Conscription and its exemption in 19th Century Japan: Incentivized family head in educational market

Using individual-level data from Keio Gijuku, this paper demonstrates that while 19th-century Japanese conscription exemptions incentivized family heads to increase their educational attendance, this surge in enrollment did not translate into a qualitative improvement in academic performance.

The Gist

Imagine 19th-century Japan as a giant, high-stakes video game. In this game, every young man has a "Game Over" screen waiting for him: Conscription. This means being forced to join the army, which was a dangerous and life-disrupting experience.

However, the game had a "cheat code." If you were a student at a private school, you could pause the game and avoid the army. But the game developers (the Meiji Government) kept changing the rules.

This paper, written by economist Eiji Yamamura, is like a detective story investigating how students reacted when the developers suddenly patched a loophole in the game.

The Plot: The "Family Head" Loophole

Here is the sequence of events the author uncovered:

1. The "Easy Mode" (Before 1883): Being a student was a golden ticket. You could skip the army just by showing up to class. Students didn't have to try very hard; they just needed to stay enrolled.
2. The "Hard Mode" Patch (1884–1888): The government changed the rules. Suddenly, regular students lost their immunity. However, there was one tiny exception: The Family Head (the boss of the family) could still avoid the army.
   • The Reaction: This was a massive shock. Young men realized, "Wait, if I become the boss of my family, I can still skip the army!"
   • The Cheat: Since you could technically become a family head even if your father was still alive (by asking him to step down), many students did exactly that. They manipulated family records to become "Family Heads" just to keep their cheat code active.
   • The Result: The number of "Family Head" students skyrocketed. But here is the twist: These students started studying harder. Why? Because they had worked so hard to become the "Family Head" (a difficult social maneuver requiring family agreement), they were highly motivated to prove they were worth the effort. They wanted to get the best grades to justify their new status and ensure they didn't drop out.
3. The "Back to Normal" Patch (1889): The government realized the loophole was too big. They closed it. Now, no one (not even the Family Head) could avoid the army just by being a student. Instead, all students got a temporary delay until age 26.
   • The Reaction: The "Family Head" students lost their special status. Their motivation to study dropped back to normal levels. The gap between them and regular students disappeared.

The Big Discovery: Quantity vs. Quality

The author found something fascinating about human behavior:

• When the rules were loose: Students treated school like a parking spot. They parked their car there to avoid the army, but they didn't actually drive the car anywhere. They showed up, but they didn't learn much.
• When the rules got tight (1884–1888): Students treated school like a gym. They had to work out hard to keep their membership. The "Family Heads" were the ones hitting the weights the hardest. Their grades went up, and they were less likely to quit.
• The Lesson: Conscription (the threat of the army) made more people go to school (quantity), but it didn't necessarily make them learn better (quality). The only time they actually learned more was when they had to fight for a specific privilege (becoming a Family Head).

The "Tokyo Commoner" Twist

The study also looked at who was studying.

• Samurai (Warrior Class): They were used to old ways and didn't have much money left.
• Commoners in Tokyo: These were the new business class. They saw education as a way to make money in the booming city of Tokyo.
• The Result: When the rules got tough, the Tokyo Commoner Family Heads were the absolute champions. They had the most to gain (money and business success) and the most to lose (the army). They studied the hardest and dropped out the least.

The Analogy: The "Free Lunch" vs. The "Burger Challenge"

Think of the school exemption like a free lunch at a restaurant.

• Phase 1: The restaurant gives everyone a free lunch. People show up, but they just sit there and eat the free food. They don't care about the menu or the quality of the meal.
• Phase 2: The restaurant says, "Free lunch is cancelled! But, if you can prove you are the Manager of the Table, you still get a free meal."
  • Suddenly, everyone scrambles to become the "Manager."
  • Once they are Managers, they take the job seriously. They study the menu, they order the best dishes, and they make sure they don't get kicked out of the restaurant. They are actually learning about food now.
• Phase 3: The restaurant says, "Okay, no more free meals for Managers either. Everyone has to pay, but you can wait until you're older to pay."
  • The "Managers" stop trying so hard. They go back to just sitting there.

Why Does This Matter?

This paper teaches us a valuable lesson about incentives.

If you give people a reward just for showing up (like a free lunch or a student ID that avoids the draft), they will show up, but they won't necessarily work hard. But if you make them fight for that reward (by forcing them to become a Family Head), they will suddenly become very motivated and perform at their best.

The government thought they were building a smarter, more educated nation by forcing people into school to avoid the army. But the study shows that while they built a bigger school, they didn't necessarily build smarter students—unless the students had a personal, high-stakes reason to care.

In short: You can force a horse to the water (school), but you can't make it drink (learn) unless it's thirsty enough. The "Family Head" rule made them thirsty.

Technical Summary

Here is a detailed technical summary of the paper "Conscription and its exemption in 19th Century Japan: Incentivized family head in educational market" by Eiji Yamamura.

1. Research Problem and Context

The paper investigates the causal relationship between conscription policies and educational outcomes, specifically focusing on the quality (academic performance) versus quantity (attendance/duration) of education. While existing literature offers mixed evidence on whether conscription increases or decreases educational attainment, this study addresses a specific gap: the mechanism by which individuals strategically manipulate their status to avoid conscription and how this affects their motivation to learn once enrolled.

Historical Setting:
The study utilizes a unique natural experiment in Meiji-era Japan (1880–1898) following the introduction of universal conscription in 1873. The conscription exemption rules underwent three distinct phases:

1. Loose Period (1873–1883): Broad exemptions existed for students and family heads.
2. Head Exemption Period (1884–1888): Exemptions for private school students were abolished, but family heads retained exemption rights. This created a strong incentive for students to become family heads (via adoption or transfer of status) to avoid military service.
3. No Exemption Period (1889–1898): Family head exemptions were abolished, but private school students regained temporary exemptions until age 26.

2. Methodology and Data

Data Source:
The study constructs a novel, large-scale individual-level panel dataset from Keio Gijuku, Japan's leading private school at the time.

• Sample Size: Approximately 6,700 individuals and 39,694 observations (academic terms).
• Variables:
  • Academic Performance: Measured by standardized ordinal rank within a class ($1 - \frac{\text{Rank}}{\text{Class Size}}$).
  • Dropout Rate: Analyzed as a duration outcome.
  • Treatment Variables: Dummy variables for "Family Head" status, "Head Exemption Period" (1884–1888), and "No Exemption Period" (1889–1898).
  • Controls: Class size, name changes (proxy for evasion attempts), social status (Samurai vs. Commoner), and hometown (Tokyo vs. others).

Econometric Strategy:
The author employs a Difference-in-Differences (DID) approach combined with Duration Analysis to identify causal effects.

1. OLS Regression (Academic Performance):
   • Model: $Performance_{it} = \alpha_0 + \alpha_1(Head_i \times HeadExemption_t) + \dots + \text{Class Fixed Effects} + \epsilon_{it}$
   • The key coefficient of interest is $\alpha_1$, capturing the interaction between being a family head and the specific period where only heads were exempt.
2. Competing-Risk Regression (Dropout Probability):
   • Model: Proportional hazards model where "dropping out" is the failure event and "graduation" is the competing risk.
   • This allows the author to distinguish between students who stayed in school merely to avoid conscription versus those who persisted due to academic motivation.

3. Key Hypotheses

• Hypothesis 1 (Selection/Incentive Effect): During the "Head Exemption Period," family head students would exhibit significantly higher academic performance than non-heads. The logic is that becoming a family head required high transaction costs and family consensus, selecting only "promising" candidates who were also highly motivated to learn to justify the investment.
• Hypothesis 2 (Strategic Attendance): Students would be less likely to drop out if their student status provided exemption from conscription. However, without the specific "Head" incentive, performance might not improve.
• Hypothesis 3 (Heterogeneity): The effect would be strongest for Commoners born in Tokyo, as they faced higher economic returns from modern education (business opportunities) compared to Samurai (who lost stipends) or those outside the economic hub.

4. Key Results

A. Quantitative vs. Qualitative Impact

• Enrollment Surge: The rate of family heads among freshmen skyrocketed from ~1% (pre-1884) to ~13% (1884–1888), confirming that students strategically manipulated their family status to avoid conscription.
• Academic Performance:
  • During the Head Exemption Period, family heads outperformed non-heads significantly (coefficient $\approx 0.118$, or ~11.8% higher rank).
  • Once the privilege was removed (1889+), this performance gap disappeared.
  • Interpretation: Conscription rules increased the quantity of attendance but did not improve the quality of education for the general student body. Only the specific subgroup incentivized to become family heads (who were likely high-ability) showed improved performance.

B. Dropout Rates

• Family heads in the 1884–1888 period had a 60% lower probability of dropping out compared to the pre-period (Hazard Ratio $\approx 0.4$).
• Even after the privilege was removed, family heads maintained lower dropout rates than non-heads, suggesting a persistent "selection effect" where high-ability students remained.

C. Heterogeneous Effects

• Tokyo Commoners: The positive effect on performance and retention was most pronounced for commoners born in Tokyo. This group had the highest expected returns from acquiring modern business skills (accounting, etc.) offered by Keio.
• Samurai: Samurai students showed less sensitivity to these incentives, likely due to lower marginal returns from modern education compared to commoners seeking to establish new business ventures.

5. Contributions and Significance

Theoretical Contribution:
The paper challenges the binary view of conscription as either purely detrimental or beneficial to human capital. It introduces a "Strategic School-Staying" mechanism:

• Conscription can increase school attendance (quantity) by creating an incentive to enroll.
• However, without specific incentives to learn (not just to attend), the quality of human capital accumulation remains low.
• The study demonstrates that institutional loopholes (exemption rules) can distort educational investment, leading to "strategic" enrollment by high-potential individuals who are then motivated to perform, while the general population may only attend to avoid service without improving skills.

Policy Implications:

• Efficiency: Conscription-induced schooling may result in low marginal returns on education if the primary motivation is avoidance rather than skill acquisition.
• Rule Design: Policymakers should be aware that exemption rules can create unintended selection effects. If exemptions are tied to specific statuses (like family headship), they may inadvertently select and motivate a specific high-ability subgroup while leaving the broader population with lower-quality educational outcomes.

Methodological Contribution:
The study provides a rare historical natural experiment using individual-level panel data from a non-Western, early-modernizing economy. It successfully disentangles the effects of attendance incentives from learning incentives by exploiting the specific, short-lived window (1884–1888) where exemption was tied strictly to family head status.

Conclusion

Yamamura concludes that while conscription increased the quantity of educational attendance in Meiji Japan, it did not qualitatively improve academic performance for the average student. The observed improvements were driven by a specific selection mechanism: the high transaction costs of becoming a "family head" filtered for promising individuals who were then highly motivated to learn. This suggests that human capital accumulation driven by conscription avoidance is often inefficient unless the institutional design aligns the incentive to attend with the incentive to learn.