higher the percentage of corporations paying more than 5 million yen.

 

Keywords: corporate tax, corporate tax return, tax data, loss corporation, capital reduction

JEL Classification: H25, H32, M41

* This study is a part of "Joint Statistical Research Program" conducted in collaboration with the National Tax College (NTC), under the approval of the National Tax Agency (NTA) (in February 2023), in accordance with "Guideline on the Utilization of National Tax Data in the Joint Statistical Research Program." The views expressed herein are those of the authors and do not necessarily reflect the views of NTC or NTA.

Top of the page

---

Using Corporate Tax Filing Data to Measure Business Dynamism*

　In this short article, using the corporate tax filing data in Japan, we examine the two items accounting for Japanese business dynamism (i.e., entry-exit and concentration) over the periods from 2014 to 2020. The annual-frequency data comprehensively accounting for the business activity (i.e., sales) of all the business enterprises in Japan suggest the following: First, both the entry and exit rates have been low and exhibiting stable dynamics although the COVID-19 pandemic resulted in a slight hike in the exit rate. Second, the business concentration has been also stable and even showed slight decline. These empirical patterns are consistent with the theoretical exposition in Miyakawa et al. (2022). Higher metabolism accompanied by the growth of a certain number of firms would be beneficial for Japanese economy.

 

Keywords: Tax filing data, business dynamism, entry and exit, concentration

JEL Classification: D22, L11, L25

* This study is a part of "Joint Statistical Research Program" conducted in collaboration with the National Tax College (NTC), under the approval of the National Tax Agency (NTA) (in March 2022), in accordance with "Guideline on the Utilization of National Tax Data in the Joint Statistical Research Program." The views expressed herein are those of the authors and do not necessarily reflect the views of NTC or NTA.

Top of the page

---