Ideas offer innovations that enable us to do our jobs more effectively and efficiently, require fewer resources, contribute to an improved quality of life, and drive economic growth. A stylized view of economic growth that emerges from idea-based growth models suggests that economic growth is directly related to R&D productivity and the number of researchers. In this view, R&D productivity is defined as the ratio between total factor productivity (TFP) and the number of researchers engaged in the search for ideas. Unfortunately, there has been a declining R&D productivity trend, at an annual rate of 7% in the USA.
Recent studies, conducted by Professors of Economics at Stanford University, have found that despite exponential growth in the number of researchers, TFP contribution has remained flat over many decades. By referring to this ratio, it’s fair to conclude that R&D productivity has been declining. It presumes that R&D productivity falls as ideas become harder to find, because the Innovation mine is becoming deeper. Such a belief leads to a policy of increasing R&D funding for offsetting falling R&D productivity. But does this belief offer a sustainable solution, or may it fail to deliver a proportionate contribution to economic growth? Are there other reasons affecting R&D productivity that should be addressed to maximize growth out of R&D investment?
Although the declining R&D productivity trend suggests that it is becoming harder to find ideas, there has been an exponential growth in patent filings, publications, and Startups. The sole purpose of conducting R&D to produce ideas may not be to generate economic rent. A single idea may not be enough to produce economic benefit. Often, there is a need for a Flow of Ideas to accumulate in a cumulative form, fuelling a new wave of growth. However, for producing economic value, those new waves need to cross the threshold set by incumbent mature innovations. Unfortunately, in many cases, they suffer from Premature Saturation and fail to deliver an economic contribution. Hence, R&D productivity may suffer for various reasons. Therefore, there could be alternatives to increasing R&D funding for driving economic growth from ideas.
Genesis of Declining R&D Productivity
An often cited example of declining R&D productivity is Moore’s Law in the semiconductor industry. To maintain the same level of chip density growth, which, by the law, doubles every two years, the number of researchers required today (as of 2017) is more than 18 times larger than the number needed in the early 1970s. Such a reality leads to the observation that ideas are becoming increasingly difficult to find. Hence, research productivity has been declining at a rate of 7 percent per year. And as a result, declining gross margins of semiconductor companies like Intel have been justified. To offset this, there should be a more substantial allocation for R&D to engage more research professionals. But for Intel, will it work? Is there a risk that such a policy decision may lead to weakening Intel’s financial health further?
To gain further insights, it is worth examining the rise in TSMC’s financial health and gross margin. What are the underlying causes? How has TSMC been enjoying far greater R&D productivity in the same semiconductor industry? Based on my analysis, it appears that Intel has been suffering from eroding R&D productivity due to its pursuit of a declining microchip architecture and business model. Instead of just increasing R&D spending to maintain growth, Intel had to decide to switch from a maturing wave to an emerging one. The effectiveness of this reality is lacking in predominant idea-based growth models, for which Nobel Prizes in Economics were awarded in 2018 and 2025.
Presumptions of Idea-based Growth Models—are they flawed?
One of the notable contributions of the idea-based growth model is Paul Romer’s idea and object theory. It attempts to promote the assumption that the sole purpose of idea production is to enjoy monopoly rent out of innovation. Each idea, notably a patented one, offers a contribution to economic growth due to its provision of a new tool (innovation) in accomplishing tasks better. Besides, it makes the assumption that ideas are static, having infinite life. Hence, economic growth is a function of idea stock. Therefore, by increasing the supply of ideas, a firm or a nation can achieve higher economic growth.
To address the static role of ideas in creating economic value, the role of Creative Destruction has been recognized, offering Nobel Prizes in 2024 and 2025. The basic notion has been that new ideas destroy older ones. However, the formation of new waves through a flow of ideas and why they may end up in premature saturation, wasting idea stock, is not explained in those Nobel Prize-winning idea-based growth models. As a result, they do not offer insights into the diverse reasons for declining R&D productivity, as opposed to merely stating that it’s becoming harder to find big ideas.
Improving R&D Productivity—what are the avenues?
As the data suggest, R&D productivity, measured as the ratio between total factor productivity (TFP) and R&D professionals, has been declining—as shown in the Figure. The common perception has been that it has been getting harder to find big ideas. As if this is a natural phenomenon. We have no choice but to increase R&D investment. However, there are avenues for increasing R&D productivity:
1. Non-Commercial Purpose of Ideas Production through R&D: There has been a notion that R&D investment is made to generate ideas for profiting from the monopoly rent of their commercialisation. Hence, there is supposed to be a linear correlation between idea production and economic growth or TFP. However, over the decades, there have been changes. Only a small fraction of ideas is being pursued for commercialization. For example, as high as 97% patents fail to recover patenting cost, and almost 50% patents are abandoned. Hence, R&D spending for producing those patents has not contributed to TFP. One of the underlying reasons is that patented ideas are increasingly being produced to obtain other incentives, such as promotion and grants from the government, and create barriers to new entrants. Therefore, in addition to R&D funding for idea production, creating economic value from them should receive equal or higher priority to address declining R&D productivity.
2. Increasing Reinvention Barriers to Create Economic Value out of Ideas: Although economic theories presume that every idea or patent offers monopoly rent to inventors, the reality has been different. Unlike the past, a single idea does not offer profitable revenue. Often, unless ideas do add up (like the iPhone) to fuel a creative destruction wave to cross the threshold, R&D investments made for those ideas do not contribute to TFP or economic growth. The high failure rate of startups indicates that a significant number of ideas retire before making a measurable economic contribution. As the threshold has been increasing, a growing volume of ideas have been failing to contribute to TFP. For example, ideas for electric vehicles are yet to offer full benefits. Similarly, ideas for Humanoid robots or autonomous vehicles are yet to be implemented. Hence, it’s not getting harder to find ideas. Instead, due to increasing barriers to cross, ideas are facing growing hurdles to add value. Therefore, there is a reason for declining R&D productivity. To address it, we need ideas from a deeper layer of science. For example, scientific discovery has opened the path to leveraging LED idea.
3. Growing Competition Reducing Innovation Success: Competition has been reducing the innovation success rate. Additionally, the lifespan of ideas in creating economic value has also been decreasing due to increased competition. Some statistics suggest that more than 75% of innovations fail to return a profitable return before being retired. Furthermore, due to intensifying competition, innovators are facing growing challenges in sustaining their ideas in the market. Hence, before the expiration of patent life, many ideas are losing their ability to add value to TFP. Therefore, growing competition has also been reducing R&D productivity.
4. Rational Decisions Affect R&D Productivity: Despite the potential, commercialization of ideas is fraught with pervasive uncertainty. Hence, rational decisions in the midst of uncertainty matter. Due to its weakness, companies run the risk of wasting R&D investment in producing ideas. For example, due to decision-making errors, Kodak was unable to benefit from digital camera idea; thereby, the US economy was deprived. Hence, to increase R&D productivity, the focus should be on understanding innovation dynamics to improve rational decision-making accuracy.
5. Subsidy Race of Startups Wastes Ideas: Instead of refining ideas, most startups have been in a race to promote half-baked innovations with massive subsidies for inflating the valuation. As a result, existing waves suffer from unfair competition. On the other hand, when venture capital funds stop funding, all those startups disappear, wasting their ideas and potential. As a result, the likely economic contribution remains unrealized. Hence, by encouraging startups to focus on refining their ideas to offer better alternatives, we can enhance R&D productivity.
6. Monopolization Tends to Keep Ideas Unused: Ideas driving technological innovations have the potential to improve the quality and reduce the cost simultaneously. Such potential opens the door to specialization in making the quality-cost trade-off irrelevant, gaining an edge over competitors. As a result, the winner ends up monopolizing the market. Once competition becomes irrelevant, a monopoly stops investing in integrating new ideas into its products and processes, resulting in unused ideas that negatively affect R&D productivity.
7. Race of Inflating Valuation out of Disruptive innovation Narratives: It has been alarming to observe that there has been a race to get rich very quickly by spreading disruptive innovation narratives out of big ideas, as opposed to delivering them. For example, due to the spreading of the electric vehicle (EV) narrative, Tesla’s valuation rose sharply, making its CEO the wealthiest person in the world. Unfortunately, instead of driving the EV wave as a successful creative destruction, Tesla has been shifting its focus to promoting new hype around robots and AI, thereby inflating its valuation further. Such a trend appears to be detrimental to leveraging ideas to drive economic growth. Hence, such a trend must be stopped and innovation mission should be steered towards to delivering better tools to increase R&D productivity.
Declining R&D productivity is a concern for driving economic growth. However, existing theories about idea-based growth models do not offer insights into the diverse reasons, other than concluding that as the innovation mine becomes deeper, it is becoming harder to find big ideas like in the past. Hence, such models tend to recommend only one policy option: increasing R&D investment to offset declining R&D productivity. However, as explained in this article, there are several other reasons for declining R&D productivity. Hence, despite finding and pursuing big ideas such as autonomous vehicles and humanoid robots, the economy has not benefited from them.
It’s worth mentioning that ideas alone do not contribute to economic growth. To make it happen, they must be integrated into the product and processes, creating a willingness to pay among customers. Hence, increasing the supply of ideas through allocating more R&D resources is not sufficient for idea-led growth. This reality deserves consideration in addressing diverse issues, as explained in this article, related to declining R&D productivity.
