Impact of regulation on digital automation in professional services FRANCO, F .*, SALMASO, E. † , SPARACCA, S. † * Nova School of Business and Economics, Portugal † Prometeia S.p.A., Italy
OECD Online workshop on Regulatory barriers to competition in professional services - 18 November 2021
This talk presents the results of the empirical analysis conducted in the report “The impact of regulatory environment on digital automation in professional services”, prepared for the European Commission. European Commission (2021). “The impact of regulatory environment on digital automation in professional services”. DOI: 10.2873/310173. Available at: https://op.europa.eu/en/publication-detail/-/publication/8457941c-974d-11eb-b85c01aa75ed71a1/language-en. Prepared by Prometeia S.p.A., BIP Business Integration Partners – S.p.A., Economics for Policy a knowledge Center of Nova School of Business and Economics Lisboa. The authors thank BIP, especially for its contribution in the conduction of the survey.
OUTLINE
Introduction Definitions The interplay between digital automation and regulation: the channels at play
The survey Firm-level econometric models: Main results Conclusion and policy implications
Impact of regulation on digital automation in professional services | 2
The study
The objective of the study is to assess how regulation affects the digital automation adoption in the selected countries and professional services
12 EUROPEAN COUNTRIES
4 PROFESSIONAL SERVICES SECTORS
Engineering services
IE NL BE
Architectural services
all rights reserved
SE
Legal services
PL DE
FR
HR
RO
IT ES Accounting services
PT
Impact of regulation on digital automation in professional services | 3
Definitions Digital automation | technologies CROSS – SECTOR TECHNOLOGIES
SECTOR - SPECIFIC TECHNOLOGIES
Advanced robotics BIM – related solutions
Blockchain Structural Engineering Software Computer-Aided Design/Drafting 3D (CAD)
Virtual/ Augmented/ Mixed Reality 3D Printing
Machine Learning
Engineering services
all rights reserved
IoT – Internet of Things
Thermography
5G technologies Photogrammetry
Big Data Drone-enabled solutions
Artificial Intelligence Cloud computing
Impact of regulation on digital automation in professional services | 4
Definitions Digital automation | automated activities CROSS – SECTOR ACTIVITIES
SECTOR - SPECIFIC ACTIVITIES
Customer acquisition, promotion, events & networking Litigation, arbitration and in-court representation
Client information on service delivered Participation to tenders (Public Procurement)
Legal advice (e.g. legal support to decisions making)
Customer retention Invoicing and payments Support services (e.g. Administrative, HR, IT, …) Estimating efforts and fees
Legal services
Information gathering (e.g. diligence and/or discovering documents)
all rights reserved
Counterparty negotiation
Compliance check (e.g. on anti-money laundering and privacy matters)
Document redaction (e.g. legal documents, contracts, transactions, wills, etc.)
Impact of regulation on digital automation in professional services | 5
Definitions Regulatory restrictions REGULATORY AREA
Regulatory approach
SPECIFIC RESTRICTION
Reserves of activities Continuous professional development obligations
Qualification requirements Compulsory membership or registration in professional body
Other entry restrictions Limitation of the number of licences granted
all rights reserved
Qualification requirements
Restriction on corporate form
Incompatibilities of activities Restrictions on joint exercise of professions
Exercise requirements
Obligatory professional indemnity insurance Tariff restrictions Advertising restrictions Shareholding requirements and/or voting rights control Impact of regulation on digital automation in professional services | 6
The interplay between digital automation and regulation: the channels at play Five potential channels of impact on digital automation
ACCESS TO FINANCE
SKILLS all rights reserved
COMPETITION
COSTS
SIZE
Impact of regulation on digital automation in professional services | 7
The survey Online questionnaires Online questionnaires’ key features
The online survey was delivered through a cloudbased software Survey
Fill-in period
Languages
8 000+
• English
professionals reached through online surveys*
• French • German
Monkey
• Italian
110+ National and cross-national professional Associations
all rights reserved
Sample
Tool
• Portuguese
5 weeks from July 30th to September 6th 2020
• Spanish
*In the research design participants were not required to answer all questions Impact of regulation on digital automation in professional services | 8
The survey Respondents Overview If digital automation were a journey, where would your practice be on that journey as of today? [% of respondents for each step of the digital automation process]
8%
Sector Distribution
33% 6% «We Have not started yet»
«We have started but we are at the beginning/early stages »
«We have been on the journey for some time, but it is not over yet»
« Not sure »
Total respondents 8 157
«We have almost reached our destination»
all rights reserved
51%
2%
The majority of respondents (92%) state that they have started their journey to digital automation. Among these, 33% say they are only at the beginning while 51% say they have been on the journey for some time.
Impact of regulation on digital automation in professional services | 9
The survey Respondents Overview | digital automation index DIGITAL AUTOMATION INDEX Sector
Revenues
2.36
< 100k €
2.11
all rights reserved
Between 100k € and 500k €
Between 500k € and 1Mio. €
2.10 2.04
Between 1 Mio. € and 2 Mio. €
Between 2 Mio. € and 10 Mio. €
10 Mio €
Accounting
Legal
Engineering
Architectural
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
Analyzing the average value of the index on different dimensions shows that the greater the size and volume of the business, the greater the level of digital automation of professionals. At the sector level, accounting has a higher indicator value.
Impact of regulation on digital automation in professional services | 10
The survey Respondents Overview | Is regulation an obstacle to automation? To what extent the following the rules of access and conduct might constitute an obstacle to digital automation? [average perceived negative impact (out of 3) of regulation on digital automation, for each sector] Shareholding requirements and/or voting rights control 3.5 Tariff restrictions
3
Compulsory membership or registration in professional body
2.5 2
Restrictions on joint exercise of professions
1.5
all rights reserved
Limitation of the number of licenses granted
1
0.5 Continuous professional development obligations
Restriction on corporate form
0
ARCH Reserve of activities
Incompatibilities of activities
ENG ACC LEG
Qualification requirements Multiple answers admitted
Advertising restrictions
Obligatory professional indemnity insurance
The professionals who feel most hampered by the rules of access/conduct are architects and engineers. Accountants are the least hindered, while lawyers suffer from reserve of activities and advertising restrictions
Impact of regulation on digital automation in professional services | 11
Firm-level econometric models Two different (complementary) approach
AUTOMATION CHOICE (binary variable) PROBIT
0 = firm does not adopt digital automation 1 = firm adopts digital automation
EXPLANATORY VARIABLES
DATA SOURCES
REGULATION PERCEPTION VARIABLES Answering the survey question: Regarding the rules of access and conduct for your profession, to what extent the following options might constitute an obstacle to digital automation?
RESULTS FROM THE SURVEY
FIRM CHARACTERISTICS
all rights reserved
ENDOGENOUS VARIABLE
variables about size (e.g. revenues), capabilities, sectoral and geographical dummy variables, and other information from the survey (e.g. financial constraints binding automation decisions, …)
AUTOMATION INTENSITY HECKMAN
The synthetic index elaborated in task 1 was employed
INFRASTRUCTURAL VARIABLES
•
for each firm, information about internet penetration, broadband speed and other infrastructural elements which can affect the adoption of automation technologies
•
the Community survey on ICT usage in enterprises (EUROSTAT) The Community Innovation survey (EUROSTAT).
Impact of regulation on digital automation in professional services | 12
Firm-level econometric model – Probit model Pooled regression Automation choice -0.0847** (0.005)
Revenues
0.137*** (0.000)
Trends of revenues over the last 3 years
0.0283 (0.414)
Access to high speed internet
0.305*** (0.000)
Cost of digital automation (e.g. financial costs, time investment etc.)
-0.144*** (0.000)
Shortage of specific skills (e.g. ICT / Digital)
Limits of the technological infrastructure (e.g. limited access to high speed internet connection)
-0.0828** (0.006)
An increase in the perception of regulation obstacles is negatively related to the decision to engage in digital automation activities
Larger firms are more prone to adopt digital automation. Infrastructural barriers such as access to high speed internet and shortage of specific IT skills prevent firms to adopt
all rights reserved
Regulation
-0.0268 (0.250)
Broadband Speed
0.175 (0.276)
Digital skills
-0.121 (0.070)
Constant
-2.384 (0.745) 2930 Yes Yes
Observations Professional service fixed effects Country fixed effects
Country specific characteristics such as Broadband speed and digital skills available seems to be less related to the choice to adopt digital automation although the direction of the impact suggest there is a weak relationship
p-values in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001
Impact of regulation on digital automation in professional services | 13
Firm-level econometric model – Probit model results By profession and type of regulation Engineers
Accountants
Lawyers
Reserve of activities
-0.140 (0.059)
-0.056 (0.125)
-0.067 (0.076)
-0.246* (0.021)
Qualification requirements
-0.135 (0.057)
-0.047 (0.183)
-0.0848* (0.016)
-0.096 (0.350)
Continuous professional devlopment obligations
-0.086 (0.185)
-0.029 (0.380)
-0.035 (0.341)
-0.132 (0.194)
-0.195** (0.003)
-0.017 (-0.5)76
0.007 (0.843)
-0.057 (0.580)
Limitation of the number of licenses granted
-0.043 (0.516)
-0.044 (0.181)
-0.004 (0.900)
-0.142 (0.170)
Restriction on corporate form
-0.165* (0.018)
0.000 (0.998)
-0.045 (0.226)
0.004 (0.967)
Shareholding requirements and/or voting rights control
-0.149 (0.053)
-0.005 (0.885)
-0.047 (0.216)
-0.016 (0.874)
Restrictions on joint exercise of professions
-0.104 (0.142)
-0.032 (0.357)
-0.009 (0.803)
-0.095 (0.344)
Incompatibilities of activities
-0.139* (0.047)
-0.050 (0.161)
-0.011 (0.765)
-0.087 (0.428)
-0.260*** (0.000)
-0.038 (0.232)
-0.019 (0.593)
-0.059 (0.542)
-0.199** (0.004)
-0.027 (0.408)
0.019 (0.599)
0.023 (0.814)
-0.145* (0.034)
0.054 (0.125)
0.026 (0.481)
0.055 (0.554)
Compulsory membership or registration in professional body
Obligatory professional indemnity insurance Tariff restrictions Advertising restrictions
Reserve of activities appears to be the only regulation which impacts on digital automation across all sectors. The architects surveyed seems to be more sensitive to regulation barriers
While descriptive analysis seemed to indicate that tariff restrictions and limitation of the licenses granted were the most relevant barriers to digital automation, when controlling for structural and infrastructural variables, these appear to be less significant for most of sectors, except for architects
all rights reserved
Architects
Automation choice
Only coefficients and p-values of automation dummy are reported, although the specification employed for each regulation and each sector controls for the same variables as the previous two outlined models
Impact of regulation on digital automation in professional services | 14
Firm-level econometric model – Heckman model results Reserve of activities Regulation index Qualification requirements Regulation index Continuous professional devlopment obligations Regulation index Compulsory membership or registration in professional body Regulation index Limitation of the number of licenses granted Regulation index Restriction on corporate form Regulation index Shareholding requirements and/or voting rights control Regulation index Restrictions on joint exercise of professions Regulation index Incompatibilities of activities Regulation index Obligatory professional indemnity insurance Regulation index Tariff restrictions Regulation index Advertising restrictions Regulation index
Automation intensity Automation choice Automation intensity Automation choice Automation intensity Automation choice Automation intensity Automation choice Automation intensity Automation choice Automation intensity Automation choice Automation intensity Automation choice Automation intensity Automation choice Automation intensity Automation choice Automation intensity Automation choice Automation intensity Automation choice Automation intensity Automation choice
Architects
Engineers
Accountants
Lawyers
-0.0864* (0.026) -0.276** (0.007)
0.012 (0.590) -0.016 (0.733)
-0.007 (0.689) -0.036 (0.449)
0.060 (0.209) -0.103 (0.455)
-0.0803* (0.038) -0.280** (0.006)
0.016 (0.480) -0.019 (0.686)
0.006 (0.734) -0.036 (0.453)
-0.005 (0.923) -0.085 (0.547)
-0.049 (0.177) -0.302** (0.002)
0.013 (0.527) -0.029 (0.539)
0.000 (0.998) -0.042 (0.382)
-0.044 (0.361) -0.085 (0.547)
-0.039 (0.289) -0.286** (0.005)
0.017 (0.406) -0.020 (0.675)
0.014 (0.404) -0.033 (0.488)
-0.001 (0.990) -0.085 (0.547)
-0.034 (0.324) -0.281** (0.006)
0.027 (0.186) -0.006 (0.903)
0.014 (0.390) -0.023 (0.627)
0.002 (0.962) -0.080 (0.572)
0.011 (0.777) -0.262* (0.011)
0.022 (0.328) 0.073 (0.151)
0.003 (0.874) -0.012 (0.810)
0.103* (0.015) -0.072 (0.611)
0.008 (0.859) -0.229* (0.028)
0.010 (0.683) 0.103* (0.047)
0.013 (0.478) -0.007 (0.888)
0.084 (0.066) -0.012 (0.934)
0.001 (0.974) -0.244* (0.018)
0.024 (0.290) 0.065 (0.195)
0.015 (0.365) 0.000 (0.993)
0.076 (0.081) 0.002 (0.989)
0.030 (0.467) -0.222* (0.034)
-0.002 (0.926) 0.086 (0.091)
0.012 (0.469) -0.011 (0.820)
0.096 (0.068) -0.020 (0.891)
-0.002 (0.959) -0.257* (0.013)
0.007 (0.718) 0.013 (0.794)
0.005 (0.771) -0.022 (0.647)
0.026 (0.543) -0.026 (0.859)
0.0452* (0.040) -0.273** (0.008)
0.0507*** (0.000) 0.029 (0.564)
0.0301** (0.007) 0.003 (0.943)
0.011 (0.708) -0.064 (0.657)
-0.011 (0.749) -0.250* (0.016)
0.021 (0.370) 0.063 (0.208)
0.004 (0.813) -0.008 (0.875)
0.051 (0.206) -0.051 (0.717)
Comparing to the results of the probit model, when looking at specific regulations, the Heckman model provides mixed results in terms of the direction and intensity of the relationship. This suggests that the choice of adopting automation is negatively affected by the regulatory obstacles, but the same does not apply to the intensity. Only architects show an overall negative relationship between regulation and intensity
all rights reserved
Automation intensity, by profession and type of regulation
Tariff restrictions in most sectors seemed to be a relevant and is positively associated with higher levels of digital automation. Only coefficients and p-values of automation dummy are reported, although the specification employed for each regulation and each sector controls for the same variables as the previous two outlined models
Impact of regulation on digital automation in professional services | 15
Firm-level econometric model Main findings
Main findings
AUTOMATION CHOICE (binary variable)
Probit model finds that when firms perceive regulation as a greater obstacle, they are less prone to engage in digital automation activities (choose not to start the digitalization process).
AUTOMATION INTENSITY
Heckman model finds that when firms adopt digital automation their perception of regulation increases along with their automation intensity. This is explained by a stronger and more aware perception of the regulatory environment the more the firm is ahead in its automation journey.
OBJECTIVE VS SUBJECTIVE REGULATION Negative all rights reserved
Variable tested
Impact of perceived regulation on automation
Positive
Measures of objective restrictiveness (such as PROSERV or OECD PMR) and regulation perception are not always correlated. Regulatory transparency should be considered as a pre-condition to facilitate digital automation.
Impact of regulation on digital automation in professional services | 16
Summary and policy implications
POLICY IMPLICATIONS
• The survey and interviews conducted has allowed the construction of a solid body of information. 92% of our sample engaged in digital automation activities, reducing service delivery times and improved quality.
• Economies of scale work as an enabler of digital automation. Regulations that end up, even indirectly, limiting businesses in terms of size (by prohibiting access to new markets or certain corporate forms, for instance) should then be seen as discouraging digital automation.
• Regulation restrictiveness is negatively correlated to the adoption of technology: the higher the perception of regulation, the lower the probability that firms will engage in digital automation.
• As for specific forms of regulation, regulatory scenarios depicted in the study call in particular for a thorough review of the scope of reserved activities.
• Presence in the market of online providers of professional services also appears a significant stimulus for the adoption of digital automation
all rights reserved
KEY FINDINGS
• Other factors, more related to the business environment, such as broadband and IT skills availability for instance, are shown to influence firms’ digital strategies.
Impact of regulation on digital automation in professional services | 17
Contacts
Prometeia Elena Salmaso elena.salmaso@prometeia.com
all rights reserved
Nova SBE Francesco Franco ffranco@novasbe.pt
Prometeia Stefano Sparacca stefano.sparacca@prometeia.com
Impact of regulation on digital automation in professional services | 18
Annex
Impact of regulation on digital automation in professional services | 19
CROSS-SECTORAL ANALYSIS Trends of the future Which digital automation trend(s) would you bet on for the future of your business? [for each technology, % of respondents who believe that the given technology will be the future of his business] 60%
51% 50%
40%
31%
32%
5G technologies
Big Data
all rights reserved
40%
27%
30%
21% 20%
16%
16%
Blockchain
IoT - Internet of Things
12% 10%
0%
Advanced robotics
Virtual / Augmented / Mixed Reality
Machine Learning
Artificial Intelligence (AI)
Cloud computing
Multiple answers admitted
The two main technologies that professionals say they would bet on for the future of their business are those related to Cloud computing and Artificial Intelligence (AI)
Impact of regulation on digital automation in professional services | 20
CROSS-SECTORAL ANALYSIS Business model changes Has the adoption of digital automation led to changes in your business model? [% of respondents whose adoption of DA led to BM changes]
28%
If yes, which ones? [For each business model change, % of respondents who adopted that change]
Introduction of new services
79%
New ways of Clients’ retention New ways of Clients’ acquisition
37%
Decommissioning of old services
34%
Changes in the pricing model
27%
More outsourcing to external providers of standardized/repetitive activities
29% Yes No I have not adopted digital automation
21%
Less outsourcing to external providers of standardized/repetitive activities
12%
More outsourcing to external providers of non-standardized activities
10%
Less outsourcing to external providers of non-standardized activities Multiple answers admitted
all rights reserved
43%
38%
8% 0%
20%
40%
60%
80%
100%
The majority of the sample states that the adoption of digital automation has led to changes in the business model. The most common change (79% of cases) regards the introduction of new services
Impact of regulation on digital automation in professional services | 21
CROSS-SECTORAL ANALYSIS Obstacles How do you assess the relevance of the following obstacles could arise when introducing digital automation solutions? [average relevance (out of 5) of each obstacle arisen from digital automation according to respondents]
Cost of digital automation (e.g. financial costs, time investment etc.)
3.64
Shortage of specific skills (e.g. ICT / Digital)
Cumbersome public administration procedures
3.38
Specific features of the service that do not allow for digitalization
3.36
Regulatory constraints (e.g. Data protection, etc.)
3.30
Limits of the technological infrastructure
3.06
Reliability of new technologies (e.g. bias in algorithms, etc.)
2.95
Lack of commitment from personnel / top management
2.95
0.00
all rights reserved
3.51
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
4.50
5.00
The three main obstacles to digital automation according to the respondents are those related to costs, shortage of Digital/ICT skills and public administration procedures
Impact of regulation on digital automation in professional services | 22
Regulation variable Perceived regulation from firms In order to capture the regulation level of each respondent a regulation index has been elaborated reporting the average perceived negative impact of regulations on digital automation
•
Question 32 asking “To what extent the following rules of access and conduct might constitute an obstacle to digital automation?” has been selected as source of information concerning the perceived level of regulation
•
For the calculation of the index, each answer to each of the 12 rules of access and conduct listed has been associated with a parametric value from a minimum of 1 to a maximum of 5
•
For each respondent, an index value is computed. This framework allows for different types of aggregation, for example per sector or company’s size
The index ranges from 1 to 5, where 1 represents “no perceived impact of regulation” and 5 “the highest impact of regulation to digital automation”
Regulation index against PMR
all rights reserved
Methodology
The scatter between the OECD PMR index and Regulation index shows that the subjective perception of regulation by firms and the objective level of regulation do not seem correlated Impact of regulation on digital automation in professional services | 23
Firm-level econometric model – Probit model results
Architects
Engineers
Accountants
Lawyers
Regulation
-0.333*** (0.001)
-0.0493 (0.292)
-0.0466 (0.322)
-0.122 (0.369)
Revenues
0.349** (0.003)
0.120*** (0.000)
0.154*** (0.000)
0.0179 (0.830)
Trends of revenues over the last 3 years
0.258* (0.016)
-0.0613 (0.215)
0.0649 (0.283)
0.251 (0.102)
Access to high speed internet
0.210 (0.296)
0.310*** (0.000)
0.291* (0.028)
0.499 (0.071)
Cost of digital automation (e.g. financial costs, time investment etc.)
-0.148
-0.148**
-0.116*
-0.179
(0.135)
(0.001)
(0.036)
(0.187)
**
0.0232
-0.0456
-0.0707
-0.137
(0.641)
(0.103)
(0.007)
(0.851)
0.0826
-0.0477
-0.0122
-0.175
(0.246)
(0.171)
(0.750)
(0.096)
Broadband Speed
0.0359 (0.079)
0.00228 (0.833)
-0.0138 (0.230)
0.00510 (0.910)
Digital skills
0.0292 (0.091)
0.00959 (0.640)
-0.0143 (0.672)
-0.0482 (0.576)
Constant
-2.621 (0.111) 293 Yes
0.484 (0.604) 1281 Yes
2.007 (0.053) 1155 Yes
2.864 (0.218) 170 Yes
Shortage of specific skills (e.g. ICT / Digital) Limits of the technological infrastructure (e.g. limited access to high speed in
Observations Country fixed effects
Sector-specific investigation allows to confirm the direction of the relationship, which is significantly strong especially for the architects. A drilldown of the specific regulations will provide further information on the relationship under scrutiny
all rights reserved
By profession
Impact of regulation on digital automation in professional services | 24
Firm-level econometric model – Heckman model results Pooled regression Automation intensity Regulation Competition from online professional services providers Trends of revenues over the last 3 years Benefits from the implementation of digital automation solutions (Yes/ No) Shortage of specific skills (e.g. ICT / Digital) Digital skills Constant
1
Automation choice Regulation Revenues Trends of revenues over the last 3 years Access to high speed internet Cost of digital automation (e.g. financial costs, time investment etc.) Shortage of specific skills (e.g. ICT / Digital) Limits of the technological infrastructure (e.g. limited access to high speed in Broadband Speed Constant mills lambda Observations Professional service fixed effects Country fixed effects
0.0676*** (0.000) 0.0380*** (0.000) 0.0709** (0.001) 0.400*** (0.000) 0.0378* (0.035) 0.0399 (0.206) -0.0365 (0.987) -0.0757* (0.012) 0.142*** (0.000) 0.0270 (0.439) 0.295*** (0.000) -0.142*** (0.000) -0.0864** (0.004) -0.0284 (0.225) -0.0266 (0.765) 2.423 (0.687) -0.547*** (0.000) 2890 Yes Yes
Automation choice appears still negative related to regulation. Infrastructural barriers are also confirmed as a disincentive to adopt the investigated technologies.
Looking at the adopters and their degree of automation intensity, it appears that an increase in the perception of regulation obstacles is positively related to the digital automation activities
all rights reserved
2
Presence in the market of online providers of professional services also appears a significant stimulus for the adoption of digital automation
1 First step Heckman model 2 Second step Heckman model Impact of regulation on digital automation in professional services | 25
Firm-level econometric model – Heckman model results By profession Automation intensity Regulation Competition from online professional services providers Trends of revenues over the last 3 years Benefits from the implementation of digital automation solutions (Yes/ No) Shortage of specific skills (e.g. ICT / Digital) Digital skills Constant
1
Automation choice Regulation Revenues Trends of revenues over the last 3 years Access to high speed internet Cost of digital automation (e.g. financial costs, time investment etc.) Shortage of specific skills (e.g. ICT / Digital) Limits of the technological infrastructure (e.g. limited access to high speed in Broadband Speed Constant mills lambda Observations Country fixed effects Dummy Italy
Engineers
Accountants
Lawyers
0.0116 (0.859) 0.0493*
***
0.126 (0.001) 0.0432***
0.0138 (0.545) 0.0289**
0.106 (0.101) 0.0118
(0.021) 0.0696 (0.320) 0.412***
(0.000) 0.0601 (0.118) 0.443***
(0.009) 0.101*** (0.001) 0.392***
(0.669) 0.151 (0.060) 0.360**
(0.000) 0.0143 (0.797) -0.00861 (0.234) 2.632*** (0.000)
(0.000) 0.0539 (0.098) -0.0119 (0.187) 2.439*** (0.000)
(0.000) -0.00685 (0.770) 0.00874 (0.433) 1.489* (0.024)
(0.003) 0.0462 (0.408) -0.0122 (0.518) 2.459 (0.060)
-0.302** (0.002) 0.377** (0.002) 0.242* (0.026) 0.211 (0.300) -0.153
-0.0384 (0.418) 0.127*** (0.000) -0.0582 (0.243) 0.293*** (0.001) -0.147**
-0.0462 (0.327) 0.155*** (0.000) 0.0655 (0.280) 0.288* (0.031) -0.114*
-0.103 (0.455) 0.0192 (0.822) 0.252 (0.103) 0.531 (0.058) -0.180
(0.126) -0.0655 (0.509) 0.0869
(0.002) -0.0727 (0.096) -0.0497
(0.039) -0.135** (0.008) -0.0143
(0.184) 0.0197 (0.874) -0.179
(0.226) -0.126*** (0.000) 4.420 (.)
(0.157) 0.274 (0.975) -13.98 (.)
(0.709) -0.440*** (0.000) 31.31 (.)
(0.089) -0.379*** (0.000) 23.13 (.)
-0.456* (0.040) 289 No Yes
-0.880*** (0.000) 1274 No Yes
0.131 (0.299) 1157 No Yes
-0.233 (0.445) 170 No Yes
The positive relationship between intensity and regulation perception is confirmed significant for the engineers
Perceived competition from online professional service providers appears in most of the services to be an important driver for digital automation intensity
all rights reserved
Architects
2
The Lambda term (the inverse Mill’s ratio) is significant only for architects and engineers, thus we can assume there is a selection bias for these two professional services
1 First step Heckman model 2 Second step Heckman model Impact of regulation on digital automation in professional services | 26