Progress 05/01/18 to 04/30/23
Outputs
Target Audience:Researchers, innovation policy makers, and non-technical audiences. Changes/Problems:The one output the project was unable to deliver is local area statistics on innovation rates, given Census Bureau rejection of our proposal to apply small area estimation techniques to innovation data. What opportunities for training and professional development has the project provided?Postdoctoral Scholars Luyi Han and Zheng Tian continued to learn how to conduct research in a Census Research Data Center. They also continued their professional development in conceptualizing and writing up research results. The presentations and papers described here have provided learning opportunities to a countless number of audience members and readers across multiple disciplines. Similarly, general-audience readers of the news release describing our findings gained new insights into innovation and its determinants. How have the results been disseminated to communities of interest?During this reporting period, we presented our work in nine presentations at the following conferences. Federal Committee on Statistical Methodology Research and Policy Conference (Washington, DC, October 2022) North American Regional Science Council Meetings (Montreal, Canada, November 2022) Western Regional Science Association (Big Island, Hawaii, February 2023) Southern Regional Science Association (Savannah, GA, April 2023) Agricultural and Applied Economics Association (Washington, D.C., July 2023) Work from the project has also been communicated to the Annual Business Survey team comprised of both NCSES and Census Bureau employees. All of the conference presentations above (see Products) were also presented in NCSES dry runs that are a required part of the external presentation clearance process, with invitations sent to Census Bureau employees if the presentation used ABS data. Research on design orientation from the project informed development of the 2022 ABS Design Module. The 2022 ABS is approximately 30 times larger than the 2014 REIS dataset and the Design Module will provide richer data for constructing a more robust design ladder. Findings from the Regression Discontinuity Design study of potential bias in the ABS has informed how modules are ordered in future years of the ABS. The analysis suggested that placing the R&D module after the innovation module as was done in 2018 may be one factor explaining the absence of bias. In 2021 the R&D module was placed before the innovation module. All future years will place the R&D module after the innovation module. Research using the business origination question to identify user entrepreneurs was communicated to the ABS team and resulted in the question being added to the cognitive testing protocol for the 2024 survey. A decision about including the question in the survey will be made in September 2023. A recurring issue with the ABS data has been difficulty differentiating opportunity entrepreneurs having a growth orientation from the much more numerous lifestyle or necessity entrepreneurs that typically do not aggressively pursue growth. Work from the project has also informed international research on rural innovation. One of the co-PIs was invited to participate as the U.S. delegate to an OECD mission studying rural innovation in Canada. Insights from the current project informed the summary report submitted to the OECD team. The OECD will produce a report on Enhancing Innovation in Rural Regions in late 2023. One of the co-PIs is serving on the International Advisory Board to IN SITU: Place-based Innovation of Cultural and Creative Industries in Non-Urban Area, funded by the EU's Horizon Europe programme. The OECD's National Experts on Science and Technology Indicators (NESTI) working party was consulted in putting together a Federal Statistical Research Data Center proposal to produce small area estimates of innovation rates that would allow mapping innovation at the substate level. Although the possibility had been discussed within NESTI, no OECD member countries had ever conducted small area estimation to produce local area innovation statistics. The working party strongly encouraged doing this using U.S. data. Unfortunately, the FSRDC proposal was rejected after Census review of possible disclosure risk. We developed a general-audience-friendly news release describing some of our findings, and published it through Penn State News. Devlin, Kristen. 2023. "Digital Divide Hinders Rural Innovation, Study Shows." Penn State News, June 9, 2023. https://www.psu.edu/news/research/story/digital-divide-hinders-rural-innovation-study-shows/. What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
In the paper entitled "Growth Trajectories of Rural Innovative Firms," we examined the impacts of substantive and incremental innovation activities on U.S. non-metro firms' employment and wage growth. We used the 2014 Rural Establishment Innovation Survey (REIS) and the 2014 Annual Survey of Entrepreneurs (ASE) linked to the 2014-19 Longitudinal Business Database (LBD) to conduct parallel analyses. We found significant and positive contributions of substantive innovation to firm employment and wage growth. This paper was submitted to Research Policy and is under review. In the paper entitled "User Entrepreneurship and Firm Employment Growth," we compared the growth trajectories of firms with various types of entrepreneurship origins using random coefficient and quantile regression models. We found that firms founded by user entrepreneurs had higher employment growth rates in the first five years, but slower growth rates in subsequent years. The manuscript was shared with the ABS team at NCSES and the Economic Reimbursable Surveys Division at the Census Bureau to justify including the business origination question developed in the Kauffman Firm Survey in cognitive testing for the 2024 survey instrument. This paper was submitted to Small Business Economics and is under review. In the paper entitled "Are Microbusiness Innovation Self-Reports in the Annual Business Survey Biased? A Regression Discontinuity Test," we aim to assess the quality of innovation data on microbusinesses in the Annual Business Survey by testing for a potential source of bias identified in earlier studies-- underreporting of innovation in joint innovation-R&D surveys compared to innovation-only surveys. Our RDD paper was presented at the 2022 Federal Committee on Statistical Methodology (FCSM) Research & Policy Conference, the 2023 Western Regional Science Association (WRSA) Annual Meeting, and to the NCSES/Census Bureau ABS team at the National Science Foundation. This paper was submitted to PLoS ONE and is under review. The paper entitled "An Examination of the Informational Value of Self-Reported Innovation Questions" uses a restricted innovation survey designed to differentiate incremental innovators from more far-ranging innovators and compares it to responses in the ASE and the Business R&D and Innovation Survey (BRDIS) to examine the informational value of these positive innovation measures. The analysis examines the association between the incremental innovation measure in the REIS and a measure of the inter-industry buying and selling complexity. A parallel analysis using BRDIS and ASE reveals such an association may vary among surveys, providing additional insight on the informational value of various innovation profiles available in self-reported innovation surveys. The findings enrich the innovation profiles research that has been done by the European Union using the Community Innovation Survey. This paper was published as a CES working paper (see products). The paper entitled "Experimenting in the Cloud: The Digital Divide's Impact on Innovation" is published in Telecommunications Policy (August 2023, 47(7): 102578). We explicitly test for an innovation enabling effect of the cloud using confidential firm-level data from the 2018 Annual Business Survey (ABS). Our findings revealed that firms using cloud services are more inclined to engage in innovation, with the positive effects varying across different types of innovation. Notably, the positive effects were more pronounced in marketing innovation compared to product innovation. Moreover, we conducted sub-sample estimations based on different industries, firms with different employment sizes, and firms located in different rural-urban continuum codes (RUCC) to assess the influence of rurality. Explicit evidence on the effects of the digital divide is provided by a Oaxaca-Blinder decomposition that estimates that up to half the difference in observed marketing innovation rates between urban and rural firms can be attributed to lower broadband availability. The paper was presented at the 69th Annual North American Meetings of the Regional Science Association International in 2022. In the paper entitled "Internationalization of the Rural Nonfarm Economy and the Cloud: Evidence from US Firm-level Export Data," we discuss the factors that are associated with firms' export performance. The Broadband Equity, Access, and Deployment Program's push for universal broadband availability brings both advantages and challenges to rural communities. On one hand, it may lead to increased leakage of local spending to online purchases, but on the other hand, it offers better opportunities to access remote markets. This study utilizes confidential data to investigate how subscription to cloud computer services, which relies on high-speed broadband, impacts export propensity and intensity. Using a Cragg hurdle regression model to address the potential endogeneity due to truncated firms' export values due to selection bias, we find that firms with foreign-born owners, owners with STEM degrees, owners with a bachelor's degree or above, multiple owners, or firms located in counties with higher broadband adoption are more likely to export their products. Our findings also suggest that firms with minority owners or owners with prior military experience are less likely to export. Metro firms are likely to export more than their nonmetro peers. We find negative effects on the cloud and metro interaction term, which indicates the cloud adoption has stronger effects on exports in rural areas. The paper was selected for presentation at the 2023 Agricultural & Applied Economics Association (AAEA) annual meeting. The paper entitled "Grassroots Design Meets Grassroots Innovation: Rural Design Orientation and Firm Performance" has been completed for presentation at a special session of the European Regional Science Association, organized by an EU-funded project to investigate cultural and creative industries in nonurban areas, and papers at the session will be included in a special issue of a regional science journal with a rural focus such as the Journal of Rural Studies. The paper uses the design ladder construct available in the 2014 REIS that characterizes design orientation as belonging to one of 3 rungs: 1) no systematic approach or an ad hoc approach to design; 2) design as a last finish before product launch; or 3) design integrated throughout the product development process. Linking the 2014 REIS to the 2014-2019 Longitudinal Business Database allowed estimating employment growth and wage growth trajectories as a function of design orientation. Quantile regression methods were used to detect the potential effects of the relatively rare design last finish orientation, and even rarer design-integrated orientation. Design orientation has a significant effect on employment growth in higher quantiles but the effects on wage growth were inconclusive. The results suggest that nontechnological innovation is not ignorable, which has been common practice in innovation studies. We also investigated rural-urban differences in export and innovation activities. The study presents statistical evidence on the factors contributing to the urban-rural export and innovation gap. We have identified several urban-specific factors that compensate for a less export-intensive industry mix in these areas. The findings open up intriguing avenues for future research, particularly in investigating whether public policies can play a role in narrowing the urban-rural export and innovation gap. By understanding these dynamics, policymakers can explore potential strategies to promote balanced export growth and economic development across urban and rural regions. We will present two papers from this study at the 70th Annual North American Meetings of the Regional Science Association International in November 2023.
Publications
- Type:
Journal Articles
Status:
Under Review
Year Published:
2024
Citation:
Tian, Zheng, Timothy R Wojan, and Stephan J. Goetz. Growth Trajectories of Rural Innovative Firms. Under review at Research Policy.
- Type:
Journal Articles
Status:
Under Review
Year Published:
2024
Citation:
Tian, Zheng, Luyi Han, Timothy R Wojan, Anil Rupasingha, and Stephan J. Goetz. User Entrepreneurship and Firm Employment Growth. Under review at Small Business Economics.
- Type:
Journal Articles
Status:
Under Review
Year Published:
2024
Citation:
Han, Luyi, Zheng Tian, Stephan J. Goetz, and Timothy R Wojan. Are Microbusiness Innovation Self-Reports in the Annual Business Survey Biased? A Regression Discontinuity Test. Under review at PLoS ONE.
- Type:
Journal Articles
Status:
Other
Year Published:
2024
Citation:
Wojan, Timothy R., Zheng Tian, Luyi Han, and Stephan Goetz. Grassroots Design Meets Grassroots Innovation: Rural Design Orientation and Firm Performance. Most likely target journal is Journal of Rural Studies.
- Type:
Conference Papers and Presentations
Status:
Other
Year Published:
2022
Citation:
Han, Luyi, Stephan J Goetz, and Timothy R Wojan. 2022. Experimenting in the Cloud: The Digital Divides Impact on Innovation. Presented at the North American Regional Science Council Meetings, Montreal, Canada, November 9.
- Type:
Other
Status:
Published
Year Published:
2022
Citation:
Tian, Z., Wojan, T. R., & Goetz, S. J. 2022. An Examination of the Informational Value of Self-Reported Innovation Questions (CES 22-46; CES Working Paper, p. 25). The Center for Economic Studies of the U.S. Census Bureau. https://www.census.gov/library/working-papers/2022/adrm/CES-WP-22-46.html
- Type:
Journal Articles
Status:
Published
Year Published:
2023
Citation:
Han, Luyi, Timothy R. Wojan, and Stephan J. Goetz. 2023. Experimenting in the Cloud: The Digital Divides Impact on Innovation. Telecommunications Policy 47 (7): 102578. https://doi.org/10.1016/j.telpol.2023.102578.
- Type:
Conference Papers and Presentations
Status:
Other
Year Published:
2023
Citation:
Han, Luyi, Zheng Tian, Stephan J. Goetz. 2023. How Export Performance Is Mediated by Varieties of Innovation, Owner Characteristics, and Location. Presented at the 62nd Annual Meeting of the Southern Regional Science Association, Savannah, GA, April 30.
- Type:
Conference Papers and Presentations
Status:
Other
Year Published:
2023
Citation:
Han, Luyi, Zheng Tian, Stephan J. Goetz, and Timothy R. Wojan. 2023. Are Some Innovation Self-Reports in the Annual Business Survey Biased? A Regression Discontinuity Test. Presented at the 62nd Annual Meeting of the Western Regional Science Association, Big Island, HI, USA, February 15.
- Type:
Conference Papers and Presentations
Status:
Other
Year Published:
2022
Citation:
Han, Luyi, Zheng Tian, and Stephan J. Goetz. 2022. Are Some Innovation Self-Reports in the Annual Business Survey Biased? A Regression Discontinuity Test. Presented at the 2022 Federal Committee on Statistical Methodology Research and Policy Conference, Washington, DC, October 25.
- Type:
Conference Papers and Presentations
Status:
Other
Year Published:
2022
Citation:
Han, Luyi, Zheng Tian, Stephan J. Goetz, Anil Rupasingha, and Timothy R. Wojan. 2022. Is User Entrepreneurship More Inclusive Entrepreneurship? Evidence from Two Micro Datasets. Presented at the North American Regional Science Council, Montreal, Canada, November 10.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2023
Citation:
Han, Luyi, Timothy R Wojan, and Stephan J. Goetz. 2023a. Internationalization of the Rural Nonfarm Economy and the Cloud: Evidence from US Firm-Level Export Data. Presented at the Agricultural and Applied Economics Association Annual Meeting, Washington, D.C., July 23-25, 2023.
- Type:
Conference Papers and Presentations
Status:
Other
Year Published:
2023
Citation:
Tian, Zheng, Luyi Han, Timothy R Wojan, and Stephan J. Goetz. 2023. User Entrepreneurship and Firm Employment Growth. Presented at the 62nd Annual Meeting of Southern Regional Science Association, Savannah, GA, March 30-April 1, 2023, Savannah, GA, April 30.
- Type:
Conference Papers and Presentations
Status:
Other
Year Published:
2023
Citation:
Tian, Zheng, Timothy R Wojan, and Stephan J. Goetz. 2023. Growth Trajectories of Rural Innovative Firms. Presented at the 62nd Annual Meeting of Western Regional Science Association, Big Island, HI, USA Feb 15 18, 2023, February 15.
- Type:
Conference Papers and Presentations
Status:
Other
Year Published:
2022
Citation:
Wojan, Timothy R, Zheng Tian, Luyi Han, and Stephan J Goetz. 2022. SAIRE sans Satire? The Promise and Perils of Small Area Innovation Rate Estimation. Presented at the North American Regional Science Council Meetings, Montreal, Canada, November 9.
|
Progress 05/01/21 to 04/30/22
Outputs
Target Audience: Target audience includes other scientists, practitioners, elected officials, leaders of federal, state and local public agencies, international organizations, and other applied decision makers who can benefit from this information. Changes/Problems:Several factors have delayed our progress on the project, as reported in our February 2022 No-Cost Extension request, which was granted. What opportunities for training and professional development has the project provided?We have a new Postdoctoral Scholar, Luyi Han, joined this project. He and another postdoctoral scholar, Zheng Tian, continue to learn how to conduct research in a Census Research Data Center. They also continue their professional development in conceptualizing and writing up research results. How have the results been disseminated to communities of interest?The results of comparing the innovation surveys of REIS, BRDIS, and ASE have just been disclosed by the FSRDC so that they have not been disseminated yet. The results that show the association of the broadband adoption and firm-level innovations are currently under the disclosure process. What do you plan to do during the next reporting period to accomplish the goals?We have the following work to be completed during the next reporting period. First, we will estimate a random coefficients model estimated on the panel of entrepreneurial establishments in the REIS combined with the Longitudinal Business Database (LBD) to test the effect of end user entrepreneurship on employment growth trajectories. Second, We will estimate a survival model to examine whether innovation orientation increases the probability of business survival. Third, we will investigate further whether more innovative firms are more likely to export goods or services and have higher employment and wage growth. Finally, we plan to submit manuscripts to journals such as Plos One, Research Policy, Telecommunications Policy, etc. A proposal has been submitted to the FSRDC to produce the final two deliverables in the AFRI proposal. This research will examine how process and product innovation are associated with productivity and employment growth at the firm level by merging Census and external innovation datasets with the 2012 and 2017 Economic Census. The merging of these data will also allow testing the feasibility of small area estimation applied to innovation data. The inaugural 2018 Annual Business Survey is the largest innovation dataset ever collected but is still too small to produce reliable innovation rate estimates for substate geographies. Auxiliary variables in the Economic Census will be used to model innovation behavior for firms not included in ABS, which will be combined with observed innovation data in ABS to produce composite innovation rate estimates at the county level. If small area innovation rate estimates are demonstrated to be reliable, future research could then examine the potentially countervailing effects of product and process innovation at the meso-level corresponding to commuting zone, core based statistical area, or development district.
Impacts
What was accomplished under these goals?
Drs. Tian and Wojan have completed the analysis that compares survey responses of the Rural Establishment Innovation Survey (REIS), the Annual Survey of Entrepreneurs (ASE), and the Business R&D and Innovation Survey (BRDIS) with respect to self-reported innovation activity. We encountered challenges in linking these three surveys due to their low joint sampling probability, but Drs. Tian and Wojan have addressed this problem so that we can now examine the differences in self-reported innovation activity revealed by these surveys and their association with the industry-level latent innovation index. The results of this analysis have been approved by the Census Bureau for disclosure, and we are writing a working paper that will be submitted to an academic journal, e.g. Plos One or Research Policy. Drs. Tian and Wojan continue to complete a sub-objective of this project that examines the distinct effect of substantive and incremental innovation on the growth of employees and wages and survivability of firms. We have included a new dataset, the Annual Business Survey (ABS), into this project. The ABS incorporates the innovation and entrepreneurship surveys of the ASE and BRDIS that have been discontinued by the Census Bureau. One feature of the 2018 ABS is that firms with 1-9 employees receive a long survey form that includes both R&D- and innovation-related questions, while firms with more than 9 employees receive a short form that includes only innovation-related questions. The originally proposed analysis to test for the existence of underreporting of innovation in joint innovation/R&D surveys (BRDIS) relative to innovation only survey (REIS and ASE) can now be performed within a single survey as a quasi-experimental test. Drs. Han, Wojan, and Tian take advantage of this survey design and use the regression discontinuity design (RDD) to examine if firms, at the threshold of 9-10 employee, would under-report innovation activities when they receive the R&D module. An abstract has been submitted for presentation of the findings at the Federal Committee on Statistical Methodology Research and Policy conference held in Washington, DC in October. We have obtained some results to test our hypothesis. We plan to submit our results for disclosure by the end of April 2022 or the early May. Drs. Han and Wojan have also made progress on another sub-project. In particular, we linked the Longitudinal Firm Trade Transactions Database (LFTTD) with the REIS data, and we have generated preliminary regression results. We also linked the Annual Business Survey (ABS) and the Federal Communications Commission's (FCC) broadband data and have obtained early results to address FCC broadband quality issues. Using clouding computing variables from the ABS data as a reliable measurement of the adoption of broadband, we have generated robust estimates that show how the adoption of broadband is associated with firm-level innovation. The results are currently under the disclosure process from the Census Bureau. We expect that the disclosed results will be ready by May 2022, and we plan to submit our findings to an academic journal like Telecommunications Policy. A background working paper on digitalization, cloud computing and innovation is under review at the National Center for Science and Engineering Statistics, containing descriptive statistics. The decision to release as a NCSES working paper was driven by the limited number of estimates an FSRDC project can submit for disclosure avoidance clearance.
Publications
- Type:
Other
Status:
Awaiting Publication
Year Published:
2022
Citation:
Wojan, TR and National Center for Science and Engineering Statistics (forthcoming) Digitalization, Cloud Computing, and Innovation in U.S. Businesses. Working Paper NCSES 22-2XX. Alexandria, VA: National Science Foundation. Available at https://www.nsf.gov/statistics/2022/ncses222XX/.
|
Progress 05/01/20 to 04/30/21
Outputs
Target Audience:Target audience includes other scientists, practitioners, elected officials, leaders of federal, state and local public agencies, international organizations, and other applied decision makers who can benefit from this information. Changes/Problems:As noted previously, we had the problem of small joint sample selection probability when trying to link the REIS dataset with BRDIS and ASE. Therefore, we changed our approach to examining how the association of various types of innovators--substantive innovators, incremental, or non-innovators-- and the latent innovation measure by industry would change if using different survey data. What opportunities for training and professional development has the project provided?The postdoctoral student, Zheng Tian, continues to learn how to conduct research in a Census Research Data Center. He also continues his professional development in conceptualizing and writing up research results. How have the results been disseminated to communities of interest?
Nothing Reported
What do you plan to do during the next reporting period to accomplish the goals?We are going to export the current results from the RDC and publish a paper based on these results that demonstrate differences in normative and positive measures of innovation. This paper will address the measurement issues that were a principal objective of the Census Federal Statistical Research Data Center (FSRDC) proposal. The joint probability of selection problem will not affect linking the datasets to the LBD that is a census of employer businesses. This will allow examining differences in how the various innovation measures are associated with firm performance through time. The next research proposal to the FSRDC to extend the analysis to export datasets has been favorably reviewed by Census and is now awaiting final Title 26 approval from the IRS.
Impacts
What was accomplished under these goals?
Initial attempts to link REIS to BRDIS and ASE resulted in a small number of matches. We examined the joint sample selection probability of the REIS, BRDIS, and ASE datasets to examine whether our matching algorithms were ineffective or whether there were only a small number of possible matches. We found that the probability of the same establishments being selected in the REIS dataset and in either the BRDIS or the ASE dataset is very low based on the probability of selection variable available for each observation. Merging these datasets as we proposed would result in a very small sample size unlikely to support meaningful analysis of simple response variance or reliable descriptive statistics. Therefore, we decided not to proceed with the original plan but find a work-around solution. The solution is to run similar regression models with the REIS, BRDIS, and ASE datasets separately and Anchorexamine how the association of various types of innovators--substantive innovators, incremental, or non-innovators-- and the latent innovation measure by industry would change if using different survey data. What we have done includes the following: A working paper by Tim Wojan, "Innovation versus Continuous Improvement: Using Secondary Input-Output Data to Examine the Difference", uses regression analysis of REIS data to benchmark the ensuing analysis with the BRDIS and ASE data. The results of this working paper show that the measure of latent innovation is positively associated with "incremental innovators" but negatively associated with both "substantive innovators" and "non-innovators". With the BRDIS data, we defined "incremental innovators" as having new product to the company, "substantive innovators" as having new product to the market, and "non-innovator" as no response to either of both questions. Further, we defined three finer and mutually exclusive categories, "new product to company only", "new product to market only", and "new product to both company and market". The regressions with the BRDIS data give us different results than those in the working paper. With all firms in the BRDIS dataset, the coefficients on the latent innovation measure for both incremental and substantive innovators are negative and positive for non-innovators. When excluding companies with no innovation, incremental innovators are positively associated with the latent innovation measure, while substantive innovators are negatively associated, which is consistent with the working paper. With the ASE dataset we are only able to examine differences between product innovation and process innovation, which was also done in the REIS working paper. The results show that coefficients on the latent innovation measure for companies with various types of product innovation are significantly negative. In contrast, the coefficients for companies with various types of process innovation are much smaller in magnitude than those for product innovation and insignificant or positively significant. These results are generally consistent with the REIS working paper.
Publications
|
Progress 05/01/19 to 04/30/20
Outputs
Target Audience:Target audience includes other scientists, practitioners, elected officials, leaders of federal, state and local public agencies,international organizations, and other applied decision makers who can benefit from this information. Changes/Problems:We are currently delayed because we cannot access the federal data center, as a result of the COVID-19 shutdown. What opportunities for training and professional development has the project provided?As noted, two post-docs are currently being trained and developing their professional skills in analyzing complex data sets. How have the results been disseminated to communities of interest?
Nothing Reported
What do you plan to do during the next reporting period to accomplish the goals?We are doing what we can to advance the project within the constraint of not being able to access the data lab. We are working on the next research project application(s).
Impacts
What was accomplished under these goals?
We have made significant progress (although it has taken longer than initially planned) in securing Special Sworn Status for postdocs Meadowcroft and Zheng. They have been able to access various Census files and started to merge data from different surveys. Together they were starting to explore the merging of BRDIS and REIS before the COVID-19-related access restrictions prevented them from visiting the secure lab. We also met as a team with our DC-based collaborator to discuss next steps on the project. External resources to illuminate the possible innovation self-reporting differences between BRDIS and ASE included review of a comprehensive cognitive testing report conducted by OECD (Peric and Galindo-Rueda 2014). The report found that American respondents generally did not perceive "significant improvement" as qualifying as "innovation" despite the self-reported innovation question asking about "new or significantly improved" products and processes. This might create a downward bias in innovation self-reports in BRDIS relative to ASE given the detailed questions on R&D and scientific and engineering employees in BRDIS that may frame a more stringent construct of innovation for respondents. To investigate this possibility empirically, REIS was used to examine correlates with the incremental innovator classification. A measure of latent innovation (Goetz and Han 2020), derived from the purchasing and selling complexity of detailed industries, was found to be positively associated with the incremental innovator classification but negatively associated with both non-innovators and substantive innovators. The latent innovation measure may be associated with both greater opportunity and greater capability for continuous improvement, explaining its strong association with incremental innovation. The latent innovation measure may provide an explanation for why innovation self-reports in ASE contradict negative innovation self-reports in BRDIS. Examination of preliminary findings from the 2018 Annual Business Survey reinforces hypotheses regarding the greater prevalence of women and minorities in the ranks of user entrepreneurs to be investigated by linking REIS to SBO and ASE. Despite lower overall rates of entrepreneurship relative to majority male business owners, female- and minority-owned businesses demonstrated significantly higher rates of self-reported innovation. The possibility that this innovation arises from own-use is suggested by the significantly lower incidence of formal R&D activities among female- and minority-owned innovative businesses relative to their male majority-owned innovative business peers. The linking of REIS to SBO and ASE will allow empirical confirmation of these suggestive results. References cited: Goetz, Stephan J. and Han, Yicheol (2020) "Latent Innovation in Local Economies," Research Policy 49:103909. Peric, Susan and Galindo-Rueda, Fernando (2014) "The Cognitive Testing of Innovation Survey Concepts, Definitions and Questions," Final report of Improving the Measurement of Innovation: Supporting International Comparisons project. Paris: OECD Directorate of Trade and Industry.
Publications
|