## 19. Freeman and Oostendorp Occupational Database, Comparator Groups

## Source details

**Canonical URL:** [19. Freeman and Oostendorp Occupational Database, Comparator Groups](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp0760.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp0760.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp0760.pdf.json)

---

### Introduction
- Context: examines effect of tariff policy and enforcement on customs duty evasion in India, exploiting the Indian tariff reform of the 1990s.
- Key background facts and trends:
  - Average tariffs: "nearly 100 percent in 1987" → "80 percent in 1991" → "about 25 percent at the turn of the century."
  - Standard deviation of tariffs: "50 percent to 40 percent and to about 10 percent" over same period.
  - Customs seizures: declined "from nearly 70,000 cases in 1990 to about 45,000 in 2004."
  - Evasion: "evasion hovers around 120 percent for the late 1980s and early 1990s" → "about 85 percent in 2002-03."
- Three stated contributions:
  - Refine identification of tariff effects on evasion using variation across products and over time.
  - Show how enforcement-related characteristics affect the evasion elasticity.
  - Illustrate methodology to quantify institutional (customs enforcement) quality over time and compare India with China.

### Defining evasion
- Four measures of evasion:
  - EvV (value-based, matched observations): EvV = log(1+XV) − log(1+MV) (equation (1) in source).
  - EvV (extreme smuggling measure): EvV2 = log(1+XV) − log(1+MV) with unmatched imports coded as zero (equation (2) in source).
  - Two analogous quantity-based measures: evq and evq2.
- Sample size notes:
  - Extreme smuggling specification increases sample by "over 100,000 observations" relative to matched-only.
  - Fisman and Wei (2004) sample "at most about 1700 observations" vs. authors' "between 222,000 and 320,000 observations."

### Data
- Main sources:
  - WITS (UN COMTRADE) partner-reported exports and Indian-authority imports, annual 1987-2003, HS 6-digit (~5000 products).
  - Tariff data compiled in Topalova (2004).
  - Additional: ADCs (selected years), port shares from Tips Software Services, salaries and computers from Ministry of Finance, Govt of India.
- Partner coverage and match rates:
  - Top 40 trading partners accounting for "about 92 percent of total trade."
  - Weighted average match rate = "65 percent."
- Final sample sizes:
  - Extreme smuggling sample "exceeds 325,000 observations."
  - Alternative specification "about 222,000 observations."

### Empirical strategy
- Core estimating equation:
  - Evptc = β*Tpt + product/year/country fixed effects + error (β is semi-elasticity of evasion w.r.t tariffs; see equation (3)).
- Identification:
  - Relies on within-product (HS 6-digit) over-time variation.
  - Standard errors clustered at the 6-digit product level.
- Enforcement identification:
  - Interact tariffs with enforcement proxies E_x to estimate how enforcement affects the evasion elasticity (see equation (4)).
  - Time dummies and product characteristics used to interpret changes in elasticity as changes in enforcement under maintained assumptions.

### Core empirical results
- Main elasticity estimates and robustness:
  - Core specification (Column 7 Table 2): a one percentage point increase in tariffs increases evasion by about "0.12 percent" (value measure).
  - Authors summarize: "a one percentage point increase in tariffs increases evasion by about 0.1 percent."
  - Alternate core estimate used in decomposition: "0.081".
  - Comparison to Fisman and Wei (2004): their estimate is "about one-thirtieth" larger than the India core estimate; product-sample composition explains substantial part of difference.
- Validation of extreme-smuggling assumption:
  - Panel E Table 2: tariff coefficient positive and significant where dependent is a dummy for exports with no corresponding import.
  - Magnitude (Column 7): "a ten percentage point increase in tariffs is associated with about 0.24 percentage points higher probability that there is no corresponding import."
- Misclassification results (Table 3):
  - Holding own tariff constant, a one percentage point decrease in average tariff on similar products leads to "about a 0.26 percent (Column 2) increase in evasion."
  - Inclusion of "tariff-on-similar-products" increases the own-tariff coefficient from "about 0.12 to 0.38" (Column 2).
- Other taxes:
  - Including Additional Duty of Customs (ADCs) and summing Customs+Excise Tariff (Table 4, Panel B) yields a positive and significant coefficient roughly similar to core ("0.102**" Tariff in Panel A; Customs+Excise Tariff "0.102**" in Panel B for evasion).
- Measurement error and product groups (Table 5):
  - For basic, capital and intermediate goods, tariff effect increases by roughly "50 percent" relative to full-sample estimate (from 0.12 to 0.17 in cited specification).
  - Consumer and consumer durables show coefficient "not statistically different from zero."
- Aggregate decomposition:
  - Average evasion declined by "0.06 (or 6 percentage points) between 1988 and 2001."
  - Average tariffs declined by "0.66 (or 66 percentage points)."
  - Using elasticity "0.081" from Table 2, change in evasion explained by tariffs = "0.053 (=0.081*0.66)", i.e., "more than 90 percent of the change in evasion."
  - Using second evasion measure, "more than three-quarters" of the change explained by tariff decline.

### Enforcement and determinants of evasion elasticity
- Product characteristics (Table 6):
  - Rauch differentiation and unit-value dispersion:
    - Products with above-median standard deviation in log unit value have much higher evasion elasticities; interaction coefficient "~0.26" (statistically significant) implying effect more than twice core specification.
  - Bulkiness:
    - Interaction Tariff X Bulkiness negative and significant in some specifications (e.g., -2.916**, -4.068*** in Panel A columns), consistent with bulky homogeneous goods being easier to enforce.
- Mode of entry (Table 7):
  - Interaction Tariff*Air coefficient negative in several specifications (negative in "5 out of 6 cases, and significant in three of them"), suggesting lower evasion elasticity for goods routed via airports versus seaports.
  - Interpretation: enforcement (e.g., computerization) more advanced at airports.
- Salaries of customs staff (Table 8 and discussion):
  - Real monthly salaries increased "80-100 percent in 1997" for customs inspectors (Fifth Pay Commission implementation).
  - Interaction of tariffs with relative wages of inspectors or commissioners generally "not statistically different from zero."
  - Stylized magnitudes: "average value of customs transactions handled by the typical customs officer in India is about Rs. 29 million per month"; monthly salary for a customs inspector ~ "Rs. 9000 per month" (2003 data).
  - Authors note possible reasons for null wage effect: low relative salaries versus transaction values; insufficient magnitude of wage change; data/identification limits.

### Enforcement quality over time
- Period interactions (Tables 9a and 9b, Figure 3):
  - No evidence of statistically significant improvement in the evasion elasticity over the 1990s; some suggestive evidence of deterioration in own-tariff elasticity.
  - Elasticity with respect to tariffs on similar products increases sharply in later periods (statistically significant), interpreted as deterioration in enforcement ability to prevent misclassification.
- Collection efficiency (Pritchett and Sethi / Zee measure):
  - Defined: ratio of average duty collection rate (effective tariff = collected import duties / value of imports) to average statutory rate.
  - For India 1990-2001 (Table 10): collection efficiency "rises sharply in the early part of the 1990s and then declines in the late 1990s," with 2001 collection efficiency "lower than at the start of the reform process."
  - Decline since 1997 despite declining tariffs is "consistent with a decline in enforcement quality."
- Institutional context and reforms:
  - Customs Act of 1962: penalties and prison terms described; "fines and penalties ... have not changed significantly over time."
  - Computerization (ICES) initiated in 1992; implementation took "four years to complete the project and another 5 years to bring all the major ports under this scheme"; as of March 2001 "only half the customs revenues were covered by the ICES."
  - Valuation reforms: move to transaction-value based valuation in 1988; later changes (1995, 2001) "gave greater rather than less discretion to customs officials."
  - Real wages: inspectors and commissioners grew on average "about 3 percent and 5 percent, respectively between 1990 and 2001."
- Perception-based measures:
  - ICRG bureaucratic quality, KKZ government effectiveness and KKZ control of corruption series show "broadly a picture of institutional stagnation," consistent with evasion-elasticity-based findings.

### India versus China: reconciling evasion-elasticity differences
- Baseline differences and decomposition (Table 11):
  - Replicating Fisman-Wei (FW) for China: Tariff coefficient ≈ "2.637***" (close to FW).
  - Using India cross-section 1998 (no partner dimension): coefficients ≈ "0.913" and "0.51" (depending on evasion definition).
  - Restricting Indian sample to FW commodity list raises coefficients (e.g., to "1.189**" in one specification).
  - FW sample biased toward differentiated goods which have higher evasion elasticity.
  - After accounting for sample composition and methodology, India’s evasion elasticity remains "less than half of China’s," implying India’s customs enforcement may have been "potentially twice as effective as that of China" around 1998.
- Cross-check:
  - Collection efficiency ratio for 1998 is "five times higher for India than China" (Table 10), qualitatively consistent.

### Concluding findings and policy implications
- Main empirical findings:
  - Tariffs have a statistically significant but small semi-elasticity effect on evasion: "a one percentage point increase in tariffs increases evasion by about 0.1 percent."
  - Product characteristics matter: differentiated goods and goods with high variance of unit price have substantially higher evasion elasticities (interaction coefficient "~0.26").
  - Mode of entry matters: goods entering via air have lower evasion elasticity than via seaports.
  - Aggregate decline in average evasion (1988-2001) largely due to tariff decline: authors attribute "nearly 90 percent of the decline in average evasion" to the "66 percentage point reduction in average tariffs."
  - No evidence that evasion elasticity improved over the 1990s; enforcement proxy suggests stagnation or possible deterioration.
  - Comparative assessment: India’s customs enforcement potentially twice as effective as China’s around 1998, but gap narrowing over time.
- Policy implications:
  - Tariff liberalization substantially reduces evasion via rent reduction—trade policy can have major anti-rent-seeking effects.
  - Enforcement improvements that reduce discretion and raise detectability (e.g., broader and faster computerization, reducing valuation discretion) could lower evasion elasticities, especially for differentiated goods.
  - Wage increases for customs officers would need to be large relative to the value of transactions handled to materially alter corruption incentives (illustrative numbers: typical customs officer handles ~ "Rs. 29 million per month"; inspector salary ~ "Rs. 9000 per month" in 2003).
  - Monitor and reduce within-4-digit tariff inconsistencies that generate misclassification incentives; address tariffs on similar products to limit misclassification-driven evasion.

*Source: _wp0760 - 19. Freeman and Oostendorp Occupational Database, Comparator Groups ...........................37*

### References..............................................................................................................

### _wp0760 - References

### Figures
- 1. Evolution of Tariffs in India ..................................................................................................5
- 2. Customs Seizures and Evasion Over Time, 1988-2004.........................................................5
- 3. Elasticity of Evasion wrt Tariffs Over Time, 1988-2001-95 Percent Confidence Bands ...20
- 4. Alternative Indices of Institutions, India .............................................................................23

### Tables
- 1. Summary Statistics...............................................................................................................26
- 2. Evasion and Tariffs ..............................................................................................................27
- 3. Evasion, Tariffs and Tariffs on Similar Products ................................................................28
- 4. Evasion, Customs and Excise Tariffs ..................................................................................28
- 5. Evasion, Tariffs and Industry Use-Type ..............................................................................29
- 6. Evasion, Tariffs and Differentiated Goods ..........................................................................29
- 7. Evasion and Tariffs: Share of Transactions, Sea vs. Air .....................................................30
- 8. Evasion, Tariffs and Wages of Customs Inspectors and Commissioners............................30
- 9a. Evasion and Tariffs: Period Interactions............................................................................31
- 9b. Evasion and Tariffs, Controlling for Tariffs of Similar Products: Period Interactions .....31
- 10. Average Statutory and Effective Tariff Rates (China and India), 1990-2001 ...................32
- 11. Evasion and Tariffs: China and India ................................................................................32

### Appendix Tables
- 12. Match Rates of Products and Values Across Different Trading Partners..........................33
- 13. Evasion and Tariffs at the Product Level...........................................................................34
- 14. Evasion and Tariffs: Same Set of Products Over Time .....................................................34
- 15. Evasion, Tariff and Squared Tariff ....................................................................................35
- 16. Evasion and Tariffs: Flexible Functional Form .................................................................35
- 17. Evasion and Tariffs-Sample Split by Institutional Quality of Partner ...............................36
- 18. Characteristics of Products Included in Fisman-Wei Sample............................................36

*Source: _wp0760 - References (page numbering and itemization as provided in the source content).*

### 19. Freeman and Oostendorp Occupational Database, Comparator Groups ...........................37

### 19. Freeman and Oostendorp Occupational Database, Comparator Groups

### Introduction
- Context: paper examines the effect of tariff policy and enforcement on customs duty evasion in India, exploiting the Indian tariff reform of the 1990s.
- Key background facts and trends:
  - Average tariffs: "nearly 100 percent in 1987" → "80 percent in 1991" → "about 25 percent at the turn of the century."
  - Standard deviation of tariffs: "50 percent to 40 percent and to about 10 percent" over same period.
  - Customs seizures: declined "from nearly 70,000 cases in 1990 to about 45,000 in 2004."
  - Evasion (example): "evasion hovers around 120 percent for the late 1980s and early 1990s" → "about 85 percent in 2002-03."
- Three stated contributions:
  1. Refine identification of tariff effects on evasion using variation across products and over time.
  2. Show how enforcement-related characteristics affect the evasion elasticity.
  3. Illustrate a methodology to quantify institutional (customs enforcement) quality over time and compare India with China.

### Defining evasion
- Four measures of evasion used:
  - EvV (value-based, matched observations): EvV = log(1+XV) − log(1+MV) (equation (1) in source).
  - EvV (extreme smuggling measure): EvV2 = log(1+XV) − log(1+MV) with unmatched imports coded as zero (equation (2) in source).
  - Two analogous quantity-based measures (evq and evq2).
- Sample size notes:
  - Extreme smuggling specification increases sample by "over 100,000 observations" relative to matched-only.
  - Fisman and Wei (2004) sample "at most about 1700 observations" vs. authors' "between 222,000 and 320,000 observations."

### Data
- Main sources:
  - WITS (UN COMTRADE) for exports (partner-reported) and imports (Indian authorities), annual 1987-2003, HS 6-digit (~5000 products).
  - Tariff data compiled in Topalova (2004).
  - Additional: ADCs (selected years), port shares from Tips Software Services, salaries and computers from Ministry of Finance, Govt of India.
- Partner coverage: top 40 trading partners accounting for "about 92 percent of total trade"; weighted average match rate = "65 percent."
- Final sample sizes:
  - Extreme smuggling sample "exceeds 325,000 observations."
  - Alternative specification "about 222,000 observations."

### Empirical strategy
- Main estimating equation (core):
  - Evptc = β*Tpt + product/year/country fixed effects + error (see equation (3)); β is semi-elasticity of evasion w.r.t tariffs.
- Identification:
  - Relies on within-product (HS 6-digit) over-time variation.
  - Standard errors clustered at the 6-digit product level.
- Enforcement identification strategy:
  - Focus on how enforcement proxies (E_x) affect the evasion elasticity by interacting tariffs with E_x (see equation (4)).
  - Time dummies and product characteristics used to interpret changes in elasticity as changes in enforcement (under maintained assumptions).

### Core empirical results
- Main elasticity estimates and robustness:
  - Core specification (most general, Column 7 Table 2): a one percentage point increase in tariffs increases evasion by about "0.12 percent" (value measure).
  - Authors summarize: "a one percentage point increase in tariffs increases evasion by about 0.1 percent."
  - Alternate core estimate cited elsewhere: "0.081" used to decompose changes in average evasion (see summary below).
  - Comparison to Fisman and Wei (2004): their estimate is "about one-thirtieth" larger than the India core estimate; authors find product-sample composition explains substantial part of difference.
- Validation of extreme-smuggling assumption:
  - Panel E Table 2: tariff coefficient positive and significant where dependent is a dummy for exports with no corresponding import.
  - Magnitude interpretation (Column 7): "a ten percentage point increase in tariffs is associated with about 0.24 percentage points higher probability that there is no corresponding import."
- Misclassification results (Table 3):
  - Holding own tariff constant, a one percentage point decrease in average tariff on similar products leads to "about a 0.26 percent (Column 2) increase in evasion."
  - Inclusion of "tariff-on-similar-products" increases the own-tariff coefficient from "about 0.12 to 0.38" (Column 2).
- Other taxes:
  - Including Additional Duty of Customs (ADCs) and summing Customs+Excise Tariff (Table 4, Panel B) yields a positive and significant coefficient roughly similar to core ("0.102**" Tariff in Panel A; Customs+Excise Tariff "0.102**" in Panel B for evasion).
- Measurement error and product groups (Table 5):
  - For basic, capital and intermediate goods, tariff effect increases by roughly "50 percent" relative to full-sample estimate (from 0.12 to 0.17 in cited specification).
  - Consumer and consumer durables show coefficient "not statistically different from zero" (suggesting mismeasurement due to QRs earlier in sample).
- Aggregate decomposition:
  - Average evasion declined by "0.06 (or 6 percentage points) between 1988 and 2001."
  - Average tariffs declined by "0.66 (or 66 percentage points)."
  - Using elasticity "0.081" from Table 2, change in evasion explained by tariffs = "0.053 (=0.081*0.66)", i.e., "more than 90 percent of the change in evasion."
  - Using second evasion measure, "more than three-quarters" of the change explained by tariff decline.

### Enforcement and determinants of evasion elasticity
- Product characteristics (ease of enforcement) — Table 6 highlights:
  - Rauch differentiation and unit-value dispersion:
    - Products with above-median standard deviation in log unit value have much higher evasion elasticities; e.g., interaction coefficient "~0.26" (statistically significant) implying effect more than twice core specification for such products.
  - Bulkiness:
    - Interaction Tariff X Bulkiness negative and significant in some specifications (e.g., -2.916**, -4.068*** in Panel A columns), consistent with bulky homogeneous goods being easier to enforce.
- Mode of entry (Table 7):
  - Interaction Tariff*Air coefficient negative in several specifications (negative in "5 out of 6 cases, and significant in three of them"), suggesting lower evasion elasticity for goods routed via airports vs seaports.
  - Interpretation: enforcement (e.g., computerization) more advanced at airports.
- Salaries of customs staff (Table 8 and discussion):
  - Real monthly salaries increased "80-100 percent in 1997" for customs inspectors (Fifth Pay Commission implementation).
  - Interaction of tariffs with relative wages of inspectors or commissioners generally "not statistically different from zero."
  - Stylized magnitude: "average value of customs transactions handled by the typical customs officer in India is about Rs. 29 million per month"; monthly salary for a customs inspector ~ "Rs. 9000 per month" (2003 data) — implication: marginal wage increases may be small relative to corruption rents.
  - Authors note possible reasons for null wage effect: low relative salaries versus transaction values; insufficient magnitude of wage change; data/identification limits.

### Enforcement quality over time
- Period interactions (Tables 9a and 9b, Figure 3):
  - No evidence of statistically significant improvement in the evasion elasticity over 1990s; some suggestive evidence of deterioration in own-tariff elasticity.
  - Elasticity with respect to tariffs on similar products increases sharply in later periods (statistically significant), interpreted as deterioration in enforcement ability to prevent misclassification.
- Collection efficiency (Pritchett and Sethi / Zee measure):
  - Defined as ratio: average duty collection rate (effective tariff = collected import duties / value of imports) to average statutory rate.
  - For India 1990-2001 (Table 10): collection efficiency "rises sharply in the early part of the 1990s and then declines in the late 1990s," with 2001 collection efficiency "lower than at the start of the reform process." Decline since 1997 despite declining tariffs is "consistent with a decline in enforcement quality."
- Institutional context and reforms:
  - Customs Act of 1962: penalties and prison terms described; "fines and penalties ... have not changed significantly over time."
  - Computerization (ICES) initiated in 1992; implementation took "four years to complete the project and another 5 years to bring all the major ports under this scheme"; as of March 2001 "only half the customs revenues were covered by the ICES."
  - Valuation reforms: major move to transaction-value based valuation in 1988; later changes (1995, 2001) "gave greater rather than less discretion to customs officials."
  - Real wages: inspectors and commissioners grew on average "about 3 percent and 5 percent, respectively between 1990 and 2001."
- Comparison with perception-based measures:
  - Figure 4: ICRG bureaucratic quality, KKZ government effectiveness and KKZ control of corruption series show "broadly a picture of institutional stagnation," consistent with evasion-elasticity-based findings.

### India versus China: reconciling evasion-elasticity differences
- Baseline difference: authors' India core evasion elasticity ≈ "0.12" vs Fisman and Wei (2004) China estimate ≈ "nearly thirty-fold" larger in initial comparison; after decomposition authors find much of the gap due to sample choices.
- Reconciliation exercises (Table 11):
  - Replicating Fisman-Wei (FW) for China: Tariff coefficient ≈ "2.637***" (close to FW).
  - Using India cross-section 1998 (no partner dimension): coefficients ≈ "0.913" and "0.51" (depending on evasion definition).
  - Restricting Indian sample to FW commodity list raises coefficients (e.g., to "1.189**" in one specification).
  - Conclusion: FW sample biased toward differentiated goods (which have higher evasion elasticity); after accounting for sample composition and methodology, India’s evasion elasticity remains "less than half of China’s," implying India’s customs enforcement may have been "potentially twice as effective as that of China" around 1998.
- Cross-check: collection efficiency ratio for 1998 is "five times higher for India than China" (Table 10), qualitatively consistent.

### Concluding findings and policy-relevant points
- Main empirical findings:
  - Tariffs have a statistically significant but small semi-elasticity effect on evasion: "a one percentage point increase in tariffs increases evasion by about 0.1 percent."
  - Product characteristics matter: differentiated goods and goods with high variance of unit price have substantially higher evasion elasticities.
  - Mode of entry matters: goods entering via air have lower evasion elasticity than via seaports (consistent with greater computerization at airports).
  - Aggregate decline in average evasion (1988-2001) largely due to tariff decline: authors attribute "nearly 90 percent of the decline in average evasion" to the "66 percentage point reduction in average tariffs."
  - No evidence that evasion elasticity improved over the 1990s; enforcement proxy suggests stagnation or possible deterioration.
  - Comparative assessment: India’s customs enforcement potentially twice as effective as China’s around 1998, but gap narrowing over time due to improvements in China and lack of improvement in India.
- Policy implications (as presented or implied in text):
  - Tariff liberalization substantially reduces evasion via rent reduction—trade policy can have major anti-rent-seeking effects.
  - Enforcement improvements that reduce discretion and raise detectability (e.g., broader and faster computerization, reducing valuation discretion) could lower evasion elasticities, especially for differentiated goods.
  - Wage increases for customs officers would need to be large relative to the value of transactions handled to materially alter corruption incentives (illustrative numbers: typical customs officer handles ~ "Rs. 29 million per month"; inspector salary ~ "Rs. 9000 per month" in 2003).
  - Monitoring misclassification incentives (tariffs on similar products) matters: policy should address within-4-digit tariff inconsistencies that generate misclassification incentives.

*Source: _wp0760 - 19. Freeman and Oostendorp Occupational Database, Comparator Groups ...........................37*

### References

### References

### Tax evasion, tax avoidance, and tax administration
- Allingham, Michael G., and Agnar Sandmo, 1972, “Income Tax Evasion: A Theoretical Analysis,” Journal of Public Economics Vol. 1, pp. 323-38.
- Fisman, Raymond and Shang-Jin Wei, 2004, “Tax Rates and Tax Evasion: Evidence from ‘Missing Imports’ in China,” Journal of Political Economy No. 112 Vol. 2, pp. 471-496.
- Kopczuk, Wojciech, 2005, “Tax Bases, Tax Rates and the Elasticity of Reported Income,” Journal of Public Economics 89, pp. 2093-2119.
- Slemrod, Joel, 2003, “The System-Dependent Tax responsiveness of Cigarette Purchases: Evidence from Michigan,” Mimeo, University of Michigan.
- Slemrod, Joel and Shlomo Yitzhaki, 2000, “Tax Avoidance, Evasion, and Administration,” Handbook of Public Economics No. 3, pp 1423-1470.
- Slemrod, Joel and Wojciech Kopczuk, 2002, “The Optimal Elasticity of Taxable Income,” Journal of Public Economics No. 84, pp. 91-112.
- Zee, Howell H., 2005, “A Summary Score of Tax Performance for Comparing Tax Systems Across Countries and Over Time,” mimeo, Fiscal Affairs Department, International Monetary Fund.

### Trade, customs valuation, and tariff evasion
- Bhagwati, Jagdish, 1964, “On the Underinvoicing of Imports,” Bulletin of the Oxford University Institute of Economics and Statistics, 26 (November), pp. 389-397.
- Chaturvedi, Sachin, 2006, “Customs Valuation in India: Identifying Trade Facilitation-related Concerns,” Asia-Pacific Research Training Network on Trade, Working Paper Series, No. 25. (http://www.unescap.org/tid/artnet/pub/wp2506.pdf)
- Javorcik, Beata, S., and Gaia Narciso, 2006, “Differentiated Products and Evasion of Import Tariffs,” mimeo, World Bank, Washington D.C.
- Pritchett, Lant and Geeta Sethi, 1994, “Tariff Rates, Tariff Revenue, and Tariff Reform: Some New Facts,” World Bank Economic Review 8, pp. 1-16.
- Topalova, Petia, 2004, “Trade Liberalization and Firm Productivity: the Case of India,” IMF Working Paper No. 04/28, (Washington: International Monetary Fund).
- Fisman, Raymond and Shang-Jin Wei, 2004, “Tax Rates and Tax Evasion: Evidence from ‘Missing Imports’ in China,” Journal of Political Economy No. 112 Vol. 2, pp. 471-496.

### Governance, institutions, and measurement
- Kaufmann, Daniel, Aart Kraay, and Massimo Mastruzzi, 2006, “Governance Matters V: Aggregate and Individual Governance Indicators for 1996-2005, ” Mimeo, (September).
- ———, and Pablo Zoido-Lobatón, 1999, “Aggregating Governance Indicators,” World Bank Policy Research Working Paper No. 2195, (Washington: World Bank).
- Krueger, Ann O., 1974, “The Political Economy of the Rent-Seeking Society,” American Economic Review, Vol. 64, No. 3.
- Giuliano, Paola, Antonio Spilimbergo, and Giovanni Tonon, 2006, “Genetic, Geographical, and Cultural Distances,” CEPR Discussion Paper No. 5807, (September).
- Subramanian, Arvind, 2006, “The Intriguing Relationship Between Growth and Institutions in India,” Oxford Review of Economic Policy, forthcoming.

### Networks, markets, and production functions; education and incentives
- Rauch, James E., 1999, “Networks Versus Markets in International Trade,” Journal of International Economics No. 48, pp. 7-35.
- Muralidharan, Karthik and Venkatesh Sundararaman, 2006, “Teacher and Non-Teacher Inputs in the Education Production Function: Experimental Evidence from India,” Mimeo, Harvard University.
- Muralidharan, K, and V. Sundararaman, 2006, “Teacher Incentives in Developing Countries: Experimental Evidence from India,” mimeo, Harvard University.
- EPW Research Foundation, 2005, “A Note on India’s Balance of Payments: Concepts, Compilation and Recent Scenario: 1950-51 to 2003-04.” Economic and Political Weekly Research Foundation.

*References — _wp0760 - References*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp0760.pdf_