## 1. Timeline of e-Invoicing Adoption Waves in Peru

## Source details

**Canonical URL:** [1. Timeline of e-Invoicing Adoption Waves in Peru](https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019231-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2019/wpiea2019231-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2019/wpiea2019231-print-pdf.pdf.json)

---

### Introduction and context
- Digitalization is transforming tax administrations and e-invoicing enables automatic transfer of billing information between firms and the tax authority.
- More than 50 countries have implemented e-invoicing, including ten countries in Latin America and the Caribbean (EY, 2018; Barreix and Zambrano, 2018).
- Peru introduced VAT e-invoicing gradually, with early waves focusing on larger firms and priority sectors (such as mining) and later waves including smaller firms.

### Data, sample, and identification
- Administrative monthly panel dataset from SUNAT covering all small, medium and large private-sector firms operating in Peru between 2010 and 2017.
- Balanced sample panel used in the analysis: 78 thousand firms that were mandated to adopt e-invoicing between 2014 and 2018, which account for over 80 percent of domestic VAT collections in Peru.
- Identification strategy exploits sequential (wave-based) mandatory rollout to estimate causal impacts (intent-to-treat and instrumental variable analyses).

### Main empirical findings — Intent-to-Treat (ITT) (first year after mandate)
- Reported taxable sales increase by 6.57 percent (Treatment indicator (1 year) = 0.0657***).
- Reported taxable purchases increase by 4.53 percent (0.0453***).
- Reported taxable value-added increases by 5.89 percent (0.0589***).
- Reported VAT liabilities increase by 7.22 percent (0.0722***).
- New VAT credits increase by 4.70 percent (0.0470; not statistically significant).
- VAT credit stock decreases by 4.32 percent (-0.0432; not statistically significant).
- Total VAT payments increase by 4.45 percent (0.0445; *p*-value around 0.1 in Panel A, stronger in dynamic results).
- Measured productivity increases by 6.04 percent (0.0604***).
- Interpretation caveat: data cannot distinguish between real increases in production versus increased reporting of previously unreported activity.

### Dynamics (up to four quarters post-mandate)
- Pre-treatment six-quarter coefficients are statistically insignificant → supports parallel pre-trends.
- By the fourth quarter following mandated adoption:
  - Reported sales, purchases and value-added increase by 10 and 15 percent relative to untreated firms (text summary: "by the fourth quarter following treatment, reported sales, purchases and value-added increase by 10 and 15 percent relative to untreated firms.").
  - VAT payments: statistically significant increase in the fourth quarter after treatment in the order of 10 percent.
- Small measured employment increases (just over 2 percent by fourth quarter) relative to value-added increases (~15 percent) — could reflect increased reporting of employment.
- No statistically significant change in the ratio of taxable to total sales or purchases around reform → evidence against simple re-classification of nontaxable sales/purchases.

### Instrumental Variable (LATE) estimates — effects for compliers (actual adopters)
- First stage: Treatment indicator (1 year after) = 0.448***.
- Second-stage (adoption) impacts:
  - Taxable sales: +0.147*** (15 percent increase).
  - Taxable purchases: +0.101*** (10 percent increase).
  - Taxable value-added: +0.132*** (13 percent increase).
  - VAT liabilities (reported): +0.161*** (16 percent increase).
  - New VAT credits: +0.105 (not statistically significant).
  - VAT credit stock: -0.0965 (not statistically significant).
  - Total VAT payments: +0.0993 (positive but not statistically significant).
- Interpretation: LATE magnitudes exceed ITT magnitudes → non-compliance among mandated firms and voluntary adoption among controls bias ITT downward; effects for actual adopters are substantially larger.

### Heterogeneity of effects
- By firm size:
  - Treatment effects driven primarily by relatively smaller firms.
  - Interaction with initial log sales (2013Q2) shows negative and statistically significant interaction → larger effects for smaller firms.
  - Among larger firms, impact on VAT payments is weaker and pooled-sample results are muted by inclusion of large firms.
- By VAT-credit status at sample start:
  - Nearly two-thirds of sample firms had no past VAT credits at estimation start.
  - Firms without past VAT credits:
    - Experience stronger increases in VAT liabilities and pay more VAT after the e-invoicing deadline.
  - Firms with positive initial VAT credit stock:
    - Increase purchases more than sales, do not experience an increase in VAT liabilities, and accumulate more VAT credits post-reform.
- By sector:
  - Statistically significant increases in taxable value-added observed in:
    - Construction (Treatment indicator (1 year): taxable value-added = 0.192***).
    - Transportation (0.0737**).
    - Retail (0.0446**).
    - Other services (0.0604***).
  - No significant treatment effect in manufacturing and hotels/restaurants for taxable value-added.
  - VAT payment responses differ by sector:
    - Other services: significant increase in VAT payments (0.160*** in Table 3, column for total VAT payments).
    - Retail: increase in new VAT credits (0.227***) and VAT credit stock (0.0669), with a decrease in VAT payments (-0.223***).
  - Interpretation: upstream sectors (other services) show stronger increases in VAT payments; downstream sectors (retail) accumulate credits and may offset liabilities.
- Firm survival (extensive margin):
  - Survival defined as having positive sales in the next quarter.
  - Aggregate pattern:
    - Drop in survival rate in the year before mandatory adoption (consistent with firms exiting in anticipation).
    - Rise in survival rate after mandatory adoption (possible selection: more profitable firms remain; control firms pressured to adopt may exit).
  - Sectoral survival heterogeneity:
    - Construction: largest pre-reform drop in survival probability.
    - Transportation and other services: similar but smaller pre-reform drops.
    - Manufacturing: reduction in survival pre-reform and no post-reform improvement.

### Wave-specific and regression highlights (balanced sample / selected coefficients)
- Panel A. Taxable Value Added — Treatment indicator (1 year after) by wave:
  - Wave 1: 0.0805 ( 0.0541 )
  - Wave 2: -0.0702 ( 0.0957 )
  - Wave 3: -0.0512* ( 0.0275 )
  - Wave 4: -0.0295 ( 0.0502 )
  - Wave 5: 0.0425*** ( 0.0147 )
  - Wave 6: 0.0560 ( 0.158 )
- Panel B. VAT payments — Treatment indicator (1 year after) by wave:
  - Wave 1: 0.131 ( 0.125 )
  - Wave 2: 0.293 ( 0.253 )
  - Wave 3: -0.0170 ( 0.0563 )
  - Wave 4: -0.109 ( 0.113 )
  - Wave 5: -0.04590 ( 0.0382 )
  - Wave 6: 0.130 ( 0.579 )

### Aggregate interpretation and contribution
- On average, mandatory e-invoicing increased firm-reported sales, purchases and value-added by over 5 percent in the first year after adoption; effects grow over time and are larger for actual adopters (LATE).
- Effects are concentrated among smaller firms and firms at higher risk of tax non-compliance; largest firms show statistically insignificant impacts on reported tax liabilities.
- E-invoicing appears to improve reporting and compliance among specific taxpayer groups via lowered compliance costs and stronger deterrence.
- Aggregate VAT collections did not rise sharply by 2017 because:
  - Measured effects concentrated in relatively smaller firms that account for less than a fifth of value added in the dataset.
  - Effects build up gradually; SUNAT had not yet fully exploited e-invoicing information flows by 2017.
  - Weaknesses in VAT credit refund/offset mechanisms can dampen translation of reported value-added increases into VAT payments.

### Policy-relevant implications and recommendations
- E-invoicing can be an effective tool to increase voluntary compliance among small and medium-sized firms and those with higher non-compliance risk.
- To maximize revenue and compliance gains:
  - Improve enforcement and monitoring strategies to exploit rich transaction-level data generated by e-invoicing.
  - Strengthen control of VAT credits and refund mechanisms to prevent misuse and to reinforce the self-enforcing nature of the VAT.
  - Provide targeted support and lower digitalization costs for firms in sectors vulnerable to exits (e.g., construction), to avoid unintended transitions to informality.
  - Monitor sectoral heterogeneity to direct state support and heightened oversight where needed.
- Further research suggested:
  - Study spillover effects of e-invoicing adoption on upstream and downstream firms.
  - Assess differential impacts across multiple VAT withholding and collection regimes.

_Italic: Source — wpiea2019231-print-pdf, “1. Timeline of e-Invoicing Adoption Waves in Peru” and section V (Results) extracted content._

### 1. Timeline of e-Invoicing Adoption Waves in Peru _______________________________________________ 8

### 1. Timeline of e-Invoicing Adoption Waves in Peru

### Introduction
- Digitalization is transforming tax administrations and e-invoicing enables automatic transfer of billing information between firms and the tax authority.
- More than 50 countries have implemented e-invoicing, including ten countries in Latin America and the Caribbean (EY, 2018; Barreix and Zambrano, 2018).
- Peru introduced VAT e-invoicing gradually, with early waves focusing on larger firms and priority sectors (such as mining) and later waves including smaller firms.

### Data, sample, and identification
- Administrative monthly panel dataset from SUNAT covering all small, medium and large private-sector firms operating in Peru between 2010 and 2017.
- Balanced sample panel used in the analysis: 78 thousand firms that were mandated to adopt e-invoicing between 2014 and 2018, which account for over 80 percent of domestic VAT collections in Peru.
- Identification strategy exploits sequential (wave-based) mandatory rollout to estimate causal impacts (intent-to-treat and instrumental variable analyses).

### Main empirical findings
- Intent-to-treat (being mandated to adopt e-invoicing) effects in the first year after the mandatory date:
  - Reported taxable sales increase by 7 percent.
  - Reported purchases increase by 5 percent.
  - Value-added growth is not accompanied by a commensurate increase in labor input, suggesting reported output share increased rather than productivity driven by more labor.
  - Large accumulations of past VAT credits allow some firms to offset VAT liabilities, lowering VAT payments in the first year for those firms.
- Heterogeneous VAT collection effects by firm size:
  - Among relatively smaller firms, VAT collections increase by over 5 percent in the first year after adoption.
  - Among large firms, the effect on VAT collections is close to zero and not statistically significant.
- Sectoral heterogeneity:
  - Larger impacts in sectors with traditionally low compliance: retail, business services and construction.
  - Firms in these sectors show stronger responses to e-invoicing adoption and higher exit rates after announcement but before implementation deadlines, consistent with e-invoicing reducing non-compliance and raising effective tax rates for less profitable firms.
- Adoption dynamics and IV results:
  - The rate of e-invoicing adoption increases steadily around the mandatory date in every reform wave, supporting the use of mandate timing as a strong instrument.
  - Instrumental Variable (IV) regressions yield qualitatively similar and quantitatively stronger results than intent-to-treat estimates:
    - Actual e-invoicing adoption increased reported value-added, VAT liabilities and VAT payments by over 10 percent.

### Interpretation and contribution
- The observed increases in reported sales and purchases likely reflect greater compliance and reporting to tax authorities rather than solely real productivity gains.
- E-invoicing affects firm behavior partly by increasing perceived scrutiny, inducing higher reporting and causing some low-compliance firms to exit.
- This study complements prior country studies showing gradual increases in reported sales and VAT collection following e-invoicing adoption, and benefits from a quasi-natural experiment created by sequential mandatory rollout across firms and sectors.

_Italic: Source — wpiea2019231-print-pdf, “1. Timeline of e-Invoicing Adoption Waves in Peru” (extracted content)._

### section 5 discusses the main results. The last section concludes.

### V. RESULTS

### A. Baseline findings
- Method: panel difference-in-differences (Equation (1)) on balanced panel (~78,000 firms; collapsed to quarterly frequency), treatment = mandated e-invoicing (indicator = 1 in quarter mandated and first four quarters after).
- Average estimated impacts (ITT, first year after mandate, controlling for firm variables):
  - Taxable sales: increase of 6.57 percent (Treatment indicator (1 year) = 0.0657***).
  - Taxable purchases: increase of 4.53 percent (0.0453***).
  - Taxable value-added: increase of 5.89 percent (0.0589***).
  - VAT liabilities (reported): increase of 7.22 percent (0.0722***).
  - New VAT credits: increase of 4.70 percent (0.0470; not statistically significant).
  - VAT credit stock: decrease of 4.32 percent (-0.0432; not statistically significant).
  - Total VAT payments: increase of 4.45 percent (0.0445; *p*-value around 0.1 in Panel A, stronger in dynamic results).
  - Productivity: increase of 6.04 percent (0.0604***).
- Dynamic specification (Equation (2), up to four quarters post-mandate):
  - Coefficients in pre-treatment six quarters are statistically insignificant → supports parallel pre-trends.
  - By the fourth quarter following mandated adoption:
    - Reported sales, purchases and value-added increase by 10 and 15 percent relative to untreated firms (text summary: "by the fourth quarter following treatment, reported sales, purchases and value-added increase by 10 and 15 percent relative to untreated firms.").
    - VAT payments: statistically significant increase in the fourth quarter after treatment in the order of 10 percent.
- Interpretation caveats:
  - Data cannot distinguish between real increases in production versus increased reporting of previously unreported activity.
  - No statistically significant change in the ratio of taxable to total sales or purchases around reform → evidence against simple re-classification of nontaxable sales/purchases.
  - Small measured employment increases (just over 2 percent by fourth quarter) relative to value-added increases (~15 percent) — could reflect increased reporting of employment.

### B. Heterogeneity across firms: size, VAT-credit status, sector, survival
- By firm size:
  - Treatment effects driven primarily by relatively smaller firms.
  - Interaction with initial log sales (2013Q2) shows negative and statistically significant interaction → larger effects for smaller firms.
  - Among larger firms, impact on VAT payments is weaker and pooled-sample results are muted by inclusion of large firms.
- By VAT-credit status at sample start:
  - Nearly two-thirds of sample firms had no past VAT credits at estimation start.
  - Firms without past VAT credits:
    - Experience stronger increases in VAT liabilities and pay more VAT after the e-invoicing deadline.
  - Firms with positive initial VAT credit stock:
    - Increase purchases more than sales, do not experience an increase in VAT liabilities, and accumulate more VAT credits post-reform.
- By sector (dynamic results and Table 3):
  - Statistically significant increases in taxable value-added observed in:
    - Construction (Treatment indicator (1 year): taxable value-added = 0.192***).
    - Transportation (0.0737**).
    - Retail (0.0446**).
    - Other services (0.0604***).
  - No significant treatment effect in manufacturing and hotels/restaurants for taxable value-added.
  - VAT payment responses differ by sector:
    - Other services: significant increase in VAT payments (0.160*** in Table 3, column for total VAT payments).
    - Retail: increase in new VAT credits (0.227***) and VAT credit stock (0.0669), with a decrease in VAT payments (-0.223***).
  - Interpretation: upstream sectors (other services) show stronger increases in VAT payments; downstream sectors (retail) accumulate credits and may offset liabilities.
- Firm survival (extensive margin):
  - Survival defined as having positive sales in the next quarter.
  - Aggregate pattern:
    - Drop in survival rate in the year before mandatory adoption (consistent with firms exiting in anticipation).
    - Rise in survival rate after mandatory adoption (possible selection: more profitable firms remain; control firms pressured to adopt may exit).
  - Sectoral survival heterogeneity:
    - Construction: largest pre-reform drop in survival probability.
    - Transportation and other services: similar but smaller pre-reform drops.
    - Manufacturing: reduction in survival pre-reform and no post-reform improvement.

### C. IV (LATE) estimates of actual e-invoicing adoption
- First stage: assignment to treatment strongly predicts actual adoption (First stage coefficient: Treatment indicator (1 year after) = 0.448***).
- Second stage (LATE — effects for compliers who adopted e-invoicing):
  - Adoption indicator impacts (second-stage estimates):
    - Taxable sales: +0.147*** (15 percent increase).
    - Taxable purchases: +0.101*** (10 percent increase).
    - Taxable value-added: +0.132*** (13 percent increase).
    - VAT liabilities (reported): +0.161*** (16 percent increase).
    - New VAT credits: +0.105 (not statistically significant).
    - VAT credit stock: -0.0965 (not statistically significant).
    - Total VAT payments: +0.0993 (positive but not statistically significant).
- Interpretation: LATE magnitudes exceed ITT magnitudes → non-compliance among mandated firms and voluntary adoption among controls bias ITT downward; effects for actual adopters are substantially larger.

### D. Summary interpretation of main results
- On average, mandatory e-invoicing increased firm-reported sales, purchases and value-added by over 5 percent in the first year after adoption; effects grow over time and are larger for actual adopters (LATE).
- Effects are concentrated among smaller firms and firms at higher risk of tax non-compliance; largest firms show statistically insignificant impacts on reported tax liabilities.
- E-invoicing appears to improve reporting and compliance among specific taxpayer groups via lowered compliance costs and stronger deterrence.
- Aggregate VAT collections did not rise sharply by 2017 because:
  - Measured effects concentrated in relatively smaller firms that account for less than a fifth of value added in the dataset.
  - Effects build up gradually; SUNAT had not yet fully exploited e-invoicing information flows by 2017.
  - Weaknesses in VAT credit refund/offset mechanisms can dampen translation of reported value-added increases into VAT payments.

### Policy-relevant implications and recommendations (as concluded)
- E-invoicing can be an effective tool to increase voluntary compliance among small and medium-sized firms and those with higher non-compliance risk.
- To maximize revenue and compliance gains:
  - Improve enforcement and monitoring strategies to exploit rich transaction-level data generated by e-invoicing.
  - Strengthen control of VAT credits and refund mechanisms to prevent misuse and to reinforce the self-enforcing nature of the VAT.
  - Provide targeted support and lower digitalization costs for firms in sectors vulnerable to exits (e.g., construction), to avoid unintended transitions to informality.
  - Monitor sectoral heterogeneity to direct state support and heightened oversight where needed.
- Further research suggested:
  - Study spillover effects of e-invoicing adoption on upstream and downstream firms.
  - Assess differential impacts across multiple VAT withholding and collection regimes.

*Source: IMF working paper section V and VI (Results; Conclusion & Recommendations) contained in the provided PDF content.*

### References

### wpiea2019231-print-pdf - References

### References cited
- Ali, Merima, Abdulaziz Shifa, Abebe Shimeles and Firew Woldeyes. 2015. Information Technology and Fiscal Capacity in a Developing Country: Evidence from Ethiopia, Institute of Development Studies Working Paper No. 11180.
- Artana, Daniel and Ivana Templado. 2018. Análisis del impacto de la Factura Electrónica en la Argentina. Inter-American Development Bank Discussion Paper No. 562.
- Barreix, Alberto D. and Raul Zambrano. eds. 2018. Electronic Invoicing in Latin America. Inter-American Development Bank. doi: http://dx.doi.org/10.18235/0001038.
- Bérgolo, Marcelo, Rodrigo Ceni, and María Sauval. 2018. Factura electrónica y cumplimiento tributario: Evidencia a partir de un enfoque cuasi-experimental. Inter-American Development Bank Discussion Paper No. 561.
- Castro, Hugo J. F., Andrés Z. Carrillo, Sara B. Cortés, Grisel A. Aragón, and María E. S. Diez. 2016. Impacto en la Evasión por la Introducción de la Factura Electrónica. Instituto Tecnológico y de Estudios Superiores de Monterrey. No. IA-006E00002-E37-2016.
- Eissa, Nada, Andrew Zeitlin, Saahil Karpe, and Sally Murray. 2014. Incidence and Impact of Electronic Billing Machines for VAT in Rwanda. Working Paper.
- EY. 2018. Worldwide electronic invoicing survey. https://go.ey.com/2XLaBBB.
- Fan, Haichao, Yu Liu, Nancy Qian, and Jaya Wen. 2018. The Dynamic Effects of Computerized VAT Invoices on Chinese Manufacturing Firms. NBER Working Paper No. 24414.
- Gupta, Sanjeev, Michael Keen, Alpa Shah, and Genevieve Verdier. eds. 2017. Digital Revolutions in Public Finance. International Monetary Fund. doi: http://dx.doi.org/10.5089/9781484315224.071.
- IMF. 2015. Peru: Article IV staff report – Selected Issues. Washington, DC.
- IMF. 2018. Fiscal Monitor (April): Digital Government. Washington, DC.
- Keen, Michael. 2013. The Anatomy of the VAT. National Tax Journal, 66(2), pp 423-446.
- Kochanova, Anna, Zahid Hasnain, and Bradley Larson. 2016. Does e-government improve government capacity? Evidence from tax administration and public procurement. World Bank Policy Research Working Paper No. 7657.
- Lee, Hyung Chul. 2016. Can electronic tax invoicing improve tax compliance? A case study of the Republic of Korea's electronic tax invoicing for value-added tax. World Bank Policy Research Working Paper No. 7592.
- Naritomi, Joana. 2019. Consumers as Tax Auditors. American Economic Review, forthcoming.
- OECD. 2017. Technology Tools to Tackle Tax Evasion and Fraud. Paris: OECD Publishing.
- Okunogbe, Oyebola, and Victor Pouliquen. 2018. Technology, taxation, and corruption: evidence from the introduction of electronic tax filing. World Bank Policy Research Working Paper No. 8452.
- Ramírez, José, Nicolás Oliva, and Mauro Andino. 2018. “Facturación Electrónica en Ecuador: Evaluación de impacto en el cumplimiento tributario”. Inter-American Development Bank Discussion Paper No. 563.
- Yılmaz, Fatih, and Jacqueline Coolidge. 2013. Can e-filing reduce tax compliance costs in developing countries? World Bank Policy Research Working Paper No. 6647.

### Appendix — figures and notes (captions and notes preserved)
- Figure A.1 – e-Invoicing Adoption Rate for Taxpayers with Operaciones No Reales  
  - Note: This figure shows the rate of e-invoice adoption across high-risk firms (ONR firms), using data from SUNAT. The month in which e-invoicing became mandatory is defined as time 0.
- Figure A.2 – Evolution of e-Invoicing Adoption by Wave and by Initial Adoption Deadline  
  - Note: This figure shows the rate of e-invoice adoption for selected waves and their corresponding initial deadline for adoption (the vertical dotted line), using data from SUNAT.
- Figure A.3 – e-Invoicing Adoption Rates Around the Initial and Final Deadlines  
  - Panel A: non ONR waves  
  - Panel B: ONR waves  
  - Note: This figure shows the rate of e-invoice adoption for different waves, using data from SUNAT.
- Figure A.4. Percent Change in Taxable Value-Added Around the Mandatory Dates of Adoption  
  - Note: Graph of the coefficient estimates of value-added and their confidence interval from the dynamic difference-in-differences regression for the balanced sample. Each graph corresponds to the sample of the firms mandated in the specified wave and those mandated in 2018.
- Figure A.5 Percent Change in VAT payments Around the Mandatory Dates of Adoption  
  - Note: Graph of the coefficient estimates of value-added and their confidence interval from the dynamic difference-in-differences regression for the balanced sample. Each graph corresponds to the sample of the firms mandated in the specified wave and those mandated in 2018.
  - (percent change by quarter from the adoption deadline)

### Appendix — selected table highlights and exact statistics
- Table A.1 – Summary Statistics for the Entire Database Over 2014-2017: main variables (1/2) — Values are in thousand 2014 soles unless otherwise specified; in 2014, the exchange rate was approximately 0.34 US$ per soles.
  - total sales: 1,316,000 ( 1,768,000 ); 659 ( 5,551 ); 158,100 ( 183,100 ); 102,400 ( 262,500 ); 15,450 ( 125,700 ); 700 ( 1,181 ); 2,678 ( 13,890 ); 759 ( 3,634 )
  - total purchases: 1,047,000 ( 1,733,000 ); 606 ( 5,334 ); 114,000 ( 142,900 ); 87,220 ( 215,800 ); 11,430 ( 87,320 ); 643 ( 1,183 ); 2,213 ( 14,730 ); 629 ( 4,243 )
  - value added: 268,900 ( 944,700 ); 53 ( 309 ); 44,120 ( 126,900 ); 15,220 ( 154,100 ); 4,024 ( 70,050 ); 57 ( 263 ); 466 ( 8,287 ); 131 ( 3,667 )
  - gross VAT: 165,400 ( 243,200 ); 115 ( 999 ); 22,000 ( 23,290 ); 13,820 ( 20,860 ); 1,721 ( 8,805 ); 112 ( 196 ); 372 ( 1,875 ); 102 ( 516 )
  - invoices issued electronically in 2018 (%): 0.85 ( 0.317 ); 0.63 ( 0.408 ); 0.91 ( 0.243 ); 0.94 ( 0.200 ); 0.93 ( 0.212 ); 0.82 ( 0.315 ); 0.34 ( 0.370 ); 0.18 ( 0.339 )
  - VAT credit stock: 20,540 ( 100,800 ); 4 ( 76 ); 2,433 ( 19,950 ); 3,523 ( 16,240 ); 477 ( 27,400 ); 3 ( 28 ); 90 ( 8,625 ); 27 ( 568 )
  - number of workers: 1,729 ( 2,766 ); 4 ( 15 ); 710 ( 1,280 ); 287 ( 474 ); 103 ( 601 ); 2 ( 5 ); 14 ( 44 ); 6 ( 32 )
  - observations: 9344, 7343, 3,078, 2,073, 45,075, 1,950, 294,434, 309,498
  - share of total value-added: 32% 0% 17% 4% 23% 0% 18% 5%
  - Waves identified (labels preserved): Firms never mandated or mandated after (October 2014) wave 7; (January 2015) wave 1; (July 2015) wave 2; (July 2016) wave 3; (December 2016) wave 4; (January 2017) wave 5; (many 2018 deadlines) wave 6.
- Table A.2 – Summary Statistics for the Entire Database Over 2014-2017: sector and risk distributions (2/2)
  - Total observations: 9344, 7343, 3,078, 2,073, 45,075, 1,950, 294,434, 309,498
  - Total value-added (million soles): 262,332; 256,141; 144,32 (sic as in source) [presented as 262,332256141,14432,852189,278115143,80341,905 in table layout]
  - High risk: 26.9 23.9 18.6 16.8 26.5 25.7 29.2 13.0 17.1 20.1 16.8 18.7 25.1 21.0 24.8 17.8 (table shows multiple wave-specific risk shares)
  - Medium risk and Low risk rows preserved in source formatting (percentages listed per wave).
  - VAT credit stock in 2013Q2: 45.7 39.8 30.5 30.1 29.6 20.1 43.5 27.7 38.5 19.6 24.6 31.9 28.1 33.3 24.7 (values as presented)
- Table A.3 / Table A.4 – Balanced Panel sample summaries (selected)
  - total sales: 1,315,000 ( 1,755,000 ); 168,300 ( 182,900 ); 101,800 ( 217,900 ); 18,780 ( 138,300 ); 2,883 ( 8,641 )
  - total purchases: 1,049,000 ( 1,735,000 ); 121,100 ( 144,200 ); 83,200 ( 143,900 ); 13,700 ( 92,370 ); 2,262 ( 7,534 )
  - value added: 266,000 ( 940,600 ); 47,140 ( 125,700 ); 18,610 ( 156,400 ); 5,080 ( 79,820 ); 622 ( 4,506 )
  - invoices issued electronically in 2018 (%): 0.85 ( 0.318 ); 0.91 ( 0.239 ); 0.94 ( 0.199 ); 0.93 ( 0.208 ); 0.32 ( 0.358 )
  - observations (balanced panel): 9002, 2,756, 1,768, 33,792, 173,508
  - share of total value-added: 35% 19% 5% 25% 16% (per wave labels)

### Table A.5 — Regression results by wave for selected variables (balanced sample)
- Panel A. Taxable Value Added (in log of soles unless otherwise indicated)
  - Treatment indicator (1 year after): Wave 1: 0.0805 ( 0.0541 ); Wave 2: -0.0702 ( 0.0957 ); Wave 3: -0.0512* ( 0.0275 ); Wave 4: -0.0295 ( 0.0502 ); Wave 5: 0.0425*** ( 0.0147 ); Wave 6: 0.0560 ( 0.158 )
  - Number of workers: Wave 1: 0.338*** ( 0.0139 ); Wave 2: 0.321*** ( 0.0140 ); Wave 3: 0.321*** ( 0.0132 ); Wave 4: 0.335*** ( 0.0151 ); Wave 5: 0.367*** ( 0.0142 ); Wave 6: 0.337*** ( 0.0158 )
  - Wage bill: Wave 1: -0.0209*** ( 0.00392 ); Wave 2: -0.0183*** ( 0.00400 ); Wave 3: -0.0157*** ( 0.00392 ); Wave 4: -0.0100** ( 0.00478 ); Wave 5: -0.00380 ( 0.00477 ); Wave 6: -0.00348 ( 0.00524 )
  - Capital stock: Wave 1: 0.00627*** ( 0.00202 ); Wave 2: 0.00490*** ( 0.00158 ); Wave 3: 0.00134 ( 0.00157 ); Wave 4: 0.00220 ( 0.00161 ); Wave 5: 0.00250* ( 0.00148 ); Wave 6: 0.00212 ( 0.00157 )
  - Constant terms: Wave 1–6 all ~9.565*** to 9.721*** with standard errors shown in parentheses.
  - Observations: Wave 1: 394,989; Wave 2: 396,471; Wave 3: 403,355; Wave 4: 402,353; Wave 5: 479,439; Wave 6: 362,841
- Panel B. VAT payments (in log of soles unless otherwise indicated)
  - Treatment indicator (1 year after): Wave 1: 0.131 ( 0.125 ); Wave 2: 0.293 ( 0.253 ); Wave 3: -0.0170 ( 0.0563 ); Wave 4: -0.109 ( 0.113 ); Wave 5: -0.04590 ( 0.0382 ); Wave 6: 0.130 ( 0.579 )
  - Number of workers: Wave 1: 0.547*** ( 0.0372 ); Wave 2: 0.524*** ( 0.0360 ); Wave 3: 0.538*** ( 0.0379 ); Wave 4: 0.543*** ( 0.0357 ); Wave 5: 0.626*** ( 0.0337 ); Wave 6: 0.562*** ( 0.0385 )
  - Wage bill: Wave 1: -0.0506*** ( 0.00890 ); Wave 2: -0.0457*** ( 0.00914 ); Wave 3: -0.0473*** ( 0.00943 ); Wave 4: -0.0278*** ( 0.00980 ); Wave 5: -0.0176* ( 0.00986 ); Wave 6: -0.0224** ( 0.0106 )
  - Capital stock: Wave 1: -0.0004050 ( 0.00526 ); Wave 2: 0.003160 ( 0.00416 ); Wave 3: 0.003500 ( 0.00472 ); Wave 4: 0.0111** ( 0.00479 ); Wave 5: 0.00762* ( 0.00442 ); Wave 6: 0.00521 ( 0.00480 )
  - Constant terms: Wave 1–6 ~4.732*** to 4.964*** with standard errors shown in parentheses.
  - Observations: Wave 1: 479,621; Wave 2: 479,600; Wave 3: 484,726; Wave 4: 482,009; Wave 5: 570,053; Wave 6: 434,300

*Italic source attribution: wpiea2019231-print-pdf - References*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019231-print-pdf.pdf_