Abstract
This study proposes a new approach of measuring compliance costs of rules and regulations by focusing on labor input, and estimates the compliance costs in Japan based on a survey of workers. According to the results, the working hours required to comply with rules and regulations account for more than 20% of total labor input. By industry, this cost is higher in the finance and insurance industry followed by the health and welfare industry, and by firm size, it is higher in large firms. If these costs were halved, overall economic productivity would increase by about 8%.
Keywords: Regulation, Compliance cost, Productivity, Labor input
Introduction
Improving productivity growth is a major policy issue in advanced economies. Many studies have shown that government regulation of businesses, such as entry restrictions and price control, negatively impacts productivity, whereas deregulation increases it. In line with such studies, deregulation has been promoted in advanced countries since the 1970s. However, in the United States, it has been indicated that economic deregulation in sectors such as telecommunications, transportation, and energy, has been more than offset in areas such as the environment and occupational safety, with the total amount of regulations increasing at an average rate of 3.5% annually (Dawson & Seater, 2013).
Similarly, in Japan, industry-specific economic regulations, such as those applied to telecommunications, electricity and gas, and retail trade have been gradually eased for more than 30 years, since the 1980s. Conversely, labor market, environmental, and consumer protection regulations (hereinafter referred to as “social regulations”) have been on the rise. According to the Current Status of Licenses and Permits (Ministry of Internal Affairs and Communications), the number of licenses and permits increased by 47% (2.5% per year) between 2002 and 2017, with ministries such as the Financial Services Agency, Ministry of Health, Labour and Welfare, and Ministry of the Environment contributing significantly to this trend (see Table 1).
Table 1.
The number of licenses and permits by ministries in Japan
| 2002 | 2017 | Annal rate (%) | Contribution (%) | |
|---|---|---|---|---|
| Cabinet office | 219 | 294 | 2.0 | 1.5 |
| Financial services agency | 1421 | 2353 | 3.4 | 19.2 |
| Ministry of internal affairs and communications | 575 | 718 | 1.5 | 2.9 |
| Ministry of justice | 237 | 360 | 2.8 | 2.5 |
| Ministry of foreign affairs | 47 | 43 | − 0.6 | − 0.1 |
| Ministry of finance | 727 | 842 | 1.0 | 2.4 |
| Ministry of education, culture, sports, science and technology | 566 | 473 | − 1.2 | − 1.9 |
| Ministry of health, labour and welfare | 1543 | 2451 | 3.1 | 18.7 |
| Ministry of agriculture, forestry and fisheries | 1114 | 1770 | 3.1 | 13.5 |
| Ministry of economy, trade and industry | 1866 | 2261 | 1.3 | 8.1 |
| Ministry of land, infrastructure, transport and tourism | 2042 | 2805 | 2.1 | 15.7 |
| Ministry of the environment | 229 | 1075 | 10.9 | 17.4 |
| Ministry of defence | 35 | 30 | − 1.0 | − 0.1 |
| Total | 10,621 | 15,475 | 2.5 | 100 |
Calculated from the Current Status of Licenses and Permits (Ministry of Internal Affairs and Communications). The figures of the Cabinet Office include those of the Cabinet Secretariat, Fair Trade Commission, and National Public Safety Commission
However, the number of legal licenses and permits does not necessarily indicate the strength of regulations or their economic impact. To cope with regulations, firms are forced to perform many complicated tasks, such as preparing legal documents, creating and storing inspection data, and negotiating with government agencies. The quantity of tasks varies significantly depending on the specific regulations.
In addition to legal rules and regulations, various types of compliance are required by administrative guidance, rules within the industry, and firms’ internal rules. During the COVID-19 pandemic, for example, various administrative guidance was provided to business activities by central and local governments. Legal regulations, guidance, and rules have a negative impact on productivity through increased compliance costs that do not directly contribute to production. Moreover, they may suppress productivity growth in the medium- to long-term through their negative impact on firms’ risk-taking and innovation. However, it is difficult to obtain a comprehensive and accurate picture of the number and strength of rules and regulations, including those that do not target specific industries.
The OECD developed and published indicators of product market regulation (PMR) based on questions posed to national governments regarding regulation-related laws, and, more recently, indicators of the impact of regulation on downstream industries (REGIMPACT). However, the indicators were created from the perspective of barriers to new entry and competition, which do not capture the overall compliance costs. Furthermore, industry-specific indicators in the non-manufacturing sector are limited to energy, transportation, telecommunications, and professional services.
In the United States, there have been new attempts to quantify federal government regulations, such as a page count of the Code of Federal Regulations (CFR; Dawson & Seater, 2013), and industry-specific regulatory indicators (RegData) based on textual analysis of the CFR (Al-Ubaydli & McLaughlin, 2017; McLaughlin & Sherouse, 2019). Recent studies use the RegData to analyze the impact of regulations on the declining business dynamism in the United States (e.g., Bailey & Thomas, 2017; Goldschlag & Tabarrok, 2018; Gutiérrez & Philippon, 2019), although conclusions are not uniform across studies.
However, it is highly debatable whether the quantity of regulation itself captures impacts and costs to industries and firms. To overcome this limitation, Calomiris et al. (2020) applied natural language processing to data from the corporate earnings calls of listed firms in the United States to create a firm-level cost of regulation index and demonstrate that regulation increases the cost of capital for firms. This is a unique approach to measuring the cost of regulation for individual firms; however, due to the nature of the data, it does not cover non-listed firms.
This study proposes a novel approach that focuses on the amount of labor input required to comply with rules and regulations, and provides an estimate of the actual compliance costs in Japan using data collected through an original survey of workers. According to the results, the working hours required to comply with rules and regulations account for more than 20% of total labor input. By industry, this cost is greater in the finance and insurance industry followed by the health and welfare industry, and by firm size, it is greater in large firms. A sizable portion of working hours of high-wage earners is devoted to these tasks, and if these labor costs could be halved, it would increase the productivity of the economy by about 8%.
The remainder of this study is structured as follows. Section 2 provides a brief overview of the related literature. Section 3 explains the survey data and tabulation method. Section 4 reports the results. Section 5 concludes the paper and discusses future directions for research.
Literature review
Previous studies have shown that economic regulations that restrict competition, including entry and price regulations, negatively impact productivity and economic growth, and eliminating or relaxing these regulations positively impact productivity (see, for example, Winston, 1993; Crafts, 2006; Holmes and Schmitz, 2010, for surveys). Nicoletti and Scarpetta (2003) and Haidar (2012) are examples of studies using cross-country data to indicate that regulatory reforms that promote competition positively affect economic growth. OECD regulatory indicators (PMR and REGIMPACT) are frequently used in empirical analyses using cross-country data. Andrews and Cingano (2014) combined firm data from several countries with OECD regulatory indicators and found that regulation impedes resource allocation efficiency and negatively impacts productivity. This study is unique because it covers not only product market regulations but also labor market regulations.
Studies that comprehensively cover economic and social regulations in the United States include Dawson and Seater (2013) and Coffey et al. (2020). Dawson and Seater (2013) conducted a unique study that used the number of CFR pages as a measure of regulation quantity. They demonstrated that the total amount of federal government regulation had been trending upward, mainly through an increase in environmental and safety regulations. They estimated that the increase in regulation had significantly reduced output and total factor productivity (TFP), with a magnitude that would reduce the postwar U.S. economic growth rate by two percentage points per year. Coffey et al. (2020), based on a text analysis of the CFR, estimated that regulatory restrictions reduced the US economic growth rate by approximately 0.8% points per annum since 1980.
Many studies have addressed the impact of social regulations that are not specific to an industry. In the following, we focus on studies dealing with the effects of such regulations on productivity and innovation. Empirical studies on labor market regulations showing negative impacts on productivity include Cingano et al. (2010), Andrews and Cingano (2014), Cette et al. (2016), Égert (2016), and Amoroso and Martino (2020). In the area of labor market regulations, several countries adopt “size-dependent regulations” that impose strict labor market regulations on publicly listed or large firms, while exempting or reducing the regulations on unlisted or small firms. Studies on French policy, where stringent labor regulations are imposed on firms with over 50 employees, have found that the policy negatively impacts productivity at the aggregate level (Garicano et al., 2016; Gourio & Roys, 2014). Aghion et al. (2021) reported that this size-dependent regulation reduces innovation.1
Regarding land-use regulations, Cheshire et al. (2014) in the United Kingdom and Herkenhoff et al. (2018) in the United States indicated that land-use regulations result in reduced productivity. Studies on environmental regulations include Greenstone et al. (2012), Albrizio et al. (2017), and Feng et al. (2021), although conclusions about the impacts of environmental regulations on productivity are mixed. Recent studies analyzing the privacy regulation of EU General Data Protection Regulation (GDPR) indicate its negative impacts on firm performance and innovation (e.g., Janßen et al., 2022; Jia et al., 2021; Johnson, 2022).
To summarize, many studies have analyzed the impact of rules and regulations on productivity, and each has its own strengths and weaknesses. In particular, the lack of quantitative data on compliance costs, covering all rules and regulations, has been an important limitation. This study attempts to capture, quantitatively, compliance costs in Japan by employing a novel approach that measures the labor input needed to comply with rules and regulations through a survey of workers.Although it is a straightforward method, this study is unique because it presents broad compliance costs by industry and worker characteristics, covering all rules and regulations, including cross-industry social regulations as well as rules formulated by industry associations and firms’ internal rules.
Outline of the survey
The survey data used in this study were retrieved from the “Survey of Life and Consumption under the Changing Economic Structure” designed by the author and conducted by Rakuten Insight, Inc. The Survey of Life and Consumption under the Changing Economic Structure was a recurring survey that began in 2017 and was conducted also in 2020 and 2021.2 Rakuten Insight, Inc., a subsidiary of Rakuten, Inc., which is the largest online retailer in Japan, is a representative Internet research company. The 2021 survey was conducted in July with monitors (potential respondents) aged 20 years or older registered with Rakuten Insight, Inc. More than two million people were registered with Rakuten Insight, Inc. The composition of registered monitors by individual characteristics is confidential; however, the number of registered monitors is sufficiently large to obtain responses that match the overall gender and age composition in Japan.
The 2021 survey was conducted to obtain more than 8000 responses, including approximately 4985 respondents from the 2020 survey and additional respondents. Additional respondents were chosen to ensure that the composition of respondents by gender and age was consistent with that of the Japanese population at the time of the survey.3 The total number of respondents was 8909 (4479 repeat respondents and 4430 new respondents); however, after excluding those who were not working, data from 5707 workers were used in this study. The composition of respondents by worker characteristics (gender, age, employment type, industry, occupation, and firm size) and the composition of the workforce in Japan (Labor Force Survey (LFS) in 2021 conducted by the Statistics Bureau, Ministry of Internal Affairs and Communications) are shown in "Appendix Table 5".4 Compared with the composition of the workforce in Japan, male respondents were overrepresented, and respondents in their 20 s were underrepresented in the 2021 survey. As approximately half of the respondents participated in the 2017 survey, the aging of respondents in subsequent years affected their age distribution. However, correcting for gender and age composition using LFS weights did not significantly affect the results (described in Sect. 4).
Table 5.
Composition of the responded workers
| (1) N | (2) Percentages (%) | (3) LFS (2021) (%) | ||
|---|---|---|---|---|
| Gender | Male | 3496 | 61.3 | 55.3 |
| Female | 2211 | 38.7 | 44.7 | |
| Age | Age 20 s | 568 | 10.0 | 15.3 |
| Age 30 s | 951 | 16.7 | 17.8 | |
| Age 40 s | 1420 | 24.9 | 23.5 | |
| Age 50 s | 1359 | 23.8 | 21.6 | |
| Age 60 or older | 1409 | 24.7 | 21.9 | |
| Employment type | Executives | 333 | 5.8 | 5.2 |
| Self-employed and family workers | 677 | 11.9 | 10.1 | |
| Standard employee | 2964 | 51.9 | 54.4 | |
| Non-standard employee | 1733 | 30.4 | 30.3 | |
| Industry | Construction | 324 | 5.7 | 7.2 |
| Manufacturing | 938 | 16.4 | 15.6 | |
| Information and communications | 255 | 4.5 | 3.8 | |
| Transport | 233 | 4.1 | 5.3 | |
| Wholesale and retail | 592 | 10.4 | 16.0 | |
| Finance and insurance | 219 | 3.8 | 2.5 | |
| Real estate | 163 | 2.9 | 2.1 | |
| Accommodations and restaurants | 162 | 2.8 | 5.5 | |
| Health care and welfare | 639 | 11.2 | 13.3 | |
| Education | 322 | 5.6 | 5.2 | |
| Other services | 915 | 16.0 | 14.7 | |
| Public services | 346 | 6.1 | 3.7 | |
| Other industries | 599 | 10.5 | 5.1 | |
| Occupation | Managerial | 638 | 11.2 | 1.9 |
| Professional and engineering | 1176 | 20.6 | 18.8 | |
| Clerical | 1188 | 20.8 | 20.7 | |
| Sales & trade-related | 651 | 11.4 | 12.6 | |
| Service | 682 | 12.0 | 12.0 | |
| Production | 227 | 4.0 | 12.9 | |
| Other occupations | 1145 | 20.1 | 21.1 | |
| Firm size | 99 or smaller | 2807 | 49.2 | 40.7 |
| 100–299 | 686 | 12.0 | 18.8 | |
| 300–499 | 329 | 5.8 | ||
| 500–999 | 394 | 6.9 | 7.2 | |
| 1000 or larger | 1233 | 21.6 | 24.3 | |
| Government | 258 | 4.5 | 9.0 | |
| Total | 5707 | 100.0 | 100 | |
The figures in columns (1) and (2) are the worker subsamples of the respondents to the “Survey of Life and Consumption under the Changing Economic Structure” conducted in July 2021. The percentages in column (3) are calculated from the Labor Force Survey (LFS) in 2021
The distribution of the respondents’ employment type, industry, and firm size did not significantly differ from that of the LFS. However, the composition of respondents’ occupations differed from that of the LFS, with more respondents were in managerial occupations and fewer are in production occupations. This difference possibly reflected the composition of the registered monitors. As the working hours associated with complying with rules and regulations are likely to be longer for managers than production workers, the average compliance cost may be overestimated. However, the quantitative impact was limited, as explained in Sect. 4.
The key question used in this study was, “What share of your total working hours is spent to comply with rules and regulations, including rules formulated at the industry or firm level?” This broad definition was adopted, as ordinary workers may not be able to distinguish government regulations from industrial and internal rules. The seven response choices were “100%,” “50–99%,” “25–49%,” “10–24%,” “5–9%,” “1–4%,” and “I do not perform such work.” Although we would prefer specific figures, it is difficult for individual workers to provide exact percentages. Therefore, we set up multiple choices. In surveys for individual participants, wordings such as “most,” “quite a bit,” or “somewhat” are often used; however, in such a design the answer is affected by respondents’ subjective senses and interpretations of the wordings.
The ratio of working hours devoted to compliance (hereinafter, “compliance-related working hours”) was tabulated based on worker characteristics. Specifically, we calculated the means and standard deviations by gender, age (10-years intervals), educational background, employment type, industry, occupation, firm size, weekly working hours, and wage level (annual earnings from work). The median value for each choice was used, and the choice “I do not perform such work” was treated as zero. However, as part-time workers were included, we also measured the absolute amount of labor input rather than merely the share of hours worked. As the survey inquired regarding weekly working hours, compliance-related working hours per week were calculated as the weekly working hours for each worker multiplied by the ratio of compliance-related working hours.5
Furthermore, when discussing the effect of compliance costs on productivity, it is preferable to consider worker wages. As the survey examined annual employment earnings, figures weighted by workers’ hourly wages were calculated.6 These figures were aggregated to calculate the share of compliance-related working hours to the total labor input, weighted by wages, for the entire economy, as well as by industry and firm size.
Results
Table 2 presents the results of tabulating the share of time spent complying with rules and regulations to the total hours worked. Nearly half of the respondents (46.1%) reported that they did not perform such tasks (“0%”). Excluding these responses, a large number of respondents chose “10–24%,” “25–49%,” and “50–99%.”
Table 2.
Share of compliance-related working hours
| Share of compliance-related hours (%) | Percentages (%) |
|---|---|
| 100 | 3.5 |
| 50 ~ 99 | 12.2 |
| 25 ~ 49 | 13.6 |
| 10 ~ 24 | 14.8 |
| 5 ~ 9 | 6.4 |
| 1 ~ 4 | 3.3 |
| 0 | 46.1 |
N = 5707. “I do not perform such work” was treated as 0%
The share of compliance-related working hours by major worker characteristics is summarized in Column (1) of Table 3. Column (2) shows the deviations from the baseline categories. The detailed tabulation results are shown in "Appendix Table 6", in which the means and standard deviations of the share (Column (1)) and average hours per week (Column (2)) of compliance-related work are reported. For all workers, the average share of compliance-related working hours is 20.7% and the average number of hours per week is 7.93. Therefore, on average, approximately one working day per week is devoted to such work. However, the standard deviations are large (28.9% and 11.98 h), indicating that there is a large dispersion of the regulatory burden among workers.
Table 3.
Worker characteristics and compliance-related working hours
| (1) Mean share (%) | (2) Diff | (3) Conditional diff | (4) Zero hour (%) | (5) N | |||
|---|---|---|---|---|---|---|---|
| All workers | 20.7 | – | – | 46.1 | 5707 | ||
| Gender | Male | 22.2 | – | – | 40.4 | 3496 | |
| Femle | 18.3 | − 3.9 | − 2.0 | 55.2 | 2211 | ||
| Age | 20 s | 27.5 | 6.4 | 6.0 | * | 36.4 | 568 |
| 30 s | 24.0 | 3.0 | 2.5 | * | 41.9 | 951 | |
| 40 s | 21.0 | – | – | 45.4 | 1420 | ||
| 50 s | 20.5 | − 0.5 | − 0.7 | 45.6 | 1359 | ||
| 60 or older | 15.7 | − 5.4 | − 1.7 | 54.2 | 1409 | ||
| Education | High school | 19.6 | – | – | 55.2 | 1461 | |
| Vocational school | 18.8 | − 0.8 | − 2.7 | * | 50.0 | 658 | |
| Junior (2-year) college | 18.8 | − 0.8 | − 0.9 | 53.9 | 579 | ||
| 4-year university | 22.1 | 2.5 | − 2.7 | * | 40.4 | 2582 | |
| Graduate school | 21.3 | 1.7 | − 5.1 | * | 32.9 | 420 | |
| Employment type | Executives | 23.3 | − 2.3 | 1.0 | 27.6 | 333 | |
| Self-employed and family workers | 13.6 | − 12.0 | − 3.7 | * | 58.8 | 677 | |
| Standard employee | 25.6 | – | – | 35.6 | 2964 | ||
| Non-standard employee | 14.6 | − 11.0 | − 6.2 | * | 62.8 | 1733 | |
| Industry | Manufacturing | 21.9 | – | – | 38.2 | 938 | |
| Wholesale and retail | 16.8 | − 5.1 | − 2.4 | 53.7 | 592 | ||
| Finance and insurance | 30.0 | 8.1 | 6.2 | * | 31.5 | 219 | |
| Accommodations and restaurants | 16.5 | − 5.4 | 0.4 | 57.4 | 162 | ||
| Health care and welfare | 23.2 | 1.3 | 5.0 | * | 44.4 | 639 | |
| Public services | 37.5 | 15.6 | 14.7 | * | 30.6 | 346 | |
| Occupation | Managerial | 24.6 | − 0.4 | − 0.1 | 24.0 | 638 | |
| Professional and engineering | 21.7 | − 3.3 | − 3.3 | * | 39.8 | 1176 | |
| Clerical | 25.0 | – | – | 42.9 | 1188 | ||
| Service | 16.5 | − 8.5 | − 4.0 | * | 54.0 | 682 | |
| Production | 17.6 | − 7.4 | − 5.5 | * | 59.9 | 227 | |
| Other occupations | 16.3 | − 8.7 | − 4.6 | * | 61.7 | 1145 | |
| Firm size | 99 or smaller | 16.3 | – | – | 55.1 | 2807 | |
| 100–299 | 21.2 | 4.9 | 2.5 | * | 44.5 | 686 | |
| 300–499 | 23.0 | 6.7 | 4.3 | * | 37.4 | 329 | |
| 500–999 | 24.3 | 7.9 | 4.3 | * | 39.6 | 394 | |
| 1000 or larger | 25.7 | 9.4 | 6.9 | * | 33.1 | 1233 | |
To reduce the size of the table, figures for some categories are omitted. The full results are reported in "Appendix Tables 6 and 7". Column (2) indicates deviations from the reference categories expressed as “–.” Column (3) shows the coefficients obtained from OLS estimations. *Significantly different from zero (p < 0.05). The reference categories for age, education, industry, occupation, and firm size are 40–49 years of age, high school, manufacturing, clerical occupation, and fewer than 100 employees, respectively. Column (4) shows the percentage of workers with zero compliance hours
Table 6.
Compliance-related working hours by worker characteristics
| (1) Share (%) | (2) Absolute hours | (3) Zero hour (%) | (4) N | ||||
|---|---|---|---|---|---|---|---|
| Mean | Std. Dev | Mean | Std. dev. | ||||
| All workers | 20.7 | 28.9 | 7.93 | 11.98 | 46.1 | 5707 | |
| Gender | Male | 22.2 | 28.9 | 9.04 | 12.55 | 40.4 | 3496 |
| Femle | 18.3 | 28.8 | 6.18 | 10.79 | 55.2 | 2211 | |
| Age | 20 s | 27.5 | 31.2 | 10.46 | 13.41 | 36.4 | 568 |
| 30 s | 24.0 | 30.0 | 9.53 | 12.89 | 41.9 | 951 | |
| 40 s | 21.0 | 28.9 | 8.32 | 12.16 | 45.4 | 1420 | |
| 50 s | 20.5 | 28.7 | 8.24 | 12.37 | 45.6 | 1359 | |
| 60 or older | 15.7 | 26.3 | 5.15 | 9.39 | 54.2 | 1409 | |
| Education | High school | 19.6 | 30.1 | 7.25 | 12.07 | 55.2 | 1461 |
| Vocational school | 18.8 | 28.3 | 7.18 | 11.87 | 50.0 | 658 | |
| Junior (2-years) college | 18.8 | 29.1 | 6.27 | 10.58 | 53.9 | 579 | |
| 4-year university | 22.1 | 28.6 | 8.77 | 12.18 | 40.4 | 2582 | |
| Graduate school | 21.3 | 26.4 | 8.72 | 12.04 | 32.9 | 420 | |
| Employment type | Executives | 23.3 | 26.3 | 9.60 | 12.38 | 27.6 | 333 |
| Self-employed and family workers | 13.6 | 25.3 | 5.05 | 10.90 | 58.8 | 677 | |
| Standard employee | 25.6 | 30.0 | 10.56 | 13.11 | 35.6 | 2964 | |
| Non-standard employee | 14.6 | 26.9 | 4.24 | 8.58 | 62.8 | 1733 | |
| Industry | Construction | 21.0 | 26.0 | 8.17 | 10.87 | 38.6 | 324 |
| Manufacturing | 21.9 | 27.9 | 8.96 | 11.94 | 38.2 | 938 | |
| Information and communications | 20.5 | 25.5 | 8.14 | 10.16 | 34.5 | 255 | |
| Transport | 22.8 | 31.2 | 9.14 | 14.13 | 47.6 | 233 | |
| Wholesale and retail | 16.8 | 26.7 | 6.33 | 10.74 | 53.7 | 592 | |
| Finance and insurance | 30.0 | 32.9 | 12.05 | 14.20 | 31.5 | 219 | |
| Real estate | 19.3 | 28.3 | 6.15 | 9.91 | 48.5 | 163 | |
| Accommodations and restaurants | 16.5 | 27.4 | 5.64 | 10.49 | 57.4 | 162 | |
| Health care and welfare | 23.2 | 29.9 | 8.77 | 12.75 | 44.4 | 639 | |
| Education | 19.1 | 28.9 | 7.38 | 13.40 | 51.2 | 322 | |
| Other services | 17.3 | 26.8 | 6.26 | 10.27 | 51.5 | 915 | |
| Public services | 37.5 | 36.0 | 14.68 | 15.18 | 30.6 | 346 | |
| Other industries | 13.8 | 25.6 | 4.88 | 9.94 | 61.1 | 599 | |
| Occupation | Managerial | 24.6 | 26.8 | 10.35 | 12.13 | 24.0 | 638 |
| Professional and engineering | 21.7 | 28.2 | 8.74 | 12.51 | 39.8 | 1176 | |
| Clerical | 25.0 | 31.5 | 9.49 | 12.89 | 42.9 | 1188 | |
| Sales and trade-related | 20.8 | 28.2 | 8.08 | 11.68 | 44.9 | 651 | |
| Service | 16.5 | 26.9 | 5.38 | 9.65 | 54.0 | 682 | |
| Production | 17.6 | 29.0 | 7.46 | 12.52 | 59.9 | 227 | |
| Other occupations | 16.3 | 28.4 | 5.67 | 11.02 | 61.7 | 1145 | |
| Firm size | 99 or smaller | 16.3 | 26.7 | 5.85 | 10.66 | 55.1 | 2807 |
| 100–299 | 21.2 | 28.5 | 8.33 | 12.00 | 44.5 | 686 | |
| 300–499 | 23.0 | 28.7 | 9.24 | 12.32 | 37.4 | 329 | |
| 500–999 | 24.3 | 29.3 | 9.85 | 12.29 | 39.6 | 394 | |
| 1,000 or larger | 25.7 | 29.9 | 10.24 | 12.71 | 33.1 | 1,233 | |
| Government | 35.2 | 36.6 | 13.92 | 15.64 | 36.0 | 258 | |
| Working hours (weekly) | Shorter than 35 h | 16.4 | 27.7 | 3.27 | 5.94 | 58.6 | 1994 |
| 35–42 | 23.3 | 29.8 | 8.96 | 11.45 | 41.2 | 1732 | |
| 43 h or longer | 22.8 | 28.7 | 11.72 | 15.04 | 37.9 | 1981 | |
| Annual earnings | Less than 3 million | 15.9 | 27.7 | 5.03 | 9.72 | 60.9 | 2504 |
| 3–5.99 | 24.1 | 30.0 | 9.84 | 13.16 | 39.1 | 1916 | |
| 6–9.99 | 25.6 | 29.1 | 11.04 | 13.10 | 30.4 | 969 | |
| 10 million or more | 23.0 | 24.7 | 9.77 | 12.05 | 20.4 | 318 | |
Column (3) is the percentages of workers responded that they did not perform compliance-related tasks
As explained in Sect. 3, male respondents were overrepresented and respondents in their 20 s were underrepresented in the survey. After the gender and age composition is corrected using the LFS (2021) weights, the weighted means are slightly larger (21.0% and 7.99 h); however, the differences are negligible. Managers were overrepresented and production workers were underrepresented in the survey. When the occupational composition is adjusted by the LFS weights, the weighted means are somewhat smaller (19.9% and 7.48 h) than those of the unadjusted figures; however, the result is largely unchanged.
Worker characteristics related to relatively high compliance-related working hours are men, young adults (20–39 years of age), highly educated, executives, standard (full-time regular) employees, finance and insurance industry, public service, management, clerical jobs, and large firms (see Table 3). In contrast, self-employed, non-standard employees, wholesale and retail, accommodations and restaurants, service jobs, production jobs, and small firms are associated with low compliance-related working hours. However, there is a large dispersion (standard deviation) within the same category (see "Appendix Table 6"). Although not reported in Table 3, workers with shorter weekly working hours and low wages (annual earnings) spend fewer compliance-related working hours (see "Appendix Table 6").7
When weighted by wages, the mean share of compliance-related working hours is somewhat larger (23.3%), as low-wage workers tend to spend less time complying with rules and regulations. To assess the impact on productivity, it is appropriate to examine the weighted figure. As total employee compensation out of GDP is approximately 289 trillion yen (FY2021) in Japan, a mechanical calculation of macroeconomic compliance costs yields a huge figure of approximately 67 trillion yen (289 × 0.233), corresponding to 12.4% of GDP.8
"Appendix Table 7" presents the results of ordinary least squares (OLS) estimations that explain compliance-related working hours (share in Column (1) and absolute hours in Column (2)) by worker characteristics. Column (3) of Table 3 selectively presents the estimated coefficients, which indicate conditional differences from the baseline categories. Most of the coefficients confirm the simple cross-tabulation results. The estimated coefficients for 20–39 years of age, finance and insurance industry, healthcare and welfare, public service, and larger firm size categories are positive and significant, indicating that these characteristics are positively associated with compliance-related working hours. By contrast, the coefficients for self-employed, non-standard employees, professional and engineering occupations, service occupations, and production occupations are negative and significant. These results are generally consistent with the observations from simple cross-tabulations, even after controlling for other worker characteristics, although the conditional differences are smaller than the raw differences.
Table 7.
Worker characteristics and compliance-related working hours
| (1) Share | (2) Absolute hours | (3) Zero hour | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Coef. | Robust SE | Coef. | Robust SE | dF/dx | Robust SE | ||||
| Female | − 1.953 | (0.947) | ** | − 0.543 | (0.376) | 0.045 | (0.017) | *** | |
| Age 20 s | 6.008 | (1.511) | *** | 2.137 | (0.612) | *** | − 0.102 | (0.026) | *** |
| Age 30 s | 2.453 | (1.206) | ** | 1.128 | (0.501) | ** | − 0.035 | (0.022) | |
| Age 50 s | − 0.741 | (1.070) | − 0.250 | (0.442) | 0.016 | (0.020) | |||
| Age 60 or older | − 1.721 | (1.102) | − 0.503 | (0.412) | − 0.001 | (0.021) | |||
| Vocational school | − 2.701 | (1.359) | ** | − 1.038 | (0.547) | * | − 0.017 | (0.025) | |
| Junior (2-year) college | − 0.895 | (1.460) | − 0.456 | (0.520) | − 0.028 | (0.026) | |||
| 4-year university | − 2.665 | (1.007) | *** | − 0.976 | (0.397) | ** | − 0.035 | (0.018) | * |
| Graduate school | − 5.097 | (1.682) | *** | − 2.248 | (0.710) | *** | − 0.038 | (0.032) | |
| Executives | 0.962 | (1.679) | 0.906 | (0.776) | − 0.070 | (0.033) | ** | ||
| Self-employed and family workers | − 3.695 | (1.332) | *** | − 0.753 | (0.550) | 0.060 | (0.026) | ** | |
| Non-standard employee | − 6.158 | (1.203) | *** | − 1.488 | (0.455) | *** | 0.105 | (0.021) | *** |
| Construction | 0.725 | (1.757) | 0.086 | (0.729) | − 0.032 | (0.035) | |||
| Information and communications | − 2.223 | (1.874) | − 1.162 | (0.751) | 0.019 | (0.039) | |||
| Transport | 1.955 | (2.330) | 1.106 | (0.988) | 0.031 | (0.041) | |||
| Wholesale and retail | − 2.402 | (1.603) | − 0.624 | (0.633) | 0.098 | (0.031) | *** | ||
| Finance and insurance | 6.164 | (2.420) | ** | 2.728 | (1.001) | *** | − 0.004 | (0.042) | |
| Real estate | 0.565 | (2.450) | − 0.239 | (0.884) | 0.099 | (0.047) | ** | ||
| Accommodations and restaurants | 0.357 | (2.563) | 0.365 | (0.977) | 0.069 | (0.049) | |||
| Health care and welfare | 4.961 | (1.630) | *** | 2.418 | (0.664) | *** | 0.002 | (0.030) | |
| Education | 3.073 | (1.962) | 2.157 | (0.832) | ** | 0.031 | (0.037) | ||
| Other services | 0.171 | (1.465) | 0.123 | (0.586) | 0.043 | (0.029) | |||
| Public services | 14.746 | (2.731) | *** | 6.129 | (1.131) | *** | − 0.119 | (0.042) | *** |
| Other industries | − 2.116 | (1.593) | − 0.358 | (0.627) | 0.094 | (0.031) | *** | ||
| Managerial | − 0.085 | (1.603) | 0.048 | (0.691) | − 0.098 | (0.029) | *** | ||
| Professional and engineering | − 3.261 | (1.323) | ** | − 1.495 | (0.537) | *** | 0.034 | (0.024) | |
| Sales and trade-related | − 1.530 | (1.504) | − 0.728 | (0.589) | 0.003 | (0.028) | |||
| Service | − 4.021 | (1.525) | *** | − 1.917 | (0.570) | *** | 0.026 | (0.029) | |
| Production | − 5.545 | (2.263) | ** | − 2.086 | (0.941) | ** | 0.203 | (0.039) | *** |
| Other occupations | − 4.563 | (1.325) | *** | − 1.983 | (0.513) | *** | 0.121 | (0.024) | *** |
| 100–299 | 2.454 | (1.207) | ** | 1.097 | (0.481) | ** | − 0.061 | (0.022) | *** |
| 300–499 | 4.301 | (1.662) | ** | 1.838 | (0.674) | *** | − 0.122 | (0.029) | *** |
| 500–999 | 4.273 | (1.612) | *** | 1.630 | (0.658) | ** | − 0.073 | (0.028) | ** |
| 1000 or larger | 6.870 | (1.122) | *** | 2.604 | (0.439) | *** | − 0.132 | (0.019) | *** |
| Government | 4.526 | (2.996) | 1.574 | (1.283) | 0.006 | (0.047) | |||
| Shorter than 35 h | − 0.878 | (1.045) | − 3.823 | (0.345) | *** | 0.034 | (0.019) | * | |
| 43 h or longer | − 1.163 | (0.967) | 2.657 | (0.433) | *** | 0.002 | (0.018) | ||
| Less than 3 million | − 2.036 | (1.081) | * | − 0.681 | (0.444) | 0.079 | (0.019) | *** | |
| 6–9.99 million | − 0.467 | (1.228) | 0.040 | (0.538) | − 0.029 | (0.022) | |||
| 10 million or more | − 1.908 | (1.717) | − 0.367 | (0.800) | − 0.125 | (0.034) | *** | ||
| Nobs | 5700 | 5700 | 5700 | ||||||
| R-squared, pseudo R-squared | 0.0792 | 0.1428 | 0.1018 | ||||||
OLS estimations (columns (1) and (2)) and probit estimation (column (3)) with robust standard errors in parentheses. ***p < 0.01, **p < 0.05, *p < 0.10. The dependent variable of column (3) is the dummy for workers with zero compliance hour. The reference categories for age, education, industry, occupation, firm size, weekly working hours, and annual earnings are age 40 s, high school, manufacturing, clerical occupation, less than 100 employees, 35–42 h, and 3–5.99 million yen, respectively
However, unlike the cross-tabulations for education, the coefficients for university and graduate degrees are negative and statistically significant. In other words, the cross-tabulation results are affected by the fact that employment type, industry, and occupation differed substantially according to educational background. The coefficient for female is marginally significant (at the 10% level) but quantitatively small for the share of compliance-related working hours and is not significantly related to the absolute amount of compliance-related hours.9
Therefore, compliance-related hours (percentages and weekly hours) differ based on worker characteristics. For instance, the mean figures are 25.7% and 10.23 h for men working in large firms in the finance and insurance industry, 17.8% and 6.65 h for men working in small firms in the manufacturing industry, and 16.4% and 5.86 h for women working in small firms in the wholesale and retail industry.
As shown in Table 2, nearly half of the respondents (46.1%) reported that they did not perform such tasks (expressed as “0%”). The percentages by worker characteristics are presented in Column (4) of Table 3. Probit estimation result to explain zero compliance-related hours is reported in Column (3) of "Appendix Table 7". In this estimation, signs of the coefficients (marginal effects) must be interpreted in the opposite manner to Columns (1) and (2). The coefficients for 20–29 years of age, executives, public service, managerial occupations, and larger firms were negative and significant. The coefficients for female, self-employed, non-standard employees, wholesale and retail industry, real estate industry, production occupation, and low wages were positive and significant. These results are generally consistent with those presented in Columns (1) and (2).
Table 4 presents the percentage of total compliance-related working hours to total working hours, aggregated by industry and firm size. Different from the figures presented in Table 3, which show simple averages per worker, the figures in this table are aggregated by industry and firm size categories, with Column (1) weighted by hours worked and Column (2) further weighted by wages. For all industries, 21.7% and 23.3% of the labor input is spent on compliance when weighted by working hours and wages, respectively. By industry, as expected, the percentages of public services are particularly large. Additionally, finance and insurance, followed by medical care and welfare, have large percentages, likely reflecting a large number of industry-specific regulations. For instance, in the finance and insurance industry, more than 30% of the total labor input is devoted to compliance. Conversely, the percentages for wholesale and retail trade, restaurants and accommodations, and other services are relatively small. Even in industries with few industry-specific regulations, the share exceeds 10%, indicating the importance of compliance costs arising from cross-industry social regulations. The differences by firm size are clear. Larger firms tend to spend more working hours complying with rules and regulations, suggesting the influence of size-dependent regulations or that larger firms tend to have stricter internal rules. Whether weights are used do not substantially affect the relative rankings by industry and firm size.
Table 4.
Percentages of compliance-related working hours by industry and firm size
| (1) Weighted by working hours (%) | (2) Weighted by wages (%) | (3) N | |
|---|---|---|---|
| All | 21.7 | 23.3 | 5707 |
| Construction | 20.6 | 23.6 | 324 |
| Manufacturing | 22.2 | 23.1 | 938 |
| Information and communictions | 20.3 | 21.0 | 255 |
| Transport | 22.8 | 25.4 | 233 |
| Wholesale and retail | 17.7 | 19.2 | 592 |
| Finace and insurance | 31.4 | 30.0 | 219 |
| Real estate | 19.6 | 18.7 | 163 |
| Accommodations and restaurants | 17.1 | 20.5 | 162 |
| Health and welfare | 25.3 | 25.7 | 639 |
| Education | 22.4 | 22.4 | 322 |
| Other services | 17.5 | 19.0 | 915 |
| Public services | 38.4 | 38.9 | 346 |
| Other industries | 14.8 | 17.1 | 599 |
| 99 employees or less | 16.9 | 18.2 | 2807 |
| 100–299 employees | 22.3 | 23.2 | 686 |
| 300–499 employees | 24.7 | 24.7 | 329 |
| 500–999 employees | 25.0 | 25.1 | 394 |
| 1,000 or more employees | 26.2 | 27.0 | 1,233 |
| Government | 35.5 | 36.4 | 258 |
Column (2) shows percentages weighted by each worker’s hourly wages
Finally, the productivity implications of the results are discussed. As shown in the first row of Table 4, the share of compliance-related working hours weighted by wages is 23.3%.10 As we measure labor input, the relationship with TFP depends on the labor share (contribution) in production. Assuming a macroeconomic labor share of 2/3, if the working hours for compliance with rules and regulations are halved, which would lead to a reduction in labor input, the impact would be an increase in TFP of approximately 7.8% (23.3% × 0.5 × 2/3).11 As Japan’s current TFP growth rate is approximately 0.4–0.5% per year, the impact is large, being equivalent to 15–20 years of cumulative TFP growth. Although this is only a mechanical estimate, it suggests the potential importance of deregulation and increased efficiency of regulation enforcement as growth policies. This calculation is the impact on the level of TFP and cannot be simply compared due to differences in analytical methods and coverage of rules and regulations, but it is consistent with studies showing that the negative impact of federal regulations on economic growth in the United States is large (e.g., Coffey et al., 2020; Dawson & Seater, 2013).
It should be noted that this estimate deals only with the effect of labor cost reductions. Some firms may invest in equipment and software to comply with rules and regulations that cannot be captured by labor input. In addition, the possibility that rules and regulations may inhibit risk-taking and innovation in firms and their negative impact on productivity through the misallocation of resources are beyond the scope of this calculation. If such channels are included, the effect on productivity would likely be considerably greater.
Conclusion
This study proposed a novel approach to measure compliance costs by focusing on working hours devoted to complying with rules and regulations. It quantitatively estimated compliance costs in Japan based on worker survey data.
The results indicated that the costs of compliance with rules and regulations, including social regulations, covering all sectors of the economy were remarkably high. A sizable proportion of the working hours of high-wage earners was devoted to these tasks, and it is estimated that productivity would increase by approximately 8% if these costs were halved. These results suggest the importance of reducing compliance costs by using strategies such as streamlining social regulations and digitization. Many social regulations aimed to ensure safety, security, and other values that differed from economic efficiency. Therefore, we could not determine whether they should be reduced or eliminated. However, considering the quantitative magnitude of compliance costs, it is necessary to consider a desirable level of regulations under trade-offs.
This study has focused on the negative impact of compliance on productivity. However, rules and regulations may have positive or offsetting impacts on macroeconomic productivity through, such as spillover effect on compliance-related industries, such as law firms, accounting offices, and consultants, or dynamic reallocation of resources from highly regulated to less regulated sectors between and within firms.
The analysis in this study depended on a multiple-choice survey, which has limitations in terms of measurement accuracy. In addition, the industries and occupations of the workers were categorized only at the one-digit classification level. A large-scale survey with more disaggregated information may enable a more detailed analysis of the relationship with productivity. Furthermore, the analysis was limited by its cross-sectional design. It would be useful to conduct future surveys periodically to observe regulatory changes over time. Finally, similar surveys could be conducted in other countries to assess the extent of compliance costs in Japan from an international comparative perspective.
Appendix
Acknowledgements
I am grateful to an anonymous referee for his/her helpful comments and suggestions. I also thank Masahito Ambashi, Kyoji Fukao, Takeo Hoshi, Kaoru Hosono, Tomohiko Inui, Sagiri Kitao, Naomi Kodama, Keisuke Kondo, Yoko Konishi, Masato Mizuno, Daigo Nakata, Toshihiro Okubo, Kazuo Yamaguchi, and Makoto Yano for their comments on earlier versions of the paper. This study is supported by the Japan Society for the Promotion of Science (JSPS) Grants-in-Aid for Scientific Research (20H00071, 21H00720).
Author contributions
I am the sole author of this paper.
Declarations
Conflict of interest
There is no conflict of interests for this paper.
Footnotes
Although not targeting labor market regulation, Hosono et al. (2017) and Hosono et al. (2019) in Japan show that size-dependent policies, which treat large and small firms differently, may have a negative impact on firm growth and productivity.
The 2020 survey was sent via e-mail to 10,041 individuals who participated in the 2017 survey. In the 2017 survey, the sample individuals were randomly chosen from 2.3 million registered monitors of Rakuten Insight, Inc., stratified by gender, age, and region (prefecture), in proportion to the composition of the 2015 Population Census (Statistics Bureau, Ministry of Internal Affairs and Communications).
The target number of additional responses was set at the cell level (gender*age categories) that was proportional to the latest composition of the Japanese population. An invitation e-mail was sent randomly after considering the predicted response rate. If the number of responses fell short of the target at the cell level, additional invitation e-mails were sent until the target number was met.
In the table, survey categories of age (6 categories), employment type (9 categories), industry (14 categories), occupation (13 categories), and firm size (13 categories) are grouped into a small number of categories to enable comparison with the LFS. Among the worker characteristics used in this study, education (5 categories) is not available in the LFS.
In the survey, weekly working hours are classified into twelve categories (from “less than 15 h” to “75 h or more”), which are the same as the categories in the Employment Status Survey (Statistics Bureau, Ministry of Internal Affairs and Communications).
In the survey, annual employment income is classified into 18 categories (from “less than 0.5 million yen” to “20 million yen or more”), which is finer than the 16 categories in the Employment Status Survey.
As the survey includes information on respondents’ residence (prefecture), we compared residents in the Tokyo area (Tokyo, Saitama, Kanagawa, and Chiba prefectures) with those living in other prefectures; however, no significant differences are observed.
Although the figure is not directly comparable due to differences in methods and coverage, it is larger than an estimate of regulatory costs in the United States of approximately 8% of GDP (Crews, 2022).
The coefficients for working hours and annual earnings are insignificant at the 5% level (see Column (1) of "Appendix Table 7").
After excluding government workers, the share is 22.1%.
This estimate is almost perfectly consistent with the productivity effect of halving compliance costs (approximately 8%) based on a survey of Japanese firms (Morikawa, 2019).
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- Aghion, P., Bergeaud, A. & Van Reenen, J. (2021). The impact of regulation on innovation. NBER Working Paper No. 28381.
- Albrizio S, Kozluk T, Zipperer V. Environmental policies and productivity growth: Evidence across industries and firms. Journal of Environmental Economics and Management. 2017;81:209–226. doi: 10.1016/j.jeem.2016.06.002. [DOI] [Google Scholar]
- Al-Ubaydli O, McLaughlin PA. Regdata: A numerical database on industry-specific regulations for all United States industries and federal regulations, 1997–2012. Regulation and Governance. 2017;11(1):109–123. doi: 10.1111/rego.12107. [DOI] [Google Scholar]
- Amoroso S, Martino R. Regulations and technology gap in Europe: The role of firm dynamics. European Economic Review. 2020;129:103351. doi: 10.1016/j.euroecorev.2020.103551. [DOI] [Google Scholar]
- Andrews D, Cingano F. Public policy and resource allocation: Evidence from firms in OECD countries. Economic Policy. 2014;78:255–296. [Google Scholar]
- Bailey JB, Thomas DW. Regulating away competition: The effect of regulation on entrepreneurship and employment. Journal of Regulatory Economics. 2017;52(3):237–254. doi: 10.1007/s11149-017-9343-9. [DOI] [Google Scholar]
- Calomiris, C. W., Mamaysky, H. & Yang, R. (2020). Measuring the cost of regulation: A text-based approach. NBER Working Paper No. 26856.
- Cette G, Lopez J, Mairesse J. Market regulations, prices, and productivity. American Economic Review. 2016;106(5):104–108. doi: 10.1257/aer.p20161025. [DOI] [Google Scholar]
- Cheshire PC, Hilber CAL, Kaplanis I. Land use regulation and productivity - Land matters: Evidence from a UK supermarket chain. Journal of Economic Geography. 2014;15(1):43–73. doi: 10.1093/jeg/lbu007. [DOI] [Google Scholar]
- Cingano F, Leonardi M, Messina J, Pica G. The effects of employment protection legislation and financial market imperfections on investment: Evidence from a firm-level panel of EU countries. Economic Policy. 2010;61:117–163. doi: 10.1111/j.1468-0327.2009.00235.x. [DOI] [Google Scholar]
- Coffey B, McLaughlin PA, Peretto P. The cumulative cost of regulations. Review of Economic Dynamics. 2020;38:1–21. doi: 10.1016/j.red.2020.03.004. [DOI] [Google Scholar]
- Crafts N. Regulation and productivity performance. Oxford Review of Economic Policy. 2006;22(2):186–202. doi: 10.1093/oxrep/grj012. [DOI] [Google Scholar]
- Crews, C. W., Jr. (2022). Ten Thousand Commandments: An Annual Snapshot of the Federal Regulatory State, 2022 Edition. Competitive Enterprise Institute.
- Dawson JW, Seater JJ. Federal regulation and aggregate economic growth. Journal of Economic Growth. 2013;18(2):137–177. doi: 10.1007/s10887-013-9088-y. [DOI] [Google Scholar]
- Égert B. Regulation, institutions, and productivity: New macroeconomic evidence from OECD countries. American Economic Review. 2016;106(5):109–113. doi: 10.1257/aer.p20161026. [DOI] [Google Scholar]
- Feng G, McLaren KR, Yang Ou, Zhang X, Zhao X. The impact of environmental policy stringency on industrial productivity growth: A semi-parametric study of OECD countries. Energy Economics. 2021;100:105320. doi: 10.1016/j.eneco.2021.105320. [DOI] [Google Scholar]
- Garicano L, LeLarge C, Van Reenen J. Firm size distortions and the productivity distribution: Evidence from France. American Economic Review. 2016;106(11):3439–3479. doi: 10.1257/aer.20130232. [DOI] [Google Scholar]
- Goldschlag N, Tabarrok A. Is regulation to blame for the decline in American entrepreneurship? Economic Policy. 2018;93:5–44. doi: 10.1093/epolic/eix019. [DOI] [Google Scholar]
- Gourio F, Roys N. Size-dependent regulations, firm size distribution, and reallocation. Quantitative Economics. 2014;5(2):377–416. doi: 10.3982/QE338. [DOI] [Google Scholar]
- Greenstone, M., List, J. A., & Syverson, C. (2012). The effects of environmental regulation on the competitiveness of U.S. manufacturing. NBER Working Paper No. 18392.
- Gutiérrez, G. & Philippon, T. (2019). The failure of free entry. NBER Working Paper No. 26001.
- Haidar JI. The impact of business regulatory reforms on economic growth. Journal of the Japanese and International Economies. 2012;26(3):285–307. doi: 10.1016/j.jjie.2012.05.004. [DOI] [Google Scholar]
- Herkenhoff KF, Ohanian LE, Prescott EC. Tarnishing the golden and empire states: Land-use restrictions and the U.S. economic slowdown. Journal of Monetary Economics. 2018;93:89–109. doi: 10.1016/j.jmoneco.2017.11.001. [DOI] [Google Scholar]
- Holmes TJ, Schmitz JA., Jr Competition and productivity: A review of evidence. Annual Review of Economics. 2010;2:619–642. doi: 10.1146/annurev.economics.102308.124407. [DOI] [Google Scholar]
- Hosono, K., Takizawa, M., & Tsuru, K. (2017). Size-dependent policy and firm growth. RIETI Discussion Paper 17-E-070.
- Hosono, K., Hotei, M. & Miyakawa, D. (2019). Size-dependent VAT, compliance costs, and firm growth. RIETI Discussion Paper 19-E-041.
- Janßen, R., Kesler, R., Kummer, M. E., & Waldfogel, J. (2022). GDPR and the lost generation of innovative apps. NBER Working Paper No. 30028.
- Jia J, Jin GZ, Wagman L. The short-run effects of the general data protection regulation on technology venture investment. Marketing Science. 2021;40(4):661–684. doi: 10.1287/mksc.2020.1271. [DOI] [Google Scholar]
- Johnson, G. (2022). Economic research on privacy regulation: Lessons from the GDPR and beyond. NBER Working Paper 30705.
- McLaughlin PA, Sherouse O. RegData 2.2: A panel dataset on US federal regulations. Public Choice. 2019;180:43–55. doi: 10.1007/s11127-018-0600-y. [DOI] [Google Scholar]
- Morikawa, M. (2019). Evidence-based Regulatory Reform. RIETI Report, December 2019.
- Nicoletti G, Scarpetta S. Regulation, productivity and growth: OECD evidence. Economic Policy. 2003;18(1):9–72. doi: 10.1111/1468-0327.00102. [DOI] [Google Scholar]
- Winston C. Economic deregulation: Days of reckoning for microeconomists. Journal of Economic Literature. 1993;31(3):1263–1289. [Google Scholar]