Skip to main content

Advertisement

SpringerLink
Log in

The productivity effect of digital financial reporting

  • Published:
Review of Accounting Studies Aims and scope Submit manuscript

Abstract

We examine the effect of digital financial reporting on firm productivity. Information frictions represent a constraint that impedes efficient resource allocation and a major source of such frictions stems from the fact that firms’ production functions (the conversion from inputs to outputs) are not observable to corporate outsiders. Digital communication of corporate financial data fundamentally changes how firm-specific information is disclosed, released, and disseminated by mitigating information asymmetry between corporate insiders and outsiders and facilitates the processing of such information. We use the staggered implementation of the SEC’s Electronic Data Gathering and Analysis Retrieval (EDGAR) system to investigate the impact of digital financial reporting on firms’ productivity. We show that the implementation of EDGAR results in an economically meaningful and statistically significant increase on firms’ productivity, measured by total factor productivity (TFP). By focusing on the role of information dissemination in coordinating investments and production, our findings provide evidence on the real effects of “going digital” in corporate reporting.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1

Similar content being viewed by others

Notes

  1. A 1982 New York Times article quoted Maryann Wismer—then a researcher at Disclosure Incorporated, a company specializing in retrieving financial information for private customers—as exclaiming “it’s just incredible the number of problems you can run into trying to find something you need.” See the full text at: https://www.nytimes.com/1982/05/19/business/sec-data-difficult-hunt.html.

  2. The SEC’s effort to promote capital market information efficiency is ongoing. Since the adoption of the EDGAR system, the SEC has been actively planning and developing additional technological innovations to improve reporting transparency. For example, to provide a level playing field to different types of investors, in April 2009, the SEC mandated that firms use XBRL when preparing their financial statements over three phase-in periods.

  3. We carefully review variables of interest in these studies and draw their connections to ours in Sect. 4.3.

  4. Shleifer and Vishny (1986) show that, monitoring will only occur when the expected benefit of monitoring outweighs the related cost of doing so.

  5. An alternative explanation is that investors, upon observing firms’ operating outcomes, provide capital to the most efficient ones. We overcome this concern by employing a firm fixed effects regression strategy so that the coefficient on the EDGAR implementation indicator captures the within-firm trends in TFP.

  6. We thank an anonymous reviewer for this suggestion.

  7. Morris and Shin (2002) show that in the absence of private information, greater provision of public information always increases social welfare (i.e., the sum of all traders’ utilities).

  8. Results are similar if we use alternative time windows (from 1990 to 2000 and from 1960 to 2019) to estimate TFP.

  9. İmrohoroğlu and Tüzel (2014) report their estimated \({\beta }_{K}\) and \({\beta }_{L}\) as being 0.23 and 0.75, respectively, using the Olley and Pakes method (discussed in Sect. 5.1).

  10. For all of the empirical tests, this definition of Post-EDGAR allows the financial information users to observe the publicly disseminated financial information through the EDGAR system.

  11. Ideally, we would have used the marginal Tobin’s Q. However, due to the empirical challenge of estimating the marginal Tobin’s Q documented in prior studies, we instead follow Biddle et al. (2009) and use the average Tobin’s Q in the regression.

  12. When firms report missing R&D or advertising expenses, we fill a value of zero for these observations. To the extent that Koh and Reeb (2015) show that firms that indeed engage in innovation may report missing R&D expenses, we assess the sensitivity of this design choice by excluding firms that report missing R&D. Results (untabulated) and inferences are qualitatively similar.

  13. We thank an anonymous reviewer for this suggestion.

  14. In alternative specifications, we use industry and group fixed effects. Results and inferences are qualitatively similar. Because firm fixed effects are finer than and subsume both the group fixed effects and industry fixed effects, we use firm fixed effects in all tests.

  15. We also benchmark our coefficient on Post-EDGAR against To et al. (2018). In their baseline regression, the coefficient on their main independent variable, LnCoverage, is 0.037 (column (3) in Table 3, with firm-level characteristics, year, and industry fixed effects included). In their Table 2, they report that the standard deviation of LnCoverage is 0.933. That is, a one standard deviation change in LnCoverage results in an increase in TFP of 0.034. In other words, the introduction of digital reporting in EDGAR results in an increase in TFP that is very comparable to, and slightly higher than, a one standard deviation change in analyst coverage.

  16. Lev and Sougiannis (1996) show that imputed R&D capitalization contains value-relevant information for investors, but contemporaneous stock prices do not fully reflect R&D capital. Stock returns in subsequent years are also positively associated with their imputed R&D capital.

  17. In untabulated tests, we capture financial reporting quality using an alternative, disclosure-based index motivated by Chen et al. (2015). We define Disclosure as the simple average of balance sheet disclosure quality (DQBS) and income statement disclosure quality (DQIS). We follow the procedures outlined in Chen et al. (2015) to calculate DQBS and DQIS. A higher value of Disclosure signifies better disaggregation of accounting data and thus higher disclosure quality. For completeness, we also include the first-order difference of this disclosure score (∆Disclosure). The coefficient on Disclosure is positive and statistically significant at the 1% level, consistent with firms with better reporting quality experiencing higher productivity (Biddle et al. 2009). Most importantly, the coefficient on Post-EDGAR remains statistically significant at the 1% level.

  18. We first compound daily returns into weekly returns for each firm and year. We then estimate an expanded market model regression (with weekly returns) with a lead and a lag term for the market index return. The firm-specific weekly return is defined as the natural log of one plus the residual return. We define a week as a crash week when a firm experiences a firm-specific weekly return that is 3.2 standard deviations below the mean firm-specific weekly return over the entire fiscal year, with 3.2 chosen to generate a frequency of 0.1% in the normal distribution.

  19. We estimate the PIN score using a combined approach. For observations after the availability of NYSE Trade and Quote (TAQ) data (since 1993), we use the intraday transaction-level data provided by TAQ to estimate the PIN score. For observations prior to 1993 (when the transaction-level data from the Institute for the Study of Security Markets (ISSM) is not available), we directly use the PIN data (Easley, Hvidkjaer, and O’Hara 2010) provided directly from Professor Soeren Hvidkjaer’s website.

  20. Claus and Thomas (2001) employ a residual-income valuation model that adds current book value per share with discounted expected residual earnings per share up to five years. They assume residual earnings grow at an expected inflation rate minus 3% after the five-year forecast horizon and that 50% of earnings are paid out as dividends each period. We obtain the inflation data from the Federal Reserve Bank of St. Louis and then solve the valuation equation to obtain the discount factor as an estimate of the implied cost of equity.

  21. Both Gao and Huang (2020) and Chang et al. (2021) describe their (private) correspondences with the SEC staff.

  22. Shipman et al. (2017) provide excellent discussions on matching with and without replacement. Matching without replacement may result in lower-quality matches.

  23. This design choice is conservative. While not all firms in Group CF-01 are transitional filers, all transitional filers are assigned to Group CF-01.

  24. The annual subscription fee for real-time data through Mead Data Central was $150,000. The non-real-time subscription fee was $75,000. The pay-as-you-go per filing rate ranged between $20 and $30.

  25. This mechanical relation is unlikely to be the explanation for regressions with year fixed effects, while Post-EDGAR is defined using staggered adoption dates.

  26. Shroff et al. (2014) argue that TFP captures an economic construct similar to investment efficiency.

  27. In concurrent work, Goldstein et al. (2021) also explore the setting of EDGAR adoption and investigate whether managerial learning is affected by its implementation. Using investment-to-price sensitivity to proxy for “revelatory price efficiency” (i.e., the extent to which prices reveal new information to managers), they show that the EDGAR implementation leads to a decrease in investment-to-price sensitivity. Their work is incrementally different from ours in that we focus on investment efficiency (i.e., how corporate investment may deviate from its optimal level) rather than on the level of investment per se or how sensitive the level of investment is, related to firm valuation.

  28. The mean investment inefficiency in the full sample is 0.13, whereas the inefficiency measures are 0.15 and 0.11 for the over-(under-) investment subsamples, respectively.

  29. Another possible cross-sectional test is to investigate the potential moderating role of institutional ownership. This idea is motivated by the possibility that some institutional investors may have access to subscribed third-party services to observe firms’ fundamental information prior to the adoption of the EDGAR system. However, we may not observe a dampened effect of digital reporting in EDGAR on firms’ productivity for firms with high institutional ownership for at least two reasons. First, prior to the adoption of the EDGAR system, machine-readable format data provided by commercial vendors may not deliver exactly the same quality of data as the full-scale digital financial reports do. Even if institutional investors have access to such third-party data, there is still a lot of information in the 10-K filings that would have not been captured in the third-party data. For example, the most popular corporate financial database is Standard and Poor’s Compustat, and most of the data fields in Compustat are tabulated numbers that are recognized in financial statements (Schipper 2007). Instead, numbers (together with their detailed contexts) that are disclosed only in the accompanying footnotes as well as other places in the 10-K (e.g., management discussion and analysis – MD&A) are often uncaptured by such a product. As such, institutional investors still may not be able to uncover a firm’s full picture of corporate operations and financial positions based on their access to a third-party database. Second, the intensity of monitoring provided by institutional investors versus retail investors is different. There is a vast literature documenting that the institutional investors perform a more active role in monitoring a firm’s management. For example, Hartzell and Starks (2003) show that institutional ownership concentration is positively related to the pay-for-performance sensitivity of executive compensation and negatively related to the level of compensation. As such, compared with firms with concentrated institutional ownership, firms with lower institutional ownership may not experience a higher increase in TFP as they receive less intense monitoring from their investor base (primarily made up of retail investors).

  30. We also use a third-order polynomial function to represent \(\varphi ({k}_{t}\), \({l}_{t}\), \({m}_{t})\). Results and inferences are qualitatively similar.

References

  • Ackerberg, D., K. Caves, and G. Frazer. 2015. Identification properties of recent production function estimators. Econometrica 83 (6): 2411–2451.

    Article  Google Scholar 

  • Altman, E. 2013. Predicting financial distress of companies: revisiting the Z-score and ZETA® models. In Handbook of research methods and applications in empirical finance. Edward Elgar Publishing.

    Google Scholar 

  • Asthana, S., S. Balsam, and S. Sankaraguruswamy. 2004. Differential response of small versus large investors to 10-K filings on EDGAR. The Accounting Review 79 (3): 571–589.

    Article  Google Scholar 

  • Atanasov, V., and B. Black. 2016. Shock-based causal inference in corporate finance and accounting research. Critical Finance Review 5: 207–304.

    Article  Google Scholar 

  • Banker, R., R. Huang, Y. Li, and S. Zhao. 2021. Do Accounting Standards Matter for Productivity? Production and Operations Management 30 (1): 68–84.

    Article  Google Scholar 

  • Bennett, B., R. Stulz, and Z. Wang. 2020. Does the stock market make firms more productive? Journal of Financial Economics 136 (2): 281–306.

    Article  Google Scholar 

  • Bertomeu, J., E. Cheynel, E. Floyd, and W. Pan. 2020. Using machine learning to detect misstatements. Review of Accounting Studies 26 (2): 468–519.

    Article  Google Scholar 

  • Bertrand, M., and S. Mullainathan. 2003. Enjoying the quiet life? Corporate governance and managerial preferences. Journal of Political Economy 111 (5): 1043–1075.

    Article  Google Scholar 

  • Bhushan, R. 1989. Firm characteristics and analyst following. Journal of Accounting and Economics 11 (2–3): 255–274.

    Article  Google Scholar 

  • Biddle, G., G. Hilary, and R. Verdi. 2009. How does financial reporting quality relate to investment efficiency? Journal of Accounting and Economics 48 (2–3): 112–131.

    Article  Google Scholar 

  • Bloom, N., and J. Van Reenen. 2007. Measuring and explaining management practices across firms and countries. Quarterly Journal of Economics 122 (4): 1351–1408.

    Article  Google Scholar 

  • Bushman, R., and A. Smith. 2001. Financial accounting information and corporate governance. Journal of Accounting and Economics 32 (1–3): 237–333.

    Article  Google Scholar 

  • Butler, A., G. Grullon, and J. Weston. 2005. Stock market liquidity and the cost of issuing equity. Journal of Financial and Quantitative Analysis 40 (2): 331–348.

    Article  Google Scholar 

  • Chaney, T., D. Sraer, and D. Thesmar. 2012. The collateral channel: How real estate shocks affect corporate investment. American Economic Review 102 (6): 2381–2409.

    Article  Google Scholar 

  • Chang, X., S. Dasgupta, and G. Hilary. 2006. Analyst coverage and financing decisions. Journal of Finance 61 (6): 3009–3048.

    Article  Google Scholar 

  • Chang, Y., P. Hsiao, A. Ljungqvist, and K. Tseng. 2022. Testing disagreement models. Journal of Finance 77 (4): 2239–2285.

    Article  Google Scholar 

  • Chang, Y., A. Ljungqvist, and K. Tseng. 2021. Do corporate disclosures constrain strategic analyst behavior? Working Paper.

  • Chen, Q., I. Goldstein, and W. Jiang. 2007. Price informativeness and investment sensitivity to stock price. Review of Financial Studies 20 (3): 619–650.

    Article  Google Scholar 

  • Chen, S., B. Miao, and T. Shevlin. 2015. A new measure of disclosure quality: The level of disaggregation of accounting data in annual reports. Journal of Accounting Research 53 (5): 1017–1054.

    Article  Google Scholar 

  • Cheng, Q., F. Du, X. Wang, and Y. Wang. 2016. Seeing is believing: Analysts’ corporate site visits. Review of Accounting Studies 21 (4): 1245–1286.

    Article  Google Scholar 

  • Choi, J., R. Hann, M. Subasi, and Y. Zheng. 2020. An empirical analysis of analysts’ capital expenditure forecasts: Evidence from corporate investment efficiency. Contemporary Accounting Research 37 (4): 2615–2648.

    Article  Google Scholar 

  • Claus, J., and J. Thomas. 2001. Equity premia as low as three percent? Evidence from analysts’ earnings forecasts for domestic and international stock markets. Journal of Finance 56 (5): 1629–1666.

    Article  Google Scholar 

  • Das, S., R. Guo, and H. Zhang. 2006. Analysts’ selective coverage and subsequent performance of newly public firms. Journal of Finance 61 (3): 1159–1185.

    Article  Google Scholar 

  • David, J., H. Hopenhayn, and V. Venkateswaran. 2016. Information, misallocation, and aggregate productivity. Quarterly Journal of Economics 131 (2): 943–1005.

    Article  Google Scholar 

  • DeAngelo, L. 1988. Managerial competition, information costs, and corporate governance: The use of accounting performance measures in proxy contests. Journal of Accounting and Economics 10 (1): 3–36.

    Article  Google Scholar 

  • Dechow, P., and I. Dichev. 2002. The quality of accruals and earnings: The role of accrual estimation errors. The Accounting Review 77 (s-1): 35–59.

    Article  Google Scholar 

  • deHaan, E. 2020. Using and interpreting fixed effects models. Working paper. University of Washington.

  • Dugast, J., and T. Foucault. 2018. Data abundance and asset price informativeness. Journal of Financial Economics 130 (2): 367–391.

    Article  Google Scholar 

  • Durnev, A., R. Morck, and B. Yeung. 2004. Value-enhancing capital budgeting and firm-specific stock return variation. Journal of Finance 59 (1): 65–105.

    Article  Google Scholar 

  • Easley, D., and M. O’hara. 1992. Time and the process of security price adjustment. Journal of Finance 47 (2): 577–605.

    Article  Google Scholar 

  • Easley, D., S. Hvidkjaer, and M. O’Hara. 2010. Factoring information into returns. Journal of Financial and Quantitative Analysis 45 (2): 293–309.

    Article  Google Scholar 

  • Fazzari, S., R. Hubbard, and B. Petersen. 1988. Financing Constraints and Corporate Investment. Brookings Papers on Economic Activity 1: 141–195.

    Article  Google Scholar 

  • Francis, J., R. LaFond, P. Olsson, and K. Schipper. 2004. Costs of equity and earnings attributes. The Accounting Review 79 (4): 967–1010.

    Article  Google Scholar 

  • Gao, M., and J. Huang. 2020. Informing the market: The effect of modern information technologies on information production. Review of Financial Studies 33 (4): 1367–1411.

    Article  Google Scholar 

  • Glaeser, S. 2018. The effects of proprietary information on corporate disclosure and transparency: Evidence from trade secrets. Journal of Accounting and Economics 66 (1): 163–193.

    Article  Google Scholar 

  • Goldstein, I., S. Yang, and L. Zuo. 2021. The real effects of modern information technologies. Cornell University Working Paper.

  • Gomez, E. 2020, The Effect of mandatory disclosure dissemination on information asymmetry: evidence from the implementation of the EDGAR System. Working Paper.

  • Guo, F., L. Lisic, M. Stuart, and C. Wang. 2019. The impact of information technology on stock price crash risk: evidence from the EDGAR implementation. Working Paper.

  • Hall, R., and C. Jones. 1999. Why do some countries produce so much more output per worker than others? Quarterly Journal of Economics 114 (1): 83–116.

    Article  Google Scholar 

  • Hall, B. 1990. The Manufacturing Sector Master File: 1959–1987. National Bureau of Economic Research. Working paper.

  • Hann, R., H. Kim, W. Wang, and Y. Zheng. 2020. Information frictions and productivity dispersion: The role of accounting information. The Accounting Review 95 (3): 223–250.

    Article  Google Scholar 

  • Harford, J., and K. Li. 2007. Decoupling CEO wealth and firm performance: The case of acquiring CEOs. Journal of Finance 62 (2): 917–949.

    Article  Google Scholar 

  • Hart, O., and J. Moore. 1995. Debt and seniority: An analysis of the role of hard claims in constraining management. American Economic Review 85 (3): 567.

    Google Scholar 

  • Hartzell, J., and L. Starks. 2003. Institutional investors and executive compensation. Journal of Finance 58 (6): 2351–2374.

    Article  Google Scholar 

  • Hillegeist, S., E. Keating, D. Cram, and K. Lundstedt. 2004. Assessing the probability of bankruptcy. Review of Accounting Studies 9 (1): 5–34.

    Article  Google Scholar 

  • Hsieh, C., and P. Klenow. 2009. Misallocation and manufacturing TFP in China and India. Quarterly Journal of Economics 124 (4): 1403–1448.

    Article  Google Scholar 

  • Huang, A., R. Lehavy, A. Zang, and R. Zheng. 2018. Analyst information discovery and interpretation roles: A topic modeling approach. Management Science 64 (6): 2833–2855.

    Article  Google Scholar 

  • İmrohoroğlu, A., and Ş Tüzel. 2014. Firm-level productivity, risk, and return. Management Science 60 (8): 2073–2090.

    Article  Google Scholar 

  • Jensen, M. 1986. Agency costs of free cash flow, corporate finance, and takeovers. American Economic Review 76 (2): 323–329.

    Google Scholar 

  • Kambil, A., and M. Ginsburg. 1998. Public access Web information systems: Lessons from the Internet EDGAR project. Communications of the ACM 41 (7): 91–97.

    Article  Google Scholar 

  • Kim, J., Y. Li, and L. Zhang. 2011. Corporate tax avoidance and stock price crash risk: Firm-level analysis. Journal of Financial Economics 100 (3): 639–662.

    Article  Google Scholar 

  • Koh, P., and D. Reeb. 2015. Missing R&D. Journal of Accounting and Economics 60 (1): 73–94.

    Article  Google Scholar 

  • Lai, S., C. Lin, and X. Ma. 2020. Regtech Adoption and the Cost of Capital, Working Paper.

  • Lee, C., P. Ma, and C. Wang. 2015. Search-based peer firms: Aggregating investor perceptions through internet co-searches. Journal of Financial Economics 116 (2): 410–431.

    Article  Google Scholar 

  • Lehn, K., and M. Zhao. 2006. CEO turnover after acquisitions: Are bad bidders fired? Journal of Finance 61 (4): 1759–1811.

    Article  Google Scholar 

  • Leuz, C., and R. Verrecchia. 2004. Firms’ capital allocation choices, information quality, and the cost of capital. Working paper. University of Pennsylvania.

  • Lev, B., and T. Sougiannis. 1996. The capitalization, amortization, and value-relevance of R&D. Journal of Accounting and Economics 21 (1): 107–138.

    Article  Google Scholar 

  • Lin, C., C. Ma, Y. Sun, and Y. Xu. 2021. The telegraph and modern banking development, 1881–1936. Journal of Financial Economics 141 (2): 730–749.

    Article  Google Scholar 

  • Loughran, T., and B. McDonald. 2017. The use of EDGAR filings by investors. Journal of Behavioral Finance 18 (2): 231–248.

    Article  Google Scholar 

  • McNichols, M. 2002. Discussion of the quality of accruals and earnings: the role of accrual estimation errors. The Accounting Review 77 (s-1): 61–69.

    Article  Google Scholar 

  • Morris, S., and H. Shin. 2002. Social value of public information. American Economic Review 92 (5): 1521–1534.

    Article  Google Scholar 

  • Myers, S. 1977. Determinants of corporate borrowing. Journal of Financial Economics 5 (2): 147–175.

    Article  Google Scholar 

  • Myers, S., and N. Majluf. 1984. Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics 13 (2): 187–221.

    Article  Google Scholar 

  • Olley, G., and A. Pakes. 1996. The dynamics of productivity in the telecommunications equipment industry. Econometrica 64 (6): 1263–1297.

    Article  Google Scholar 

  • Palepu, K. 1986. Predicting takeover targets: A methodological and empirical analysis. Journal of Accounting and Economics 8 (1): 3–35.

    Article  Google Scholar 

  • Restuccia, D., and R. Rogerson. 2017. The causes and costs of misallocation. Journal of Economic Perspectives 31 (3): 151–174.

    Article  Google Scholar 

  • Roychowdhury, S., N. Shroff, and R. Verdi. 2019. The effects of financial reporting and disclosure on corporate investment: A review. Journal of Accounting and Economics 68 (2–3): 101246.

    Article  Google Scholar 

  • Schipper, K. 2007. Required disclosures in financial reports. The Accounting Review 82 (2): 301–326.

    Article  Google Scholar 

  • Shipman, J., Q. Swanquist, and R. Whited. 2017. Propensity score matching in accounting research. The Accounting Review 92 (1): 213–244.

    Article  Google Scholar 

  • Shleifer, A., and R. Vishny. 1986. Large shareholders and corporate control. Journal of Political Economy 94 (3–1): 461–488.

    Article  Google Scholar 

  • Shroff, N., R. Verdi, and G. Yu. 2014. Information environment and the investment decisions of multinational corporations. The Accounting Review 89 (2): 759–790.

    Article  Google Scholar 

  • Smith, C., and J. Warner. 1979. On financial contracting: An analysis of bond covenants. Journal of Financial Economics 7 (2): 117–161.

    Article  Google Scholar 

  • Solow, R. 1956. A contribution to the theory of economic growth. Quarterly Journal of Economics 70 (1): 65–94.

    Article  Google Scholar 

  • Solow, R. 1957. Technical change and the aggregate production function. Review of Economics and Statistics, 312–320.

  • To, T., M. Navone, and E. Wu. 2018. Analyst coverage and the quality of corporate investment decisions. Journal of Corporate Finance 51: 164–181.

    Article  Google Scholar 

  • Welker, M. 1995. Disclosure policy, information asymmetry, and liquidity in equity markets. Contemporary Accounting Research 11 (2): 801–827.

    Article  Google Scholar 

  • Yoshikawa, H. 1980. On the “q” theory of investment. American Economic Review 70 (4): 739–743.

    Google Scholar 

  • Yu, F. 2008. Analyst coverage and earnings management. Journal of Financial Economics 88 (2): 245–271.

    Article  Google Scholar 

Download references

Acknowledgements

We thank Stephen Penman (Editor) and two anonymous reviewers for very helpful comments. We are indebted to Michael Welker, who provided constant encouragement, supervision, and support while we worked on this project. We also thank Paul Calluzzo, Yolande Chan, Linda Myers, Nancy L. Su, Kai Sun, Changqiu Yu, Ying Zhang, Steven Zheng, and seminar participants at Queen’s University and the University of Manitoba. Zhang and Liu acknowledge supports from the Commerce ’83 Fellowship at Queen’s University and the Chartered Professional Accountants Research Fellowship at University of Manitoba, respectively. Liu also thanks the financial support provided by Queen’s University, where the initial draft was completed during his doctoral study at the Smith School of Business.

Author information

Authors and Affiliations

  1. School of Business, Hong Kong Baptist University, Hong Kong, China

    Zheng Liu

  2. Smith School of Business, Queen’s University, Kingston, ON, Canada

    Ning Zhang

Authors
  1. Zheng Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Ning Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Ning Zhang.

Additional information

Publisher's note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendix: Econometrics on the estimation of TFP

Appendix: Econometrics on the estimation of TFP

This appendix provides the technical details for the estimation of TFP.

  • Econometrics on TFP estimation

To calculate TFP, we start with the log-linearized Cobb–Douglas production function:

$$y={\beta }_{0}+{\beta }_{K}*k+{\beta }_{L}*l+u$$
(A1)

As discussed in the main text, \(u\) can be decomposed into two terms \(\omega\) and \(\varepsilon\):

$$y={\beta }_{0}+{\beta }_{K}*k+{\beta }_{L}*l+\underset{\overbrace{\underset{\begin{array}{l}observable\; to\\ the\; manager\end{array}}{\underbrace{\omega }}+\underset{\begin{array}{l}unobservable\; to\\ the\; manager\end{array}}{\underbrace{\varepsilon }}}}{{u}}$$
(A2)

When making investment decisions, the manager can choose k and l upon observing \(\omega\) or a noise measure of \(\omega\). In other words, under this situation, the independent variables (k and l) are correlated with the error term (\(u\)), making the estimation of Eq. (3) from an ordinary least squares (OLS) regression biased.

To alleviate the concern, we follow Ackerberg et al. (2015) to estimate the production function. There are several assumptions in this model:

  • Assumption 1: (Information set) The firm’s information set at time t is \({I}_{t}\). It includes both current and past production shocks but does not include any future production shock.

  • Assumption 2: (First-order Markov) Productivity shocks evolve according to the distribution \(p\left({\omega }_{t+1}\right|{I}_{t})=p\left({\omega }_{t+1}\right|{\omega }_{t})\).

  • Assumption 3: Intermediate input (\(m\)) is chosen at the same time or after \({l}_{t}\) is chosen.

  • Assumption 4: \({m}_{t}\) is strictly increasing in \({\omega }_{t}\).

Based on Assumption 3, \({m}_{t}\) is written as:

$${m}_{t}=f\left({k}_{t},{l}_{t},{\omega }_{t}\right)$$
(A3)

To the extent that \({m}_{t}\) is strictly increasing in \({\omega }_{t}\), we invert Equation (A3) and represent \({\omega }_{t}\) as a function of \({k}_{t}\), \({l}_{t}\) and \({m}_{t}\):

$${\omega }_{t}={f}^{-1}\left({k}_{t},{l}_{t},{m}_{t}\right)$$
(A4)

Thus, we rewrite equation (A2) (with time subscripts) by substituting \({\omega }_{t}\) using Equation (A4):

$${y}_{t}={\beta }_{0}+{\beta }_{K}*{k}_{t}+{\beta }_{L}*{l}_{t}+{f}^{-1}\left({k}_{t},{l}_{t},{m}_{t}\right)+{\varepsilon }_{t}$$
(A5)

Now, the output \({y}_{t}\) is now a non-parametric function of \({k}_{t}\), \({l}_{t}\), and \({m}_{t}\).

$${y}_{t}=\varphi \left({k}_{t},{l}_{t},{m}_{t}\right)+{\varepsilon }_{t}$$
(A6)

where \(\varphi ({k}_{t}\), \({l}_{t}\), \({m}_{t})\) = \({\beta }_{0} \, + {\beta }_{K}*{k}_{t}+ {\beta }_{L}*{l}_{t} \, + \, {f}^{-1}(\) \({k}_{t}\), \({l}_{t}\), \({m}_{t})\).

Apply the first moment condition here:

$$E\left[\left.{\varepsilon }_{t}\right|{I}_{t}\right]=E\left[\left.{y}_{t}-\varphi \left({k}_{t},{l}_{t},{m}_{t}\right)\right|{I}_{t}\right]=0$$
(A7)

We use a second-order polynomial function to re-express \(\varphi ({k}_{t}\), \({l}_{t}\), \({m}_{t})\):

$$\varphi \left({k}_{t}, {l}_{t}, {m}_{t}\right)={\theta }_{0}+{\theta }_{K}\text{*}{k}_{t}+{\theta }_{L}\text{*}{l}_{t}+{\theta }_{M}\text{*}{m}_{t}+{\theta }_{{K}^{2}}\text{*}{k}_{t}^{2}+{\theta }_{{L}^{2}}\text{*}{l}_{t}^{2}+{\theta }_{{M}^{2}}\text{*}{m}_{t}^{2}+{\theta }_{KL}\text{*}{k}_{t}\text{*}{l}_{t}+{\theta }_{KM}\text{*}{k}_{t}\text{*}{m}_{t}+{\theta }_{LM}\text{*}{l}_{t}\text{*}{m}_{t}$$

In the first stage regression, we estimate the above equation and obtain the fitted value \(\widehat{\varphi }({k}_{t}\), \({l}_{t}\), \({m}_{t})\).Footnote 30 Assumptions 1 and 2 imply that \({\omega }_{t}\) can be decomposed into the expected value at time t−1 and a residual term:

$${\omega }_{t}=E\left[\left.{\omega }_{t}\right|{I}_{t-1}\right]+{\mu }_{t}=E\left[\left.{\omega }_{t}\right|{\omega }_{t-1}\right]+{\mu }_{t}=g\left({\omega }_{t-1}\right)+{\mu }_{t}$$
(A8)

Therefore, Equation (A2) can be re-written (with time subscripts) into:

$${y}_{t}={\beta }_{0}+{\beta }_{K}*{k}_{t}+{\beta }_{L}*{l}_{t}+g\left({\omega }_{t-1}\right)+{\mu }_{t}+{\varepsilon }_{t}$$
(A9)

Given that \(E\left[{\mu }_{t}|{I}_{t-1}\right]\) = 0 and \(E\left[{\omega }_{t}|{I}_{t}\right]\) = 0, the second moment condition is:

$$E\left[{\mu }_{t}+{\varepsilon }_{t}|{I}_{t-1}\right]=E\left[{y}_{t}-{ \beta }_{0} - \, {\beta }_{K} \, *{k}_{t} \, - \, {\beta }_{L}*{l}_{t}- \, g\text{(}{\omega }_{t-1}\text{)|}{I}_{t-1}\right]=0$$
(A10)

Rewrite \({\omega }_{t-1}\) into:

$${\omega }_{t-1}=\varphi \left({k}_{t-1}, {l}_{t-1}, {m}_{t-1}\right)-{\beta }_{0}-{\beta }_{K}*{k}_{t-1}-{\beta }_{L}*{l}_{t-1}$$
(A11)

By inserting Equation (A11) into Equation (A10), Equation (A10) becomes:

$$E\left[\left.{y}_{t}-{\beta }_{0} - \, {\beta }_{K}*{k}_{t}-{\beta }_{L}*{l}_{t}-g\left(\varphi \left({k}_{t-1},{l}_{t-1},{m}_{t-1}\right)-{\beta }_{0}-{\beta }_{K}*{k}_{t-1}-{\beta }_{L} *{l}_{t-1}\right)\right|{I}_{t-1}\right]=0$$
(A12)

Now, replace \(\varphi ({k}_{t-1}\), \({l}_{t-1}\), \({m}_{t-1})\) with the fitted value \(\widehat{{\varphi }_{t-1}}\). Following İmrohoroğlu and Tüzel (2014), we set \({\omega }_{t}=g\left({\omega }_{t-1}\right)=\rho {\omega }_{t-1}+{\mu }_{t},\) and Equation (A12) can be re-written as:

$$E\left[\left.{y}_{t}-{ \beta }_{0} -{\beta }_{K}*{k}_{t} \, - \, {\beta }_{L}*{l}_{t}- \, \rho \left(\widehat{{\varphi }_{t-1}}-{\beta }_{0 }- \, {\beta }_{K} *{k}_{t-1}-{\beta }_{L}*{l}_{t-1}\right)\right|{I}_{t-1}\right]=0$$
(A13)

Following İmrohoroğlu and Tüzel (2014), we use the non-linear least squares model to estimate \({\beta }_{0}\), \({\beta }_{K}\), \({\beta }_{L}\), and \(\rho\).

  • Data and variable definitions in TFP estimation

To calculate TFP for the firms in our sample, we follow Bennett et al. (2020) and obtain accounting data from Compustat. We obtain the price index for gross domestic product (GDP) and the price index for private fixed investment from the Bureau of Economic Analysis. Value added (Y) is defined as sales (SALE) minus materials scaled by GDP deflator. Material (M) is defined as the difference between total expense and labor expense deflated by the GDP price index. Total expense is revenue (REVT) minus operation income before depreciation and amortization (OIBDP). We use staff expense (XLR) in Compustat to calculate total labor expense. When this value is missing, we first calculate the average wage per employee within a Fama–French-12 industry using all non-missing wages in that specific industry, then impute a firm’s labor cost using the number of employees in the firm times the industry-average wage per employee. Capital stock (K) is defined as gross property, plant, and equipment (PPEGT) scaled by the price deflator and adjusted for the age of the capital stock, following Hall (1990). Labor (L) is the number of employees.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Liu, Z., Zhang, N. The productivity effect of digital financial reporting. Rev Account Stud (2023). https://doi.org/10.1007/s11142-022-09737-6

Download citation

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1007/s11142-022-09737-6

Keywords

JEL Classification

Search

Navigation