学生研究報告会

The impact of immigration on GDP per capita and labor productivity in the OECD countries is assessed within this study, including the way in which the impacts of immigration vary based upon the skill level of immigrants. It is our intention to expand on previous discussions regarding immigrant numbers and examine the economic channels through which immigration influences productivity. By examining the difference between GDP per capita and labor productivity (i.e., GDP per hour worked), it will be possible to determine if skilled and unskilled immigrants make different contributions to an economy's overall performance.

To accomplish the goals outlined in this study, we utilize OECD panel data from the years 2000 to 2015. We utilize two econometric methods. First, we apply two-way fixed effects regression equations to account for time invariant characteristics of each country and for global shocks that may occur during the study period. Secondly, we apply a two-stage least squares (2SLS) method to address potential endogeneity issues, for example, the issue of reverse causality. Reverse causality exists when high-income countries and/or countries with high productivity levels tend to attract more immigrants. We use lagged immigration variables as instruments under the belief that prior migration will predict future migration through the persistence of networks. However, we assume that prior migration will not influence short run productivity changes.

As indicated by our fixed-effect estimates, immigration is negatively related to both GDP per capita and labor productivity. A negative relationship exists both in terms of total migrant stock and in terms of the percentage of skilled migrants, however, there is no evidence of a statistically significant relationship for the percentage of unskilled migrants. We attribute the negative relationships found in our estimates to short term friction. Newly arriving immigrants may encounter various problems associated with newness to a foreign country, for example, difficulties with finding employment due to lack of experience, unfamiliarity with the new culture, etc. Additionally, because the capital stock may take longer to adjust than the labor supply, the short run measured productivity may decline. These transitionary mechanisms likely reduce measured productivity, regardless of the underlying human capital possessed by migrants.

On the other hand, when using 2SLS to correct for endogeneity, the results of the study are reversed. The results indicate that immigration, especially skilled immigration, has a substantial, positive, and statistically significant causal effect on both GDP per capita and labor productivity. The estimates indicate that a one percent increase in migrant stock is associated with a roughly 0.24 percent increase in income, with the skilled migrant population providing equal increases in income. Similarly, the estimated causal effects of immigration on labor productivity are slightly greater. Conversely, the estimates of unskilled migrant stocks remain statistically insignificant. Furthermore, the first stage of our 2SLS model confirms that lagged migration was a strong and valid instrument, with Kleibergen-Paap F-statistics exceeding the conventional threshold.

In summary, the difference between the estimates obtained via OLS versus 2SLS, and the previously mentioned findings, demonstrate that the negative correlations observed in fixed-effect models were not reflective of adverse effects of immigration; rather, they are indicative of reverse causality. The 2SLS findings demonstrate that, after accounting for reverse causality and possible endogeneity, immigration (especially skilled migration) generates substantial improvements in economic performance.

Additionally, we recognize several limitations of the study. While lagged migration can potentially remove some forms of endogeneity, lagged migration may not completely remove all forms of endogeneity. The five year intervals in which the data was collected are too long to include the dynamic nature of the labor markets. Although our study includes broad categories of education to differentiate between skilled and unskilled migrants, the categories may not accurately represent the varying degrees of occupational skills among migrants. Furthermore, national averages may obscure regional variations in the degree to which migrants are able to integrate into the local economies. Finally, additional research is required to investigate the mechanisms (innovation, firm productivity, investment, wage effects, and demographic stability) through which skilled migrants improve productivity.

In recent years, the number of foreign visitors to Japan has grown rapidly, reaching unprecedented levels. According to data released in 2025, monthly and cumulative inbound arrivals marked historic highs, exceeding 30 million visitors by September. Although the COVID-19 pandemic caused a temporary decline, inbound tourism rebounded strongly and continues to expand. This trend raises concerns about overtourism—such as congestion and violations of local norms—while simultaneously presenting opportunities for economic revitalization. Motivated by these dynamics, this study examines the determinants of inbound visitor numbers to Japan, focusing separately on tourists and short-term non-tourist visitors such as business travelers, workers, and international students.

Previous research has employed regression analyses to identify determinants of inbound tourism or has used surveys to explore motivations for travel to Japan. While these studies offer meaningful insights, they have several limitations. Some are based on relatively short time periods or coarse regional classifications, and others rely on representative GDP values rather than country-specific data. Survey-based studies clarify motivational “pull factors” but neglect structural “push factors” such as visa policies, economic conditions, or flight accessibility. Building on these gaps, the present study adopts a more comprehensive approach by analyzing long-term panel data from 68 countries (1988–2024). It incorporates economic, institutional, cultural, and geographic factors—including GDP, GDP per capita, visa exemption, OECD membership, population, direct flights, the Global Peace Index, staple food culture, religion, EPAs, regional groupings, exchange rates, and cultural trade values.

The analysis uses PPML as the primary estimation method, with OLS included for robustness checks. Due to differences in data availability across variables, the study estimates sixteen separate models for tourists and non-tourists. Across all models, PPML and OLS produce consistent results, confirming the reliability of the findings.

For inbound tourists, the strongest determinants include staple food similarity (rice staple), Asian regional proximity, visa exemption, availability of direct flights, and GDP. Countries where rice is the staple food exhibit significantly higher tourist numbers, suggesting cultural familiarity in dietary habits. Likewise, Asian countries—geographically and culturally closer to Japan—show consistently higher inbound flows. Visa waiver agreements and direct flights also substantially increase visitor numbers, highlighting the importance of reducing administrative and travel costs. By contrast, negative effects are observed for regions such as Africa, Europe, the Middle East, and South America, where distance, limited flight connectivity, or cultural differences may pose barriers. Religious variables show similarly strong negative effects for visitors from Muslim-, Hindu-, and Jewish-majority countries.

For non-tourist visitors, the results show considerable overlap but also important distinctions. As with tourists, regional and religious factors remain influential. Residents of Asia show high visitation regardless of purpose, while those from Africa, the Middle East, and South America are less likely to visit. However, several differences emerge. First, non-tourist visitors are even more strongly influenced by regional factors, especially the negative coefficients for African, Middle Eastern, and South American countries. Second, while exchange rates affect both groups, tourists are more sensitive to currency fluctuations, suggesting greater price elasticity in leisure travel compared with work or study-related mobility. Third, the Muslim dummy shows stronger negative effects for non-tourist visitors, indicating that cultural or institutional barriers may be greater in non-tourism contexts such as business or short-term residence. Conversely, OECD membership and direct flights positively influence both groups, underscoring the structural nature of mobility networks.

Based on these findings, several policy implications emerge. Enhancing direct flight networks is crucial, particularly for distant regions where accessibility is a major barrier. Expanding halal-friendly services, prayer facilities, and multilingual support can address cultural constraints that currently deter visitors from Muslim-majority countries. For tourists, food experience programs—such as hands-on sushi or Japanese cuisine workshops—can leverage the strong positive influence of staple food familiarity. For both groups, strengthening Japan’s unique cultural experiences and hospitality may offset the deterrent effects of unfavorable exchange rates.

At the same time, rapid growth in inbound travel raises risks of overtourism, including pressure on infrastructure, environmental burden, and the erosion of local identity. Future tourism strategies must therefore balance visitor growth with sustainable management through measures such as capacity limits, dispersion policies, and community-based tourism planning.

Finally, this study faces limitations, particularly missing data for some countries and years. Future research could incorporate more complete datasets, integrate qualitative evidence, and explore how Japan can achieve sustainable inbound expansion while mitigating the negative effects of overtourism.

Stable access to energy forms the foundation that determines the quality of people’s lives and is indispensable for sustainable development in both developed and developing countries. Japan has long supported the expansion of electricity access and the introduction of renewable energy in Sub-Saharan Africa (SSA), including Kenya. Building on Japan’s experience in this field, this study aims to comprehensively examine how expanding electricity access in Kenya contributes to multidimensional poverty reduction through theoretical, econometric, and fieldwork analyses, and to propose a direction for inclusive and sustainable electrification policy.

In the theoretical analysis, an extended Solow growth model incorporating electricity capital is constructed. The model formalizes a two-stage mechanism: (1) the policy-induced accumulation of electricity capital raises electricity access rates, and (2) the accumulation of electricity capital enhances productivity and human capital, thereby improving broader measures of income. This framework serves as a foundation for testing indirect policy effects in the empirical analysis.

In the macro-econometric analysis, using WDI panel data from 48 SSA countries (2000–2023) and applying fixed effects and distributed lag models, the results suggest that while an increase in electricity access significantly improves household expenditure and educational outcomes in the medium term, short-term investment burdens may temporarily suppress these effects.

The micro-econometric analysis employs KIHBS 2005/06 and 2015/16, as well as KCHS 2020–2022, estimating the Average Treatment Effect on the Treated (ATT) using Propensity Score Matching (PSM) and Staggered Difference-in-Differences (DID). Electrified households exhibit approximately 13% higher per-adult expenditure and a 7-percentage-point higher primary school enrollment rate compared to unelectrified households. RIF-DID analysis shows stronger effects among lower- and middle-income groups, indicating a potential contribution to reducing inequality. In rural areas, improvements in lighting and communication appear to shorten household chore time, secure study time, and expand employment opportunities, particularly benefiting women and children.

Field surveys conducted in August 2025 in the Kinangop area of Nyandarua County revealed that electrified households experienced an average increase of 1.8 hours in nighttime study time and greater access to information via mobile phones and radios. In contrast, unelectrified households faced health hazards and financial instability from reliance on kerosene and candles, as well as frequent blackouts and high electricity tariffs. These findings substantiate the two-stage mechanism proposed in the theoretical model and complement the econometric results.

For policy evaluation, two key challenges identified in the analysis—(A) issues of quality and pricing in the rural solar panel market, and (B) countermeasures against transmission losses and peak-hour outages—are addressed through two policy instruments: (I) the Risk Mitigation Facility (RMF) to encourage private investment, and (II) the Time-of-Use (TOU) tariff system for demand-side management. Using fixed-effect and synthetic control models, the analysis estimates that RMF introduction could increase rural electricity access by up to 24 percentage points, while TOU adoption could reduce transmission and distribution losses by approximately 5.5 points. Robustness checks using placebo tests and MSPE ratios confirm these findings, considering statistical significance and pre-treatment fit.

Overall, expanding electricity access raises rural living standards through improvements in income, education, and health, thereby fostering genuine social inclusion. This study proposes a two-stage policy approach: in Stage 1, the RMF policy mitigates initial investment risks to “expand access,” while in Stage 2, the TOU system establishes a mechanism to “use energy efficiently.” Japan’s cooperation in geothermal power, transmission and distribution, smart meters, and grid digitalization should be re-evaluated as international collaboration supporting these two stages.

The originality of this study lies in three aspects:

 (i) estimating the causal effects of electricity access at both macro (48 SSA countries) and micro (Kenyan households) levels, and identifying which income groups benefit most;

 (ii) incorporating institutional factors such as governance quality, renewable energy policies, and tariff subsidy systems to identify the policy environments that shape electrification effects; and

 (iii) integrating fieldwork findings to complement quantitative analyses with on-the-ground realities.

Through this comprehensive approach, the study provides empirical insights that contribute to Kenya’s inclusive and sustainable development, as well as to Japan’s future international cooperation practices.

Emerging economies have increasingly sought to upgrade their position within global value chains, yet many continue to face structural constraints such as low labor productivity and limited domestic technological capabilities. To strengthen competitiveness and promote domestic manufacturing, governments have revisited traditional industrial policy tools—including tariff protection and targeted subsidies—rooted in the logic of infant-industry protection. While such measures historically played an important role in the industrialization of advanced and East Asian economies, it remains unclear whether similar strategies remain effective under today’s far more integrated and competitive global trade environment. India offers a timely and policy-relevant case. As part of the “Make in India” initiative launched in 2014, the government identified the Electronics System Design and Manufacturing (ESDM) sector as a strategic priority. Between December 2017 and February 2018, India implemented a series of tariff hikes on fully assembled mobile phones and key components, raising mobile phone tariffs to 20% and increasing tariffs on camera modules, chargers, and other parts. These interventions aimed to reduce dependence on imported smartphones, encourage local assembly, attract foreign investment, and expand domestic value addition. In the years following the tariff increases, India’s mobile phone export volume rose sharply, suggesting a potential improvement in industry competitiveness. Against this backdrop, this study aims to examine whether the smartphone-related tariff measures have effectively promoted the development of India’s ESDM sector. Employing a Difference-in-Differences (DID) framework and constructing a detailed industry- and region-level panel dataset, the analysis evaluates how the tariff policy influenced quantifiable outcomes such as production inputs, output growth, value-added expansion, and improvements in labor productivity.

The analysis draws on two primary data sources. The main dataset is the establishment-level manufacturing data from the Annual Survey of Industries (ASI), India’s official micro-census of registered factories. This study constructs a panel for 2015–2023 by merging block-level files and aggregating observations to the state × industry (4-digit NIC code) × year level using a 2-digit NIC code to identify ESDM industries. The ASI provides detailed information on output, inputs, employment, and wages, enabling the construction of key outcome variables such as gross value added (GVA), labor productivity, and export value. GVA is calculated in two forms: (i) the standard ASI-based measure using ex-factory output values, and (ii) an alternative measure using sales value to capture more accurately. Both measures are used to construct labor productivity indicators. Observations with negative GVA values are flagged, and variables with excessive missingness are filtered before aggregation. Additionally, the study incorporates Night-Time Lights (NTL) data from the VIIRS/DNB satellite series as an external proxy for regional economic activity. Monthly radiance readings are filtered and aggregated into annual averages, then spatially matched to India’s sub-district boundaries to build a sub-district × year panel capturing changes in local economic intensity. The NTL dataset complements the ASI-based indicators by providing a broader regional measure of economic activity in areas with an ESDM cluster.

The empirical analysis consists of two complementary DID approaches. The first examines the impact of tariff policy changes on manufacturing outcomes using the ASI panel aggregated to the state × industry × year level. ESDM industries are compared with other manufacturing industries over time. The measurement includes state, industry, and year fixed effects to account for time-invariant characteristics and common macroeconomic shocks. Standard errors are clustered at the industry level, and alternative clustering choices are incorporated as robustness checks. The post-policy period is defined as 2018 onwards. The second DID uses the sub-district × year NTL panel to assess regional changes in economic activity following the same policy shift. Treatment status is defined using the Indian government’s official designation of ESDM clusters: all sub-districts located within districts recognized as ESDM clusters are coded as treated, while sub-districts in non-designated districts serve as controls. The post period is likewise defined from 2018 onwards. This regional DID captures broader spatial responses associated with government-designated ESDM clusters and provides an external robustness check that complements the analysis based on ASI data.

The ASI-based results show that ESDM industries expanded markedly after 2018. GVA and man-days worked both increase significantly, indicating growth in overall production scale and labor demand. Wages also rise, suggesting improved worker compensation in ESDM industries. Sale value, used as an alternative output measure, shows a strong positive response, consistent with broad sectoral expansion. In contrast, labor productivity does not display comparable improvements. Imported input use rises significantly, indicating continued reliance on foreign components despite increased domestic activity, and export activity shows a slight decline. The NTL-based regional analysis provides complementary evidence. Sub-districts located within government-designated ESDM cluster districts exhibit a clearer post-2018 rise in nighttime luminosity compared with non-cluster areas, indicating stronger local economic activity associated with these clusters. Together, the results offer an assessment of tariff-based industrial policy, highlighting both its capacity to stimulate sectoral growth and its limitations in fostering productivity gains within globally fragmented value chains.

The Great East Japan Earthquake of 2011 led to one of the largest disruptions to modern Japanese logistics. It severely damaged coastal infrastructure and halted regional production and transportation networks. Among all impacted regions, the Tohoku region was the most severely affected, with seaports along the coastline suffering extensive damage due to flooding. Understanding how exports shifted and responded to this sudden drop in port capacity is crucial for the characterization of Japan’s economic resilience, and most importantly, for informing preparedness for future anticipated disasters. Prior literature, by Hamano and Vermeulen (2017), suggests that firms reroute export shipments to nearby undamaged ports when their usual ports are no longer capable of supporting shipments. What’s more, Tomita et al. (2013) highlighted breakwater failures that led to port failures. Deriving from these two literatures, we established two hypotheses: (1) export at directly impacted and damaged ports should experience a sharp and instant decline starting from March 2011 until recovery, and (2) looking into nearby ports, we should observe a substitution effect, which, in other words, is a measurable increase in exports since exports are now rerouted. 

To estimate our hypotheses, we extracted export statistics from the Japan Ministry of Finance and constructed a port-level monthly dataset. Moreover, using measurements from the Geospatial Information Authority of Japan, based on data continuity, economic relevance and significance, and geographical proximity, we carefully selected 20 major seaports. We then manually categorized them into three groups: damaged ports, which experienced the most destruction due to tsunami inundation and were incapable of holding exports; near ports, which are undamaged ports located close enough to the damaged ports that can be used as substitute ports; and far ports, which are unaffected ports that acted as the control group for our model.

Using monthly export data surrounding the disaster, the analysis compares outcomes across ports with differing levels of physical damage. Ports were grouped into categories based on the scale of inundation, allowing the study to identify whether the severity of damage influenced the magnitude of export loss. While Hamano examines firm-level outcomes and classifies treatment by magnitudes, this paper analyzes port-level export values and defines treatment using a damage-based near-far categorization.

To identify causal effects, the analysis defines tsunami-affected ports as the treatment group and unaffected ports as the control group, examining their performance before and after March 2011. The model adopts a Poisson fixed-effect DID framework, incorporating port and time fixed effects. Furthermore, event study specification was used to estimate treatment effects.

The results indicate a substantial short-term decline in exports for heavily damaged ports immediately after the disaster. This drop is visible in the regression estimates and supported by graphical evidence, which show a sharp negative shock followed by gradual rebound.

Importantly, the findings point to rapid recovery, highlighting a high degree of resilience in Japan’s logistics infrastructure. The economic losses were therefore concentrated in the immediate aftermath of the disaster, with limited long-term effects on trade capacity. Overall, the study demonstrates that although natural disasters impose severe short-run costs, the ability of firms and infrastructure to adapt, reroute, and restore functionality plays a central role in mitigating prolonged trade disruption.

Our DID estimation provides clear results that support our first hypothesis. Damaged ports experienced a significant decline in export value, with our results indicating a 39.7% immediate drop after the earthquake hit. This statistically significant result indicates and highlights the direct physical destruction of port facilities. However, our second hypothesis is not supported by our data and results. Nearby ports, which we expected to be used as substitute ports to absorb diverted shipments, showed an statistically insignificant decrease in export value. This absence of substitution effect not only suggests that the supply chain disruption might be larger than what we expected it to be, but also shows that damaged ports recovered rapidly, eliminating the importance and necessity of rerouting shipments.  

Our findings showed two key insights. First, the port substitution effect did occur, but it did not follow the logic of using the nearest available ports as substitutes. In some cases, far ports showed even stronger substitution effects than nearby ones, suggesting that capacity and stability are prioritized. Second, Japan’s logistics systems prioritize restoring and recovering damaged ports and infrastructure instead of developing new, alternative routes. This is reflected in the extremely fast recovery we observed, in which exports returned to pre-disaster levels within just five months. Given that, we do realize the limitations of our model. Since only monthly export data was used, it masks a likely faster weekly recovery. Future work could explore transport costs and network-based routing, which would be useful in simulating the potential economic impact of anticipated disasters, such as the Nankai Trough Earthquake. 

本研究は、世界的に問題視されている肥満の要因について、グローバル化の影響に注目しながら、パネルデータを用いた実証分析により明らかにする。近年、肥満状態にある人口の割合が増加し続けており、このまま肥満率が上昇していくと、2050年において過体重(BMIが25以上)および肥満(BMIが30以上)である成人の割合は全世界で 50%を超えると推計されている（Hay et al.2025）。肥満は先進国・途上国問わず増加しており、この状態を放置すれば、「地球」の住人は確実に太っていってしまう。肥満の原因は国や地域により異なると考えられるが、個人の食習慣が原因なだけでなく、経済的理由や社会的要因が絡む場合も多い。肥満は個人の心身の健康だけでなく、生産性の低下等を通じて経済損失を生み出し、社会に悪影響を及ぼしてしまう。各国は食品への課税や広告規制などの対策を講じているが、効果は限定的であり、肥満率の上昇を大きく抑制するまでには至っていない。肥満は世界最大の健康問題として世界保健機構も問題視しており、早期対応が求められる喫緊の課題である。

肥満の原因を実証分析した先行研究としては、米国における肥満の増加を分析した Courtemanche et al.（2016）が、所得階層、地域のレストランやフィットネスセンターの増数、喫煙などの影響を分析している。グローバル化の進行による食の多様化も、高カロリー食品の流入を通じて、肥満増加の要因になりうる。例えばBaggio and Chong(2019)は、米国のFTA 締結が米国と締結相手国の肥満を悪化させることを示しており、「自由貿易が先進国・途上国問わず肥満率の増加させる」という“Globesity”(グローバリゼーションによる肥満)の考えを裏付けている。Zhou(2016)は、所得の増加が肥満に与える影響は、経済発展やグローバル化の進展度に依存することを示している。

しかし、これらの先行研究では、分析の対象国が限られており、そのため先進国と非先進国の違いを十分に考慮できていない。そこで本研究では、非先進国を含んだ多数の国のパネルデータを収集し、二方向固定効果モデルによりグローバル化と肥満の関係とその他の肥満増加要因を再検証した。分析の結果、FTA締結により肥満が増加するのは非先進国に特有であり、先進国では有意な関係は見られなかった。一方、先進国は貿易における食品輸入割合が大きいほど肥満が増加するが、FTA締結数と食品輸入割合の交差項の係数が負であることから、FTAを通じて貿易自由化をしているほど、食品輸入による肥満増加が抑制される。これらの結果は、肥満につながりにくい食品ほど、関税等の貿易政策の水準が高く、そのため食品輸入増が肥満につながる構造を生み出していることを示唆している。また非先進国では、グローバル化以外にも教育の普及や都市インフラの不足といった国家的課題が肥満率を左右することが明らかになった。一方、先進国では結婚や食生活といった個人の行動選択が肥満率に影響を与えていることが分かった。

この結果を受け、本研究では非先進国向けの肥満対策として、肥満につながりにくい品目の関税を下げ、逆に肥満につながりやすい品目の関税をあげることにより、グローバル化の拡大と肥満率の拡大の連鎖を止めることを提案する。また、非先進国に対しては都市インフラの整備や教育制度の拡充、先進国向けの対策としては、携帯アプリ等を通じて個人の行動変化を促すことを提言する。太り続ける地球でこれらの政策を行うことで、世界規模の肥満進行の抑制につながるであろう。

本研究は日本の主食であるコメの海外輸出拡大が可能であるかを考えるものである。近年、日本は海外からの旅行者により多くのインバウンド需要が発生している。国内では日本食の需要に対応するため、代表的な日本食である寿司、うどん以外の料理にも注目が集まっている。街中でも見かけるようになった“おにぎり専門店”など日本のコメを活かした料理が、寿司に比べて安価で食することができるために一種の流行にもなった。一方、最近の日本国内のコメの価格上昇は多くの消費者を悩ませている。2024年から2025年にかけて市場販売価格は2倍ほどに上昇したため、石破政権は備蓄米を放出して市場価格の上昇を抑えることに努めた。また、従来の減反政策を転換して増産目標を発表し、コメ価格の高止まりに対応する姿勢を見せていた。しかし、2025年10月に発足した高市政権は、「需要量に応じた生産」を掲げ、政府のコメ政策の方向性が定まらない。

このような状況を受けて、日本のコメの生産体制や価格設定の過程、今後のコメ政策に関心が集まっている。そこで、私たちは日本国内のコメ生産の現状について調べ、特に輸出拡大の可能性について考えることにした。農林水産省が主導している国内のコメ生産を手助けするプロジェクトの一つに、2017年に開始された「コメ海外市場拡大戦略プロジェクト」がある。これは日本米や米加工食品の輸出量を増加させることを中心におき、生産者支援や輸出企業に対する情報提供を行うものだ。農林水産省の統計によれば2024年度のコメ輸出量は4.5万トンを記録し、直近五年間で2.6倍に増加している。現在の政権によるコメ輸出の姿勢に大きな変化がなければ、これからも日本はコメの輸出促進に取り組み、海外市場でも取扱量が増えていくかもしれない。しかし、現状の日本のコメ農家は生産者の高齢化や水田用地の統合、生産効率化技術の普及の遅れが生じている。また、国内のコメに対する需要が落ちて、パンや他の食品に需要がシフトしつつあり、政府がコメ増産に舵を切るのかも不透明である。このような問題意識から、私たちは、日本がコメの輸出を拡大し、海外市場におけるシェア拡大を望めるのかを検証する。以下に本研究の概要を述べる。

本研究は農林水産省やWTO、税関によって公開されている統計データに基づき、どのような要因で日本のコメの輸出が変化しているのかを分析している。データは2016年から2024年までのものを取り扱う。税関の品目別表をもとに、コメとコメ加工品について、品目別・相手国別に日本からの輸出額の年次データを整理した。これらのデータを用いて、日本のコメの輸出の決定要因を統計ソフトを用いて回帰分析を行った。回帰分析においてはグラビティモデルを利用し、また説明変数として輸出相手国の農業関税の割合や在外邦人数、相手国の米の消費量などを用いた。現状の分析結果として、相手国の農業関税の低下や在外邦人数の増加が日本の米の輸出を増加させるということが分かった。現在のこの結果を踏まえて、今後はより説明変数を追加し、様々な観点からの輸出増加の要因を探っていく。また日本のコメと外国で生産されるコメの価格の差である内外価格差も示して、日本のコメが海外のコメとどれほど高いかを観察していく。

コメ不足を解消し価格高騰を抑えるには、コメを増産すべきではあるが、中長期的には日本国内のコメ需要は減少していくとの見方もあり、安易な増産は将来のコメ余剰と価格暴落を招く恐れもある。国内需要が減っていく中でも国内の生産基盤を維持していくために輸出拡大が必要だろう。農林水産省も、輸出は生産者の所得を高め、国内市場の需給調整にも資すると考えているようである。また長期的には食料自給率の改善にもつながる重要な視点だ。日本の米の生産や輸出が安定することが今後望まれる。

私たちは貿易実務の違いが実際の貿易に与える影響について研究を行った。まず背景として、授業で貿易について学ぶ中で、貿易には関心を得てから実行に移すまでの貿易の流れがあり、その流れの中にある業務（以下貿易実務と呼ぶ）には国ごとに違いがあることを知った。これらのことを知り私たちは、「各国の貿易実務の違いは貿易に影響しているのか」という疑問を抱いた。

本研究は、従来の研究では捨象されがちであった「貿易実務」という実態的なプロセスに焦点を当て、二国間の実務手順や慣習の違いが、実際の貿易取引量（輸入額）にどのような影響を与えているのかを解明することを目的とした。特に、金銭的な支出ではなく、手続きの煩雑さや予見可能性の欠如といった「実務的摩擦」や業務のすり合わせによる担当者への「負担」や「ストレス」が、貿易の阻害要因となる可能性に着目した。我々が立てた仮説は、「貿易実務が似ている国同士であれば、業務手順の『当たり前』が通用するため、実務的摩擦やストレスが減少し、その結果として貿易額が増加する」というものである。

検証にあたり、各国の貿易実務の特性を定量化する独自の指標「TOI（Trade Operation Index）」を作成した。TOIは優劣を格付けするものではなく、その国の実務が持つ特性を示すものである。具体的には、取引の基盤となる安心感を捉える3つの要素（国の安全度、貿易保険、契約の確実性）と、取引を実際に進める実行力を捉える3つの要素（物流の効率性、決断の速度、リスク回避度）の計６つの要素から構成される。本研究では、このTOI単体の数値ではなく、貿易を行う二国間のTOIの差（TOIgap）を「貿易実務の類似性」とした。というものである。これに基づき、2023年のデータを用いて、グラビティモデルおよび国ダミーを使ったモデルの２つを用いて回帰分析を行った。

結果として、TOIgapの係数はいずれのモデルにおいても正で統計的に有意となり、「貿易実務の差が大きい国同士ほど二国間貿易額が大きい」という一見意外な関係が示された。また、輸出国GDPとTOIgapの交差項は負で有意となり、相手国の経済規模が大きい場合には、TOIgapが小さい、すなわち貿易実務が似ているほど貿易拡大効果が強まることも明らかになった。これらの結果から、TOIの値が低い国ほど、自国より実務環境が整った高TOIの国を相手として選好し、その結果として「TOIの差が大きい取引」が増えている可能性が示唆される一方で、実務が類似している国同士では、特に経済規模の大きな国との間でGDP成長を通じた貿易拡大が起こりやすいと解釈できる。

以上の分析から、貿易実務の違いが貿易に与える影響は「一律の負担」でも「一律の利益」でもなく、輸出国の経済規模によって役割が大きく変わることが明らかになった。 中小規模国にとっては、相手国との実務的な「違い」が市場機会の源泉となる一方、大規模国にとっては、実務の「類似性」こそがGDPのポテンシャルを最大効率で貿易額に変換するための鍵となることがわかった。

本研究は、これまでデータとして捉えにくく、重視されていなかった「貿易実務」をTOIとして数量化し、その差異が国際貿易に統計的に有意な影響を持つことを示した。とりわけ、実務環境の整備水準が異なる国同士でむしろ貿易が活発化しうるという、従来の距離や関税だけでは捉えきれない関係性を明らかにした点に独自性がある。この知見は、企業が海外市場へ参入する際、関税や物理的距離に加え、「実務環境の質的差異」を新たな判断基準として活用しうることを示唆している。今後、貿易実務に関するさらなる調査が進めば、新たな非関税障壁やリスク要因の発見も可能となり、本研究は実務・研究双方に発展の余地を残す重要な一歩であると言える。

This paper examines how the growing share of renewable energy sources (RES) in Poland’s electricity generation affects short-run movements in wholesale electricity prices. The motivation comes from Poland’s ongoing energy transition. Even though coal still supplied close to 70% of generation as recently as 2022, the structure of the system has changed in recent years. The turning point came in June 2025, when renewable production exceeded coal generation for the first time in history. This marked a clear shift in the direction of Poland’s energy mix and raised an important question: how does an increasing share of RES translate into short-run price dynamics in the wholesale market?

The data used in the study are as follows. Wholesale day-ahead electricity prices were taken from the Polish Power Exchange (TGE). Carbon prices (EU Energy Trading System), natural gas prices, and electricity load came from ARE (Agencja Rynku Energii) and the European Energy Exchange. All of these variables are available at a daily frequency. The only exception is the RES share, which is available only monthly. To make the dataset consistent and remove non-stationarity in the time series, all variables were converted to monthly differences. The differenced time series were then tested using the ADF test, confirming that all the differenced data were stationary. 

The main regression model investigates how changes in the renewable share relate to monthly changes in wholesale electricity prices after controlling for the variables noted above. The results show a clear and statistically strong effect. This indicates that if the RES share increases by one percentage point compared to the previous month, the monthly change in wholesale prices becomes about 7.5 PLN/MWh lower. This suggests that increasing renewable generation slows price growth from one month to the next. It also reflects the merit-order effect: when renewables grow relative to fossil-fuel generation, they reduce the marginal cost of supplying electricity, and this shows up directly in the month-to-month price movements.

Both EU ETS carbon prices and natural gas prices have substantial and statistically significant positive effects on changes in wholesale electricity prices. The carbon-price coefficient of roughly 6.48 implies that an increase in EU ETS prices raises monthly price changes by around 6.5 PLN/MWh. These results highlight that, despite the rise of renewables, Poland’s price formation is still shaped heavily by fossil-fuel costs. Coal and gas remain marginal generators, so changes in their operating costs show up quickly in wholesale prices.

Electricity load is not statistically significant in the differenced model. Monthly differences smooth out most of the variability that usually makes load a key driver of hourly or daily price movements. In a month-to-month context, fuel and carbon-market shocks dominate.

Two extended models introduce interaction terms to test whether the effect of renewables depends on gas prices or demand. The interaction between changes in RES share and changes in gas prices is negative, suggesting that renewables might slightly reduce the transmission of gas-price shocks into wholesale prices, but the effect is not statistically significant. The interaction between RES changes and load changes also shows no significant effect. These results suggest that the influence of renewables on short-run price movements is relatively stable across different market conditions.

The limitations of the study are as follows: because the original time series data is non-stationary, differencing was required, which led to the inability to correctly explain the long-term effects of the prices and allowed focusing only on the short-term effects. Developing another model to fully understand the long-term trends will be required to draw a broader conclusion. Furthermore, this study focuses solely on wholesale market dynamics, and the effects on consumer electricity prices may differ. The next step in the research would be to examine whether the short-run price-reducing impact of renewables translates into lower costs for households and to what extent consumers benefit from the ongoing transition in the power sector.

In summary, the results show that a growing share of renewable energy in Poland contributes to a reduction in short-term movements in wholesale electricity prices. Alongside the results on EU ETS prices, the research suggests that the EU Emissions Trading System is functioning as intended, creating meaningful incentives for Poland to accelerate its energy transition. 

本研究では近年の日本企業の海外進出トレンドを背景に、企業の国際化が生産性に与える影響を実証的に分析した。企業の国際化は、外需の獲得や生産拠点の分散、知識・技術スピルオーバーの享受などを通じて企業業績や生産性に正の影響を及ぼす可能性がある一方で、海外展開に伴う固定費用や調整コストなどにより、必ずしも短期的に生産性向上に結びつくとは限らない。こうした背景から、企業の国際化と生産性の関係については、単純な線形関係ではなく、国際化の水準や段階に応じた非線形的な関係が存在する可能性が示唆される。この関係性を検証することを目的に本研究では分析対象を海外事業を行う日本企業とし、製造業と非製造業に区分した上で、2020年から2024年までのパネルデータを基に重回帰分析を実施した。また、企業の生産性を測る指標としてOLS回帰、LP法、ACF法による生産関数の推定に基づき算出された全要素生産性を用い、企業の国際化の程度を測る指標としては海外売上比率を用い分析を行った。分析の結果、製造業においては企業の国際化と生産性の間にはU字型の非線形関係が存在し、一定の水準以下では国際化は生産性に負の影響を与えるが、その水準以上では正の影響に転じるという関係性が示された。また、企業の国際化がその企業の技術革新活動と相互に作用し、生産性の向上に寄与することも分析から示唆され、海外展開への取り組みが同水準であっても技術革新活動への取り組みに差異があった場合には生産性の向上量にも差異が生まれるといったことが示された。以上より、本研究は日本国における製造業企業において、国際化と生産性の関係性がU字型の非線形的な関係を持つこと、そして国際化は技術革新活動との相乗効果を媒介し、生産性の向上が達成されるということを実証的に示した。

本稿では、第2次トランプ政権による追加関税（以下、トランプ関税）の下での日本の自動車メーカーの直接投資に焦点を当てる。トランプ関税は、日本の自動車産業に深刻な影響を及ぼす可能性が指摘されている。主要6社の業績に対し約1兆円規模の減収効果が試算される中、日本企業はこのまま米国への輸出を継続するか、あるいは関税回避を目的として海外直接投資（FDI）による現地生産化を進めるべきかという重要な戦略判断に直面している。しかし、トランプ政権の特徴である政策運用の不確実性は、企業がFDIを決定する際のリスクを著しく高め、経営判断をいっそう困難にしている。こうした状況下では、関税率そのものだけでなく、政策の予測可能性の低下が投資判断に与える影響を体系的に検討する必要がある。

本研究の目的は、この政策不確実性が自動車産業のFDIに与える影響を定量的に検証し、現状においてFDIを進めるべきかを明らかにすることである。

本研究は、Boer & Rieth（2024）および孫（2019）を主要な先行研究として位置付ける。Boer & Rieth（2024）は関税政策が経済政策不確実性を増大させ、企業のFDI決定を抑制する可能性を示している。一方、孫（2019）は重力モデルを用いて経済連携協定や経済自由度指数等がFDIに与える影響を分析し、高関税率や地理的距離がFDI障壁となることを確認した。　　　　　　　　　　　　　　　　　　　　　　　　　　　本研究はこれらの知見を踏まえ、FDIに対する各種経済指標の影響に加え、独自に「不確実性ショックダミー」変数（第1次トランプ政権期2017〜2021年を1、それ以外を0）を導入することで、政策不確実性の効果を検証する。仮説として、名目GDPおよび経済自由度指数はFDIに正の影響、1人当たり名目GDP、自動車関税率、輸送距離および不確実性ショックは負の影響を与えると想定する。

分析では、二国間の貿易・投資量を説明する重力モデルを応用し、被説明変数を日本の自動車産業の対外FDIフロー（対数）、説明変数として名目GDP、1人当たり名目GDP、経済自由度指数、自動車関税率、輸送距離、不確実性ショックダミーを用いて推定を行った。　

その結果、「名目GDP」と「経済自由度指数」は1％水準で正に有意であり、市場規模が大きく規制の少ない国ほどFDIが促進されることが示された。「1人当たり名目GDP」は負に有意であり、高賃金国は投資先として選好されにくいことが確認された。また、「不確実性ショック」変数は10％水準で負に有意となり、政策不確実性が企業のFDI行動を抑制する要因となることが示された。

以上の結果より、日本企業はトランプ関税の発動リスクに対し、短期的な政治ショックに過剰反応してFDIを拡大すべきではないとの結論を得た。FDIを進める場合には、市場規模が大きく規制が少なく、かつ賃金水準の低い国を選定することが合理的である。具体的には、米国・メキシコ・カナダ協定（USMCA）に基づく優遇措置を活用でき、米国向け輸出において追加関税回避が期待できるメキシコが有力な候補地となる。本研究は、政策的不確実性の高い局面において、日本の自動車メーカーが経済合理性にもとづき、冷静かつ戦略的に生産拠点配置を検討する必要性を示すものである。