Dr Matthias Qian (钱铖)

How are Startups Shaping the Future of Work? The Role of AI Translators

AI translators — multidisciplinary experts who bridge business and technology expertise — reduce the coordination costs that arise with the difficulties in the communication of hyperspecialized workers who engage in the division of labor to redesign systems of decision making. The organizational inertia of incumbent firms reduces their adoption of AI translators, increasing the risk of failed AI investments and of their creative destruction. This paper asks if AI translators are the basis for the successful entry of new firms and how these VC-funded startups shape the future of work. I identify 14 million AI translator job postings using natural language processing of over one billion task descriptors extracted from the full vacancy text of the near universe of the past decades’ US online job ads. Using a sample of 11,810 venture-capital-funded US startups, I find a positive effect of AI translator use on startup performance, including on successful initial public offerings. These scaled startups rely heavily on AI translators as intermediaries: they post over four times as many AI translator job postings as incumbent firms. The lack of intellectual property protections on the task composition of jobs contributes to strong local knowledge spillover effects that explain the growing importance of AI translators in the labor market. 

Future of Professional Work: Evidence from Legal Jobs in Britain and the United States, with Mari Sako, and Jacopo Attolini, 2022, Journal of Professions and Organizations (best paper award)

This paper examines the impact of digital technology on professional work by combining insights from the future of work debate and the system of professions. With the adoption of digital technology, who ends up undertaking digital tasks depends on the nature of professional jurisdictional control, which we define as a profession’s power to maintain or shift from existing jurisdictional settlements in the face of external disturbances. Protective jurisdictional control implies that the profession engages in full or subordinate jurisdiction, delegating new tasks to subordinate semi-professionals. By contrast, connective jurisdictional control leads them to prefer settlements by division of labor or advisory links, enabling equal-status professions to work together. Using a large database of online job postings by Burning Glass Technologies, we find evidence for this hypothesis. Empirically, we deploy three ways to gauge the nature of professional jurisdictional control: first, by comparing traditional law firms and alternative business structure firms in the UK regulated legal industry; second, by contrasting the US (with protective jurisdictional control) and the UK; and third, by examining the legal sector (in which the legal profession is dominant) and non-legal sectors. Moreover, we find that protective (connective) jurisdictional control is associated with lower (higher) pay premia for new digital skills, consistent with theory. Our findings highlight the importance of the mediating role of professional jurisdictional control to inform the future of work debate.

The Impact of Artificial Intelligence on Venture Capital: A Critical Outlook, with Thomas Hellmann, Robin Gansäuer, and Junida Mulla, Oxford Intersections

This paper examines how the adoption of artificial intelligence (AI) technologies is transforming the venture capital industry. Leveraging interviews with industry practitioners who are actively working on the adoption of AI tools, it develops a critical outlook on the likely changes awaiting the industry. AI tools accelerate the sourcing and due diligence of venture deals. However, the final authority to make investment decisions remains with humans. The paper identifies socially grounded conviction and gut feelings as two key human proficiencies that AI systems struggle to replicate. Relationships embedded in their professional networks also allow venture capitalists to provide value-added support to their founders in ways that cannot be easily replaced by AI systems. As AI systems are adopted industry-wide, they will broaden founders’ access to funding and shift market power away from investors. It is not the AI infrastructure itself, but human proficiencies, such as socially grounded conviction, gut feeling, and networks, that will allow venture capital firms to competitively differentiate themselves.

Flexible Work Arrangements in Low Wage Jobs: Evidence from Job Vacancy Data, with Abi Adams-Prassel, Maria Balgova, and Tom Waters, R&R Review of Economics and Statistics

In this paper, we analyze firm demand for flexible jobs by exploiting the language used to describe work arrangements in job vacancies. We take a supervised machine learning approach to classify the work arrangements described in more than 46 million UK job vacancies. We highlight the existence of very different types of flexibility amongst low and high wage vacancies. Job flexibility at low wages is more likely to be offered alongside a wage-contract that exposes workers to earnings risk, while flexibility at higher wages and in more skilled occupations is more likely to be offered alongside a fixed salary that shields workers from earnings variation. We show that firm demand for flexible work arrangements is partly driven by a desire to reduce labor costs; we find that a large and unexpected change to the minimum wage led to a 7 percentage point increase in the proportion of flexible and non-salaried vacancies at low wages.

A Taxonomy for Technology Venture Ecosystems, with Mari Sako, SSRN working paper

We develop a taxonomy – Oxford Venture Ecosystem Taxonomy (OVET) -- to classify technology startup ventures along nine dimensions: (1) the area of work, (2) purpose of technology use, (3) technology stack, (4) platform business model, (5) type of clients, (6) value capture strategy, (7) founder and funder characteristics, (8) geographical footprint, and (9) funding cycle. This paper provides a theory and method for developing taxonomies, emphasizing the importance of clarifying the purpose for which a taxonomy is used and the determination of the appropriate level of abstraction. This approach is then applied to develop the OVET taxonomy in the context of specific sectors with AI use cases. We illustrate this application in four sectors, namely fintech, healthtech, lawtech, and proptech. In the last section, we discuss how such a taxonomy, enabling classifications and analytics, could provide valuable insights and improve the quality of decision-making by venture founders, investors, policy makers and other stakeholders in the venture ecosystem.

Information Type and the Geography of Price Discovery, with Howard Jones, and Jose Martinez, SSRN working paper

Local (São Paulo) and global (New York) markets contribute significantly to price discovery in dual-listed Brazilian shares, but their contribution varies over time. Local information shares increase by 8.4% on days when a stock experiences a significant idiosyncratic price swing, but do not similarly increase on earnings announcement days or on days when the whole local market experiences a significant price swing, despite an equally large increase in trading. Traders in the local market seem to have an advantage in collecting and processing company-specific unscheduled information, but not widely disseminated scheduled information which affects the company or the whole market.

How does Equity Allocation in University Spinouts affect Fundraising Success? Evidence from the UK, Reject and Resubmit Management Science

There is considerable controversy about the allocation of equity in university spinouts. Founder teams and outside investors frequently criticize universities for taking excessive ownership stakes, weakening entrepreneurial incentives, and making spinouts ‘uninvestable.’ Universities in turn defend their ownership rights in terms of the resources needed to generate the research in the first place. This paper uses detailed data from UK spinouts to assess the impact of university ownership on subsequent fundraising success. Perhaps surprisingly, the data suggests a positive correlation between university stakes and fundraising success, even after controlling for observable characteristics. However, this correlation appears to be partly driven by universities retaining larger stakes in their most promising spinouts. Using an instrumental variable based on the precedence set by prior spinouts within a university, we find some evidence that higher university stakes reduce the likelihood of fundraising success. A 10% larger university stake decreases the probability of raising venture capital on average by 3%. The negative effect is concentrated in less science-intensive spinouts (e.g., IT), and is statistically insignificant in the more science-intensive spinouts (e.g., engineering, or biomedical). Reductions in university stakes are also associated with increases in the spinout rate.

Knowledge Similarity Among Founders and Joiners: Impact on Venture Scaleup in Fintech and Lawtech, with Mari Sako, and Mark Verhagen, SSRN working paper

In what ways does knowledge similarity among co-founders contribute to venture scaleup? This paper addresses this question by taking account of knowledge domains among early employees and in founders’ social networks. We build a theoretical framework to predict which knowledge combinations are likely to lead to venture growth at different stages. We test our theory using a database of 315 fintech and lawtech startups (with a total of 600 founders and 328 early joiners) in three locations (London, New York City, and San Francisco Bay Area) during the period 2009-2020. Our central finding is that venture growth is explained by knowledge-similar founding teams, founders’ social ties to other founders whose knowledge domains are different from their own, and knowledge-dissimilar employees. Moreover, the positive impact of knowledge similarity in founding teams on venture growth is stronger for older ventures than young ventures, and for mature than nascent ecosystems. These findings highlight the importance of simultaneously studying knowledge domains in founding teams, their social networks, and early employees. Just as important is to analyze the impact of knowledge similarity at different stages in the entrepreneurial process from ideas generation, choice of strategy, and strategy implementation.

To name your venture after Oxbridge or not, that is the question! with Thomas Hellmann and Jundia Mulla, Canadian Journal of Economics

This paper examines the significance of including a company’s location of origin in its name, a practice known as using an eponymous location name. Our theory demonstrates that signalling an origin from a prestigious location is an appealing strategy for relatively weaker companies, whereas stronger companies tend to avoid it. We empirically test this theory on a sample of UK university spin-outs from 2010-2021. Eponymous location names are particularly prevalent among spin-outs from Oxford and Cambridge. As predicted by our theory, eponymous location names are negatively related to measures of fundraising success after controlling for spin-out origin.

The association between socioeconomic status and mobility reductions in the early stage of England’s COVID-19 epidemic, with Won Do Lee, Tim Schwanen, 2021, Health and Place

This study uses mobile phone data to examine how socioeconomic status was associated with the extent of mobility reduction during the spring 2020 lockdown in England in a manner that considers both potentially confounding effects and spatial dependency and heterogeneity. It shows that socioeconomic status as approximated through income and occupation was strongly correlated with the extent of mobility reduction. It also demonstrates that the specific nature of the association of socioeconomic status with mobility reduction varied markedly across England. Finally, the analysis suggests that the spatial differentiation in the ability to restrict everyday mobility in response to a national lockdown is an important topic for future research.

7 Strategies for Leading a Crisis-Driven Reorg, with Peter Buchas, Stephen Heidari-Robinson, and Suzanne Heywood, Harvard Business Review

The Covid-19 pandemic has forced countless companies to reorganize at an accelerated pace. To understand what makes a crisis-driven reorganization succeed or fail, we drew on our own 15 years of experience advising companies on organizational change, as well as a database compiled by Quartz Associates and HBR documenting over 2,500 reorganizations. The database shows that crisis-driven reorganizations are a net benefit in just two thirds of cases; 19% actually damage the company, and only 8% fully deliver everything they aim to in the time planned. What can leaders do to increase their chances of success?

Asymptotic Properties of the Gauge of Step-Indicator Saturation, with Bent Nielsen, Econometric Theory

We investigate the asymptotic properties of Step-indicator Saturation which is an algorithm to handle unmodelled location shifts in time series. We consider a stylized version of the algorithm that uses the split-half approach. We present asymptotic convergence and distribution results on the gauge of the algorithm which is the frequency of falsely retained step-indicators when the data generating process has no shifts. The proofs rely on empirical process results of temporal differences of residuals. Our results offer an asymptotic justification to use the gauge in choosing the tuning parameter of this statistical procedure.

Director of Oxford COVID-19 Impact Monitor, a mobility insights provider for the UK during the COVID-19 pandemic, with Adam Saunders. The Telegraph article 1, article 2; The Guardian article; Oxford Uni new article 1, article 2, article 3, article 4; The Daily Mail article, BBC news interview

Co-Investigator on ESRC project on Mapping the LawTech ecosystem, with Mari Sako, Legal IT insider article, Oxford Law article, The Times article, Legal Business article. Contributor to the Oxford-SRA Technology and Innovation in Legal Services report, with Mari Sako and Richard Parnham

Co-author of UK Electric Vehicle and Battery Production Potential to 2040, with David Howey, Adam Saunders, Oxford Uni news article; The Guardian article; the FT article, the Telegraph article

Shortlisting and interview panel member for ESRC-ADR UK No.10 data science fellowships 2021

Big Data and AI elective for the Master of Financial Economics course at the Saïd Business School of the University of Oxford.

Financial Econometrics for six years for the Master of Financial Economics course at the Saïd Business School, University of Oxford. I won the Master of Financial Economics teaching award.

Econometrics for PPE and E&M students at the Department of Economics, University of Oxford. 

Tutoring undergraduates at Lady Margaret Hall and Merton College, University of Oxford, including Nobel laureate Malala Yousafzai. 

Tutoring MBA students as part of the Creative Destruction Lab at the University of Oxford.

Supervising student projects at an innovative new  Law and Computer Science course, University of Oxford, which brings together law and computer science students.