Week 4a: Questions - Who becomes an inventor, scientist, or entrepreneur?

[jamesallenevans]

Post your (<150 word) question for class about or inspired by the following readings:

• “Tapping into Talent: Coupling Education and Innovation Policies for Economic Growth,” 2024. Ufuk Akcigit, Jeremy Pearce, and Marta Prato. Review of Economic Studies, forthcoming.
• “Brains and Business: How Inventors and Entrepreneurs Shape Economic Progress,” 2025. Ufuk Akcigit, Harun Alp, Jeremy Pearce, and Marta Prato. Working Paper.
• “Measuring the Characteristics and Employment Dynamics of U.S. Inventors,” Ufuk Akcigit and Nathan Goldschlag, Journal of Economic Growth, forthcoming.

[willowzhu]

Measuring the Characteristics and Employment Dynamics of US inventors
Inventors are more likely to be found at incumbents, and are less likely to switch jobs. How do we close the gap of innovation between entrants and incumbents? Is the clustering of big innovations within big firms contributing to the decline in dynamism in the United States?

Furthermore, what are some policies that we can implement to diversify inventor demographics? We discussed in class that it is necessary for us to make space for new ideas, and we can see from past studies that inventor success and social status is correlated. How do we grapple with these complex and interconnected problems through policies?

[JaslinAg]

Education is inherently unequal, richer communities have better schools and educational resources. As there are limited slots in PhD (and similar educational programs), rich parents are incentivized to hoard these resources and provide the best education for their children. The readings showed that a higher average innate ability in PhD slots would increase innovation, and thus economic growth. When discussing education being more equitable, most people jump to standardized testing as a “fair” way to allocate prestigious educational slots. Many standardized tests were designed around this desire - to segregate and prevent a more equal allocation of education. All current measures of IQ are socioeconomically biased. By the time, a child is school-aged, their socio-economic circumstances have affected their innate ability.

What are ways to measure innate ability? Can it be separated from socio-economic circumstances? Knowing this, what policies can be implemented to ensure equitable access to education?

[aveeshagandhi]

“Tapping into Talent: Coupling Education and Innovation Policies for Economic Growth”:
The authors of the text place emphasis on aligning education policies with more innovative approaches or strategies. This makes me think about how policymakers design education systems that are not only adept at enhancing technical skills but also cultivating creativity and interdisciplinary problem-solving. How can this be achieved? This thought stems from the reading's argument, which explores how targeted education policies are an impetus for long-term growth.

“Brains and Business: How Inventors and Entrepreneurs Shape Economic Progress” underscores the interactions between entrepreneurs and investors in their endeavors towards driving economic development. What market mechanisms / institutional methods best facilitate this collaboration, especially within Automation-driven fields?

[mskim127]

Implications of a loan market on the modeling choices in "Tapping into Talent":
How would the inclusion of a borrowing market into the model affect its conclusions on the balance between educational subsidies and R&D subsidies? On the one hand, if individuals were allowed to balance their income inter-temporally, then potentially those willing and able to conduct research can borrow against future income to become PHDs? I would expect this change to shift policy recommendations in favor of R&D subsidization as it would seemingly reduce the inefficient allocation of labor caused by financial restrictions. On the other hand, the model appears to perform well in the context of data collected in a market where loans are available and borrowing is prevalent. Could this suggest that student loans are not providing the flexibility they are expected to provide considering market behavior can be sufficiently captured by a model that considers it non-existent?

[jessiezhang39]

The paper "Tapping into Talent" highlights that R&D subsidies yield immediate yet limited growth impacts due to constraints like "time-to-build" human capital, while education policies foster more substantial long-term growth. Specifically, the model finds that education policies surpass R&D subsidies in effectiveness approximately five years after implementation, as educational investments take time to nurture talent and integrate it into the innovation ecosystem​. However, this leaves room for further inquiry: how should policymakers balance short-term innovation needs with the extended timeline required for educational policies to deliver results?

Furthermore, the paper has not addressed whether these subsidies might have diminishing returns in societies with lower inequality or how the optimal policy mix shifts as inequality fluctuates. A blended policy calibrated to inequality levels could maximize outcomes, but questions remain about implementation: how should governments allocate limited budgets across education subsidies, R&D incentives, and PhD slot expansion while holistically factoring in their respective time horizons and complementarities?

[yangkev03]

In "Measuring the Characteristics and Employment Dynamics of U.S. Inventors", we learn of various demographic and employment characteristics of inventors in the U.S. Specifically, by matching inventors to inventions through Protected Identification Keys and using LEHD data to understand the employment history of inventors, the paper finds that the most productive inventors work at older firms, while inventors at young firms produce more impactful patents. To me, this seems like an apparent contradiction in that the impact of patents seems to lose their relationship to patent productivity. One reason for this may be the fact that after producing impactful patents, inventors tend to leave younger firms for larger, older firms. Using this chain of thought, I was wondering why inventors would prefer to work at large, older firms. Some possibilities may be that inventor earnings are closely tied to inventive productivity. Therefore, as inventors create impactful patents, they may be compensated more attractively at larger, established firms. Additionally, large firms may be better suited to support the patenting process and provide resources for greater amounts of R&D.

[jesseli0]

In "Tapping into Talent: Coupling Education and Innovation Policies for Economic Growth," inherent distaste for research is given as the main reason why many unconstrained but capable people do not go into research. This preference is treated as somewhat rare and immutable. Thus, we can only get more PhD students through uncapping financial restraints. While it is important that every person with a preference for research is not financially constrained, this might not be the only way we can get more researchers. While we might not be able to alter the preferences of the individuals, their decision might not be exogenous to the environment of academia. For example, one's decision to pursue graduate studies could be improved if they had excellent professors or a generally good time pursuing academics in undergraduate. If academia paid better and had an easier path to career progression, then it might also attract more students. Is there policy we can consider to increase the "quality" of higher education, and can we endogenize it into the model?

[cskoshi]

In the "Tapping Into Talent" Paper, the model assumes that people who go into PhDs become either team members or team leaders in research. This assumption seems crucial to the link to the net quantity of ideas, q , as outlined on page 12. My question is, how would the model function if we factored in the reality that some PhDs either drop out early or do not pursue research full time. I do see it possible to perhaps factor this into the fraction term Φ, but if not, how would this impact our policy recommendations. If the true net contribution to growth, q, is less than this model predicts, how would this affect the cost-benefit analysis that governments take in deciding how much subsidy to provide.

Also, it might be interesting to look into the issue of policy credibility. The model relies on people's expectation of future income which while the government can influence through policy, is also affected by how the people perceive the policy. More specifically, lets say the government, in a bid to induce more skilled workers to take up PhDs, temporarily increases the R&D subsidy, hence giving people the impression that their future profits would increase. But then, they take it away once the economy is back to health. Would this reduce the effectiveness of future R&D policies as people become skeptical of the permanency of these measures. So, when future R&D policies get enacted, perhaps people's expectations of future income do not change as much as intended. In other words, what would a realistic and credible R&D subsidy look like?

[nsun25]

In “Measuring the Characteristics and Employment Dynamics of U.S. Inventors,” Ufuk Akcigit and Nathan Goldschlag, Journal of Economic Growth, forthcoming, the authors examine the characteristics, employment dynamics, and economic contributions of inventors by linking patent data with U.S. Census microdata. We learn about racial and gender disparities among inventors' demographics. My question is how could targeted policy interventions address these disparities? How can policymakers use these findings to design programs that promote innovation, particularly in regions or demographics with lower representation?
And in terms of our current political climate, how might Trump’s recent policies, such as the reduction of DEI initiatives, stricter immigration rules, increased AI investments, and deregulation, affect inventor demographics, employment dynamics, and the broader innovation ecosystem?

[e-uwatse-12]

From, Measuring the Characteristics and Employment Dynamics of U.S. Inventors by Prof. Akcigit and Nathan Goldschlag, a natural next question from the article is how do the interactions between R&D subsidies and higher education policy shape not only aggregate productivity in the U.S but also the distribution of opportunities across socio-economic groups, and to what extent can these policies address systemic inequalities in the access to innovation-driven career paths? Furthermore, How does the long-term effects of education policy on talent development and career choice differ in societies with varying levels of income inequality

[kbarbarossa]

Given the evidence in the reading that expanding PhD slots increases the number of researchers but also leads to a decline in the average talent of PhD enrollees, how does this approach compare to using financial aid to address this issue of talent misallocation when it comes to underprivileged students? We know that students with wealthier parents, on average, seem to have higher IQs, but how does the availability of financial aid in the PHD sense at an institution affect the talent that the individual institution attracts? Does expanding slots attract a more diverse group of students, including those with higher emotional intelligence who value inclusivity, embrace different perspectives, and are drawn to risk-taking or unconventional experiences? Or does this have the same effect of decreasing the average IQ of the admitted student body?

[chrislowzhengxi]

In "Measuring the Characteristics and Employment Dynamics of U.S. Inventors", the article mentions a decline in inventors working at young firms, even though young firms seem to produce more impactful patents. Instead, inventors are mostly concentrated in older, larger firms, which might have policies or structures that hinder innovation. At the same time, there’s a clear underrepresentation of minorities, like women and African Americans, among inventors. What policies could address these issues? For instance, could employment subsidies or funding specifically for young firms encourage inventors to join them? Similarly, could initiatives like mentorship programs, targeted grants, or improved STEM education for underrepresented groups help increase diversity in the inventor pool? What about improving innovation or impact of inventions in the already-established large firms? How can we find a balance between supporting young firms and increasing diversity to maximize innovation?

[siyakalra830]

"Measuring the Characteristics and Employment Dynamics of U.S. Inventors" reveals that inventor earnings are highly skewed, with a large percentage of both all inventors and "superstar" inventors concentrated in the top decile and even the top one percent of the earnings distribution. Given that this earning potential is closely tied to inventive productivity as measured by citation-weighted patent grants, to what extent do the current compensation structures and incentive systems within firms, particularly older, larger firms where most inventors are employed, contribute to the observed decline in inventor dynamism (such as decreased job switching, entrepreneurship, and geographic mobility) and how might alternative compensation and incentive structures impact the allocation of inventive talent and the overall rate of innovation in the U.S. economy?

[sijia23333]

In "Tapping into Talent: Coupling Education and Innovation Policies for Economic Growth," the authors reveal important insights about how different policies shape talent allocation and innovation. The paper finds that R&D subsidies are less effective than standard models predict because they cannot reach talented but financially constrained individuals. Meanwhile, education policies can strengthen innovation by enabling talented but constrained individuals to enter research, though these effects take longer to materialize. Notably, the policies work better in combination - for instance, with a 2.5% GDP budget, the optimal mix includes both R&D subsidies (58%) and education support (42% total between subsidies and slots).
The paper also reveals that education policies are particularly effective in more unequal societies where financial constraints are more binding. This raises interesting questions about policy design across different contexts. How should the optimal policy mix shift as countries develop and inequality changes? Given the finding that R&D subsidies have faster effects while education policies have larger long-term impacts, how should policymakers balance short-term and long-term priorities? Additionally, since the paper finds that talent is relatively scarce and local, what implications might this have for immigration policy as a complementary tool for expanding the researcher pool?

[vmittal27]

In Measuring the Characteristics and Employment Dynamics of U.S. Inventors, the authors use patent data to perform analysis on U.S. inventors and related characteristics. However, we've learned that not all innovation results in patents. For example, Coca-Cola purposely chooses not to patent its recipe, choosing to protect it as a trade secret instead. Other firms may not apply for patents due to application and defense costs as well (which is likely a concern for small firms rather than large ones)

Knowing this, how might we correct for any bias we incur by measuring innovation solely through patents? I would imagine that trying to directly measure all sources of innovation isn't possible, since things like trade secrets can't even be verified to be real. However, could there be other proxies or instrumental variables that allow us to better correct for these biases?

[malvarezdemalde]

The article "Measuring the Characteristics and Employment Dynamics of U.S. Inventors" highlights the underrepresentation of women and minorities among inventors, along with the growing dominance of large firms in innovation. "Tapping into Talent: Coupling Education and Innovation Policies for Economic Growth" explores how financial barriers and limited educational access hinder talented individuals from entering innovation-driven fields. What kinds of policies could help address these challenges? Could targeted scholarships for underrepresented groups, paired with incentives for small and medium-sized firms to hire diverse talent, help build a more dynamic inventor pool? Additionally, how does the balance between diversity, equity, and inclusion initiatives and maintaining perceptions of meritocracy affect the design and acceptance of these policies? Lastly, how effective have the Pell Grant and similar student financial aid programs been in expanding educational access for underrepresented groups, and could its model be adapted to support careers in innovation-intensive industries?

[Adrianne-Li]

The readings emphasize the complementarity of education and innovation policies in fostering long-term economic growth, particularly through the development and integration of human capital into the innovation ecosystem. "Tapping into Talent: Coupling Education and Innovation Policies for Economic Growth" demonstrates that while education policies address structural constraints and promote long-term innovation by nurturing talent, R&D subsidies yield faster, short-term results. Considering the varying effectiveness of these policies across different socio-economic contexts, how should policymakers balance immediate innovation demands with the extended timelines required for educational reforms, especially in regions with high inequality? Furthermore, how can public-private partnerships be leveraged to align these timelines, ensuring sustainable growth while addressing disparities in access to resources and opportunities for innovation-driven careers?

[pauline196]

Reading “Tapping into talent: coupling education and innovation policies for economic growth”, I had several questions. The paper makes the assumption of a steady state and finds the solution to model, and it makes sense for the developed countries but probably not as much for the developing. How might the removal of the steady-state assumption alter the paper’s conclusions regarding the impact of R&D subsidies, education subsidies and the expansion of educational opportunities?

Other few questions that I had in mind are:

• What frameworks could be used to analyze the allocation (rather than the quantity) of R&D and education subsidies to maximize impact?
• Regarding Figure 9, which illustrates a positive correlation between father’s income and the probability of a child obtaining a PhD, what factors could account for the similar probabilities observed between the 10th poorest percentile and the 50th–60th percentile?

Not directly related to this assignment, but where can I explore more about the calibration techniques and how they work?

[Dylanclifford]

In “Measuring the Characteristics and Employment Dynamics of US inventors” it is found that inventors at young firms produce more impactful patents, but inventors are increasingly less likely to switch jobs or start new firms. Notably, our last lecture suggests that talent misallocation occurs when skilled individuals face barriers, like education costs, to entering innovation.

How might AI tools address these barriers? Like could AI lower startup costs (IE. automating R&D tasks) to help young firms attract inventors? Or might AI make big firms even more dominant by giving them better tools to retain talent? If inventors are already staying put at older firms, could AI worsen this trend by making incumbents more efficient, or reverse it by empowering startups?

And importantly, should policies focus on using AI to support small firms or regulate how big firms use AI in innovation?

[grozdanickata]

The paper “TAPPING INTO TALENT...” by Prof. Ufuk was intriguing to me for the way in which it highlighted the significant tradeoff in the process of expanding PhD slots. It is true that increasing the number of researchers will inherently boost the level of innovation in the pool, but it also often can lead to a decline in the average talent level of researchers (according to fact 9). In economics we learn about the idea of opportunity cost— that making changes/ improvements in one aspect always comes at the cost of another.
But theoretically, is it possible to innovate policy mechanisms targeting enhancement of both the quantity and quality of researchers at the same time, to “overcome” this idea of opportunity cost? How would a collaborative program with private research sectors (which typically have less financial obstacles than public ones) look in navigating these inconsistencies in the research space?

[anacpguedes]

In 'Measuring the Characteristics and Employment Dynamics of U.S. Inventors,' the findings emphasize the role of high-impact patents in driving innovation and economic growth. With the integration of AI into the invention process, barriers to entry are significantly lowered, likely resulting in an exponential increase in invention output. However, this raises concerns about the dilution of innovation quality.

Given that AI facilitates and accelerates idea generation, how might the resulting proliferation of low-impact inventions reshape the innovation landscape? Specifically, will the disproportionate contribution of high-quality patents to aggregate productivity remain stable, or could the increased prevalence of low-impact solutions create inefficiencies by saturating markets with less meaningful advancements?

[joycecz1412]

In "Tapping into Talent", the most significant indicator of who will become an inventor is their level of education, with the majority being PhDs. Given previous discussions on the influence of AI on education, would the broader use of AI into teaching and education result in higher quality education that incentives more PhDs or could it perhaps have the opposite effect? In relation to last weeks paper on innovation prediction, could AI potentially end up reduce the demand for PhDs, as fewer people are needed to create the same level of innovation? Innovation is a pareto distribution, and with most of the financial rewards going to the very few at the tip, the less creative innovators could be replaced by more efficient predictive AI.

[rbeau12]

"Measuring the Characteristics and Employment Dynamics of U.S. Inventors" reveals that inventors in mid-size young firms generate the most impactful patents. During the last couple of years, the United States has used aggressive protectionist policies to support domestic firms in cutting-edge industries like AI and EVs. Whether by design or as an unintended consequence, these policies have mainly benefited large, incumbent firms. Could these protectionist measures ultimately suppress technological progress by disadvantaging smaller firms? How should policymakers balance the protection of strategic domestic industries with the encouragement of free market innovation? Could a policy further both goals simultaneously?

[nmkhan100]

In “Measuring the Characteristics and Employment Dynamics of U.S. Inventors,” the authors point to barriers that limit who gets to innovate and how they succeed. This raises the question of what it takes to make the innovation ecosystem more inclusive.

How can policymakers tackle the issues of socioeconomic background and limited access to networks that hold many people back? Could focusing on improving education or creating mentorship opportunities help? Or would a more comprehensive approach—one that involves partnerships, targeted investments in underserved areas, and reforms across the system—be the key to unlocking untapped potential? What kind of challenges might such a broad approach face?

[dnlchen-uc]

"Tapping in Talent: Coupling Education and Innovation Policies for Economic Growth" finds that increasing the number PhD slots seemed to encourage innovation as a whole but was associated with a decline in the average IQ of PhDs. As the Danish government intended to double PhD enrollment within the span of two years, could this rapid expansion have resulted in a decrease in the quality of education as well? It is well established in education economics that class size is strongly correlated with educational outcomes. Was the quality of doctoral education adversely impacted by institutional limitations such as resource shortages and abnormally large class sizes? Since post-docs aren't subject to any major standarized tests (how educational outcomes can be traditionally measured) how can we measure and isolate for this specific effect?

[sabrinamatsui31]

Drawing from Measuring the Characteristics and Employment Dynamics of U.S. Inventors and Tapping Into Talent: Coupling Education and Innovation Policies for Economic Growth, both highlight the significance of talent allocation in driving innovation. The first article emphasizes the demographic underrepresentation among inventors, while the second discusses how financial frictions limit access to higher education for talented individuals. How could policies that specifically address gender and racial disparities among inventors synergize with initiatives aimed at reducing financial barriers to education? What mechanisms might policymakers use to ensure that such combined strategies foster both greater inclusivity and sustained innovation over time?

[pedrochiaramitara]

The article “Tapping in Talent: Coupling Education and Innovation Policies for Economic Growth” interested me a lot, as quantifying and analyzing the different approaches leader can take to increase a countries production is very interesting. There is a concern with long term growth in the paper, yet one could argue that specially in democracies there is a very large appeal to engage in short term investments. This tension might lead to an excessive investment in short term policies. How can police makers design policies that both benefit them in the short term and have long lasting effects? Is there a way the model can integrate the concern of politicians for short term results?

[darshank-uc]

In Measuring the Characteristics and Employment Dynamics of U.S. Inventors, Akcigit and Goldschlag demonstrate that younger inventors tend to work at younger and smaller firms. I’m curious to what extent this trend is motivated by inventor preference vs. barriers to entry––how does patience and risk tolerance vary among inventor age groups? While working at larger firms may provide experienced inventors more probable short-term rents from their patents, we’ve studied previously that small firms, if successful (which is the risk), also grow to mid-size and large firms very quickly––and with fewer total inventors contributing to this process, the more profit share they claim. On the other hand, aged incumbent firms may demand higher standards for inventors to justify higher baseline wages, which older inventors may possess over younger inventors. How could we model income levels across age groups for inventors conditional on the size of the firm they worked in? A potential proxy is to investigate job-hopping rates for inventors based on age––we might expect young inventors in failing young firms to find new opportunities given wage uncertainty.

[florenceukeni]

The article Tapping into Talent made me consider what structural barriers contribute to the underrepresentation of women and minorities in the field of inventors, highlighted by their lower representation in tech (e.g., 8.7% women) compared to sectors like healthcare and education (17.8% and 14.8%, respectively)? How do these barriers interact with dynamics in specific sectors, like cultural norms or access to resources, and what exactly do these patterns tell us about systemic challenges more broadly and the opportunity and necessity to foster diversity in innovation?

[michellema02]

In "Tapping into Talent: Coupling Education and Innovation Policies for Economic Growth," Akcigit, Pearce, and Prato show that the effectiveness of R&D subsidies is significantly lower than predicted by standard endogenous growth models due to talent constraints and financial frictions. They find that education subsidies can be more effective than R&D subsidies, especially in unequal societies where talented individuals face financial constraints. However, they note that different policies operate on different time horizons---R&D subsidies have immediate effects through lab equipment investments, while education policies take longer to impact growth through human capital development. Given these temporal differences, I'm curious: what is the optimal sequencing of innovation and education policies over time? Should policymakers front-load R&D subsidies to get immediate growth effects while education policies ramp up, or would this create inefficiencies by increasing demand for researchers before the talent pipeline is ready? Moreover, since the paper shows that R&D policy effectiveness depends on the existing talent pool, would early education investments create complementarities that make later R&D subsidies more impactful than implementing them simultaneously?

[jacobchuihyc]

"Tapping into Talent: Coupling Education and Innovation Policies for Economic Growth,” and “Measuring the Characteristics and Employment Dynamics of U.S. Inventors,” both examine the works of talent and innovation, underlining systemic barriers and policy interventions that define eventual outcomes. "Measuring the Characteristics and Employment Dynamics of U.S. Inventors" points, for instance, to serious questions of equitable access by women and minorities into positions that would allow them to pursue careers in innovation. Meanwhile, "Tapping into Talent" estimates that financial constraints greatly limit the number of talented people who can take up a research career, even more so in societies with unequal opportunities.

Given this set of findings, how might policymakers design interventions that tackle both demographic underrepresentation and financing constraints in innovation systems? For instance, can scholarships targeted to underrepresented groups, together with incentives for young high-impact firms, be designed to address this twin problem and engender an inclusive and dynamic innovation landscape? How would such interventions also vary in their impact across countries at different levels of inequality or industrial development?