Kenneth A. Younge, Ph.D.

Associate Professor, EPFL

Curriculum Vitae

Professor Younge is an applied economist at the Swiss Federal Institute of Technology. He uses big data, econometrics, and machine learning to examine the strategic importance of patent portfolios, financial disclosures, knowledge spillovers, and artificial intelligence. His work has been published by RAND, IEEE, the Review of Economics and Statistics, the Journal of Economics and Management Strategy, the Journal of Economic Behavior and Organization, the Strategic Management Journal, and the National Bureau of Economic Research. He is a past winner of the Best Conference Paper of the Strategic Management Society and the BPS Outstanding Dissertation Award of the Academy of Management.

Professor Younge has co-founded five firms over the course of his career and worked as a Chief Technology Officer, Director of Development, and President. He also is a Director at Quant AI - an advisory firm focused on executive education and strategic consulting in the field of Artificial Intelligence. A recipient of numerous teaching awards, Professor Younge leads training programs and consulting engagements in Europe and the United States.


Review of Economics and Statistics
Forthcoming ( available online under Just Accepted )

Abstract: The United States patent system is unique in that it requires an applicant to cite documents they know to be relevant to the examination of their patent application. Lampe (2012) presents evidence that applicants strategically withhold 21-33% of relevant citations from patent examiners, suggesting that more than one in ten patents are fraudulently obtained. We challenge this view. We examine the institutional details of how courts identify strategic withholding and find that Lampe’s empirical design is inconsistent with both legal standards and standard operating procedures. We then compile a more up-to-date and detailed set of data to reassess the empirical basis for Lampe’s main claim. We find no evidence that applicants withhold citations from examiners.

Joint work with Jeffrey Kuhn and Alan Marco

Review of Economics and Statistics
Vol. 102, No. 3: pp. 569–582 ( 2020 )

Abstract: We investigate the willingness of individuals to persist at exploration in the face of failure. Prior research suggests that the organization's "tolerance for failure" may motivate greater exploration by the individual. Little is known, however, about how individuals persist at exploration in an uncertain environment when confronted by prolonged periods of negative feedback. To examine this question, we design a two-dimensional maze game and run a series of randomized experiments with human subjects in the game. Our results suggest that individuals explore more when they are reminded of the incremental cost of their actions, a result that extends prior research on loss aversion and prospect theory to environments characterized by model uncertainty. In addition, we run simulations based on a model of reinforcement learning, that extend beyond two-period models of decision-making to account for repeated behavior in longer-running, dynamic contexts.

Joint work with Yaroslav Rosokha

The RAND Journal of Economics
Vol. 51, No. 1: pp. 109–132 ( 2020 )

Abstract: Many studies rely on patent citations to measure intellectual heritage and impact. In this article, we show that the nature of patent citations has changed dramatically in recent years. Today, a small minority of patent applications are generating a large majority of patent citations, and the mean technological similarity between citing and cited patents has fallen considerably. We replicate several well-known studies in industrial organization and innovation economics and demonstrate how generalized assumptions about the nature of patent citations have misled the field.

Joint work with Jeffrey Kuhn and Alan Marco

IEEE - 18th International Conference on Machine Learning and Applications
DOI 10.1109/ICMLA.2019.00120 ( 2019 )

Abstract: Automatic measurement of semantic text similarity is an important task in natural language processing. In this paper, we evaluate the performance of different vector space models to perform this task. We address the real-world problem of modeling patent-to-patent similarity and compare TFIDF (and related extensions), topic models (e.g., latent semantic indexing), and neural models (e.g., paragraph vectors). Contrary to expectations, the added computational cost of text embedding methods is justified only when: 1) the target text is condensed; and 2) the similarity comparison is trivial. Otherwise, TFIDF performs surprisingly well in other cases: in particular for longer and more technical texts or for making finer-grained distinctions between nearest neighbors. Unexpectedly, extensions to the TFIDF method, such as adding noun phrases or calculating term weights incrementally, were not helpful in our context.

Joint work with Omid Shahmirzadi and Adam Lugowski

Journal of Economic Behavior & Organization
Vol. 150: pp. 162-181 ( 2018 )

Abstract: While executives play an important role in leading firm innovation, they may economize on efforts to innovate when protected from takeover threat. Middle managers may curtail the rate and scope of innovation when executives are expected to reduce their innovation involvement. We test our prediction by exploiting a natural experiment in Delaware where court rulings increased takeover protection for Delaware firms. Difference-in-differences estimates show that increased takeover protection reduced the rate of innovation by firms, and that it also reduced the scope of innovation across several key dimensions (technological, temporal, organizational, and international). Consistent with our argument, we find that the negative effect of takeover protection on innovation was weaker for larger firms, where innovation decision making authority is more likely to be delegated to middle managers and executive involvement is lower. Finally, we examine the substitutive relationship between competitive pressures from the takeover market and the product market, and find that the negative effect of takeover protection on innovation was stronger for firms facing low competitive pressure from the product market.

Joint work with Tony Tong

Journal of Economics & Management Strategy
Vol. 25: pp. 652-677 ( 2016 )

Abstract: We estimate the firm‐level returns to retaining employees using difference‐in‐differences analysis and a natural experiment where the enforcement of employee noncompete agreements was inadvertently reversed in Michigan. We find that noncompete enforcement boosted the short‐term value of publicly traded companies by approximately 9%. The effect is increasing in local competition and growth opportunities, and offset by patenting.

Joint work with Matt Marx

Strategic Management Journal
Vol. 36: pp. 686-708 ( 2015 )

Abstract: This study draws on strategic factor market theory and argues that acquirers' decisions regarding whether to bid for a firm reflect their expectations about employee departure from the firm post‐acquisition, suggesting a negative relationship between the anticipated employee departure from a firm and the likelihood of the firm becoming an acquisition target. Using a natural experiment and a difference‐in‐differences approach, we find causal evidence that constraints on employee mobility raise the likelihood of a firm becoming an acquisition target. The causal effect is stronger when a firm employs more knowledge workers in its workforce and when it faces greater in‐state competition; by contrast, the effect is weaker when a firm is protected by a stronger intellectual property regime that mitigates the consequences of employee mobility.

Joint work with Tony Tong and Lee Fleming

National Bureau of Economic Research
Adam Jaffe and Ben Jones, editors -- The Changing Frontier: Rethinking Science and Innovation Policy
Chapter 7: pp. 199 - 232 ( 2015 )

Abstract: We document three facts related to innovation and entrepreneurship in renewable energy. Using data from the US Patent and Trademark Office, we first show that patenting in renewable energy remains highly concentrated in a few large energy firms. In 2009, the top 20% firms accounted for over 40% of renewable energy patents in our data. Second, we compare patenting by venture capital-backed startups and incumbent firms. Using a variety of measures, we find that VC-backed startups are engaged in more novel and more highly cited innovations, compared to incumbent firms. Incumbent firms also have a higher share of patents that are completely un-cited or self-cited, suggesting that incumbents are more likely to engage in incremental innovation compared to VC-backed startups. Third, we document a rising share of patenting by startups that coincided with the surge in venture capital finance for renewable energy technologies in the early 2000s. We also point to structural factors about renewable energy that have led the availability of venture capital finance for renewable energy to fall dramatically in recent years, with potential implications for the rate and trajectory of innovation in this sector.

Joint work with Ramana Nanda and Lee Fleming

Wiley Blackwell Outstanding Dissertation Award ( 2013 )

Abstract: In this dissertation, I argue that employee mobility is a key consideration of the firm. Firms often rely on human assets to generate and maintain knowledge. When key individuals depart the firm, they take knowledge with them, potentially undermining the firm or helping competitors. Specifically, I theorize as to how the potential for employee departure affects firm value, and empirically examine my hypotheses in strategy contexts such as M&As, R&D, and equity investment.

Ph.D. Dissertation

National Renewable Energy Laboratory
NREL/TP-­6A20-­50624 ( 2011 )

Abstract: Low-carbon energy innovation is essential to combat climate change, promote economic competitiveness, and achieve energy security. Using U.S. patent data and additional patent-relevant data collected from the Internet, we map the landscape of low-carbon energy innovation in the United States since 1975. We isolate 10,603 renewable and 10,442 traditional energy patents and develop a database thatcharacterizes proxy measures for technical and commercial impact, as measured by patent citations and Web presence, respectively. Regression models and multivariate simulations are used to compare the social, institutional, and geographic drivers of breakthrough clean energy innovation. Results indicate statistically significant effects of social, institutional, and geographic variables ontechnical and commercial impacts of patents and unique innovation trends between different energy technologies. We observe important differences between patent citations and Web presence of licensed and unlicensed patents, indicating the potential utility of using screened Web hits as a measure of commercial importance. We offer hypotheses for these revealed differences and suggest a researchagenda with which to test these hypotheses. These preliminary findings indicate that leveraging empirical insights to better target research expenditures would augment the speed and scale of innovation and deployment of clean energy technologies.

Joint work with Thomas Perry, Mackay Miller, Lee Fleming, and James Newcomb

Working Papers

Abstract: Current measures of patent similarity rely on the manual classification of patents into taxonomies. In this project, we leverage information retrieval theory and Big Data methods to develop a machine-automated measure of patent-to-patent similarity. We validate the measure and demonstrate that it significantly improves upon existing patent classification systems. Moreover, we illustrate how a pairwise similarity comparison of any and every two patents in the USPTO patent space can open new avenues of research in economics, management, and public policy.

Joint work with Jeffrey Kuhn

Abstract: For firms to benefit from scientific research, it is often important for key scientists to remain with the firm. We develop theory and evidence of a “mobility discount” wherein markets value scientific research conducted within the firm, but also discount firm value for the potential of losing key scientists. In a sample of high-tech IPOs, we show that markets discount firm value by up to 12% when firms rely on key scientists. We also show that the mobility discount is greater in states that prohibit the enforcement of non-compete agreements and when science is more important to the firm.

Sole author

Abstract: This paper estimates the causal effect of being a first mover in the patent space by charting invention in a vector space model of technological relatedness and timing. Relative to close second movers, we find that first movers are more likely to prosecute a patent application to issuance and to pursue more follow-on invention. Invention novelty, however, is associated with being less likely to prosecute a patent application to issuance and with less follow-on invention. Moreover, the importance of being a first mover depends on how novel the invention is: the marginal effect of being a first mover increases when pursuing more novel invention, even though the overall likelihood of prosecuting a patent to issuance continues to decline with more novelty. We provide open access to our data for follow-on work by other researchers.

Joint work with Jeffrey Kuhn

Abstract: This article investigates patent citations made to published patent applications. Although citations to patent publications are conceptually indistinguishable from citations to granted patents, they are omitted from all standard measures. We find that publication citations are a large and growing portion of patent citations, and that they differ statistically from citations to granted patents on several important dimensions. We conclude that omitting publication citations is likely to generate biased measures, and that standard measures of patent citations should be corrected. We release our computer code and corrections for future use.

Joint work with Jeffrey Kuhn



MGT 432 - Data Science for Business - Syllabus

MGT 414 - Technology and Innovation Strategy - Syllabus

MGT 618 - Computational Research Methods for Social Sciences - Syllabus


Data Science for Managers - Boot Camp

Data Science for Managers - Fast Track

Previous Courses

Innovation Strategy (Exec Ed)

Technology Strategy (MBA)

Strategic Management (MBA)

Entrepreneurship and Business Plan Preparation (BA)

An award winning teacher in 2011, 2014, 2015, 2016, 2017, and 2020.


The complete code bank for the Data Science for Managers program.

The DSFM Program open sources all of code used in demos, exercises, illustrations, projects, and workshops for free use by the creative commons. You may access, download, clone, edit, reuse, and collaborate on all of the materials in the Code Bank under a free MIT License. Please see the contributing page if you would like to contribute back to the initiative.

A PyPI package to encrypt, decrypt, and automatically pass private code through a public git repository.

Use passcode to encrypt python modules on a development machine, pass the unreadable code through a git repository, pull the code into production, automatically decrypt the code on the other end, and run/reference the code in production. Do all of that with just two lines of code.

A python toolkit to match business names.

Disambiguating and matching business names can be a difficult problem when the same business can have different spellings, abbreviations, and corporate suffixes. This is a perennial problem for researchers and data scientists when there is not a unique identifier across different data sources (or even different tables within the same data source). Moreover, practitioners may want to apply different probabilistic thresholds to change the sensitivity and specificity of the match, depending on the purpose of the match. The Technology and Innovation Strategy lab at EPFL has developed a multipurpose tool to assist with this task. The tool is in beta release - please ask to be added to the BizMatch repository linked above - Access available on request.

A python toolkit to work with text datasets on top of Pandas.

Developed in the Technology and Innovation Strategy lab at EPFL by Jonathan Besomi, Texthero helps you work with text as a tabular dataset. It is simple to learn, requires little knowledge of Natural Language Programming, and works directly on top of Pandas. You can preprocess text data, map it into vectors, and visualize the vector space in just a few lines of code.

All code is provided for commercial and/or non-commercial use, subject to an open source MIT license. No permission is required for use – please just cite the repository.


Patent Citation Similarity Dataset

From: Patent Citations Reexamined (with Jeffrey Kuhn and Alan Marco)

Many studies of innovation rely on patent citations to measure intellectual lineage and impact. To create this dataset, we use a vector space model of patent similarity to compute the technological similarity between each pair of citing-cited patents. The VSM model analyzes the full text of each document to position it as a vector in a vector space that includes more than 700,000 dimensions and then calculates the angular distance between the two vectors. The dataset includes similarity values for all citations made by patents issued between 1976 and 2017 to issued patents or published patent applications.

Download (819 MB)

Patent Citation Timing and Source Dataset

From: Patent Citations Reexamined (with Jeffrey Kuhn and Alan Marco)

Innovation studies frequently distinguish between patent citation submitted by the patent examiner and those submitted by the patent application. However, publicly available citations data is often misleading, for instance by attributing a patent citation to the patent examiner when it was in fact first submitted by the patent application. This dataset uses internal USPTO data to identify the date on which each citation was first submitted as well as the party (examiner or applicant) who first submitted it. The dataset includes observations for citations made by patents issued 2001-2014, although some level of leftward truncation is evident due to limitations in internal data availability at the USPTO.

Download (292 MB)

Patent Families Dataset

From: Patent-to-Patent Similarity: A Vector Space Model (with Jeffrey Kuhn)

Patent applicants frequently file groups of patent applications linked together by priority claims. These priority claims create families of patent applications that share features such as inventors, priority dates, and technical descriptions. By analyzing these linkages, each patent can be assigned a family identifier that it shares with other patents in the same family. This data set includes two levels of family identifiers (clone for near copies, and extended for more attenuated linkages) for each patent issued 2005-2014.

Download (18 MB)

Datasets are Copyright 2017-2021. All datasets are provided for non-commercial use, subject to the Creative Commons Attribution-NonCommercial-NoDerivatives license. No co‑authorship is required to use the data in academic research – please just cite the supporting article.


I work with companies at my lab at EPFL (the Chair for Technology and Innovation Strategy). For more information, please email me: kenneth.younge ~at~

I also consult for companies through Quant AI. For more information, please email me: kyounge ~at~


I'm married, with one wonderful wife, two spunky children, and three crazy dogs. I'm also an avid climber, with ascents on 5.13 rock, WI5 ice, D8 dry tooling, A4 big walls, and adventures in Yosemite, Nepal, Bolivia, Argentina, and Chile. I still love to get out, and up, whenever I can.

I don't have a Facebook page and I don't tweet. My LinkedIn profile has zero connections and you won't find much about me online. Frankly, I'm skeptical that 'social' media is making our lives better. If you would like to connect, just email me and let's have coffee or zoom. Or stop by my office.


Kenneth Younge CV.pdf