# Acknowledged inventions: constructed illustrations the author has chosen to
# keep in quotation marks for readability (MB rulings, Aug 2026). The quote
# audit reports these as 'acknowledged' rather than NOT FOUND.
We exploit the staggered rollout of the policy across states.
The results are consistent with a causal interpretation
The evidence is consistent with a causal interpretation.
Under our identifying assumptions, the estimates imply a causal effect of 0.14.
These results can be interpreted as causal under our identifying assumptions
This literature has been fruitful, and we take the next logical step.
this policy question is urgent and needs evidence.
We find significant effects of customs modernisation on trade.
Trade barriers impose significant costs on developing economies
We exploit the staggered implementation of customs reforms across 45 countries to identify the causal effect of trade facilitation on bilateral trade flows.
We find that customs modernisation increases bilateral trade by 12-15%. Effects are concentrated in differentiated goods and are larger for country pairs with previously high trade costs.
These results suggest that administrative trade barriers are a significant impediment to trade and that customs reform can be an effective policy tool for trade promotion.
This paper uses a novel dataset of firm-level investment decisions matched to plant-level capital stocks to estimate the effect of uncertainty on investment.
our results have implications for policy
our estimates suggest that a 10% reduction in customs processing time would increase bilateral trade by 12-15%
The rest of this paper is organised as follows. Section 2 develops a simple model of investment under uncertainty.
The coefficient on education may possibly be positive, which could perhaps suggest that education might have some effect on earnings.
The coefficient on education is positive and significant at the 1% level, suggesting that an additional year of education increases earnings by approximately 8%.
Education causes an 8% increase in earnings.
Our IV estimates suggest that an additional year of education increases earnings by approximately 8%, consistent with the OLS estimate of Smith (2020).

# Bulk-acknowledged instructional micro-examples (MB readability ruling, Aug 2026)
Assumption 1 (Perfect Competition): All firms are price-takers.
The key structural parameter is θ, which we estimate using maximum likelihood.
using bilateral trade data for 120 countries over the period 1990-2015
We include education as a control because it may be correlated with both health and income
We use the 2010-2015 waves of the American Community Survey (ACS), accessed through IPUMS
Starting sample: 150,000 observations. After dropping missing education: 148,753. After age restriction: 142,862. After matching to treatment data: 98,441. Final analysis sample: 98,441.
The mean of our dependent variable (test scores) is 0.02, with a standard deviation of 0.98, confirming that the scores are approximately standardised. The treatment variable (exposure to the programme) has a mean of 0.34, indicating that roughly a third of the sample received the treatment.
We use data from the World Bank
The coefficient is economically and statistically significant
The left column is not wrong, but it signals a rather casual engagement with identification. The right column signals methodological awareness. The vocabulary should match the strength of your identification: “We identify a causal relationship
The IV estimates are consistent with attenuation bias in OLS.
We add demographic controls (age, gender, education) to address the concern that the treatment group may differ from the control group on observable characteristics.
To address the concern of pre-existing trends, we include state-specific linear time trends.
Our IV estimate of 0.14 is similar to the RCT estimate of Brown (2020), suggesting our identification successfully addresses the endogeneity that biases OLS downward.
One concern is that our results may be driven by unobserved regional characteristics. To address this, we include region fixed effects. Column 4 of Table 5 shows that the point estimate is essentially unchanged.
A second concern is non-random selection into the treatment group. We address this using propensity score matching. Table 6 reports the matched estimates.
Figure 2 shows the event study estimates. The pre-treatment coefficients are small and statistically insignificant, supporting the parallel trends assumption. After implementation, the treatment effect increases gradually, reaching 0.12 by year 3.
We assume that agents are risk-neutral
In equilibrium, the price equals marginal cost
The sample comprises 45,000 individuals observed annually from 1990 to 2010.
The survey may undercount informal workers.
The treatment increases test scores by 0.15 standard deviations
The coefficient suggests a positive association
The coefficient is statistically significant at the 1% level
This suggests that the mechanism operates through labour supply
The coefficient may be positive and perhaps significant
simple Specific past data collection Past simple
We assume that each firm produces a single good
The survey was conducted in 2011
The policy was implemented in 2003
We find that the treatment effect is positive and significant.
We found that trade barriers reduced bilateral flows by 15%.
We find that trade barriers reduce bilateral flows by 15%.
We collect data from the American Community Survey for the years 2005-2015.
We use data from the American Community Survey for the years 2005-2015.
Smith (2020) showed that returns to education varied across regions.
Smith (2020) shows that returns to education vary across regions.
We have estimated the effect of trade facilitation on bilateral flows.
We estimate the effect of trade facilitation on bilateral flows.
Column 2 reports the IV estimates. The coefficient was 0.14 and is significant at the 5% level.
Column 2 reports the IV estimates. The coefficient is 0.14 and is significant at the 5% level.
The data were collected by the BLS.
Fixed effects are commonly used to address unobserved heterogeneity.
The data were collected by the BLS
The survey was administered in 2011
Capital accumulation is crucial for growth (Solow, 1956).
Unlike Smith (2020), we use an instrumental variables approach.
In contrast to Jones (2018), our analysis accounts for general equilibrium effects.
Our results are consistent with those of Davis (2017).
Our paper relates to several strands of the literature.
(Author1, Year; Author2, Year; Author3, Year)
Table 3 reports the baseline results.
Column 4 of Table 5 presents the IV estimates.
First, we present OLS results. Second, we show IV estimates. Third, we conduct robustness checks.
One concern is reverse causality. However, the lagged specification produces similar results.
Our estimate of 0.15 is larger than Smith’s (2020) OLS estimate of 0.08, consistent with attenuation bias.
The coefficient is positive, consistent with the model’s prediction
Capital accumulation is crucial for business cycles and economic growth. A large literature has studied the determinants of investment, focusing on the roles of uncertainty, financial constraints, and adjustment costs (Author1, Year; Author2, Year; Author3, Year).
However, empirical models of investment for aggregate capital may be plagued by inherent biases. Most existing estimates rely on aggregate data that cannot distinguish between types of capital, and standard identification strategies fail to account for the endogeneity of investment opportunities.
We restrict the sample to individuals aged 25-64.
