Main takeaway

There is an inherent tension between the political economy and economics of innovation policy. The politics of innovation reconciles such tensions.

Why should you worry about innovation policy and politics?

The Texas CHIPS Act

• Bipartisan bill
• UTD is part of the Texas Semiconductor Innovation Consortium (TSIC)
• $700m allocated!

The Texas CHIPS Act case discussion

Assume you are a member of the Texas legislature. Do you think it would be in your self interest to be allocating resources to semiconductor innovation?

Assume you are a (non-partisan) economist. Do you think it is an efficient allocation of resources for the Texas legislature to be spending $700m on semiconductor technology?

Do you think the political institutions in Texas are designed so that members of Texas legislature would allocate resources to semiconductors?

The 3 bigger puzzles

Puzzle 1: Assume you are a politician in a given institutional setup. Do you think it would be in your self interest to be allocating resources to innovation? Why?

Puzzle 2: Assume you are an economist. Do you think it is efficient for the government to be allocating resources to innovation? Why?

Puzzle 3: Are political institutions designed to be incentive-compatible with decisions towards allocation of resources to innovation? Why?

Putting the puzzle pieces together

Claim 1

From an economic perspective, we can observe that innovation policy is cost-effective.

Consider an imaginary case…

CBO Director: "For every $10M spent on NIH funding, how many useful outputs to society?"

CBO Employee (You): 😕

Step 1: Understand the 3 main challenges to reckoning NIH funding impact

Step 2: Scratch your “Shoe leather” to go and talk to scientists

Step 3: Solve the attribution
challenge

Step 4: Solve the counterfactual challenge

Step 5: Give your boss some intermediate results

$$\begin{align*} Patents_{\widetilde{dst}} = \alpha_0 + \alpha_1Funding_{dst} + \beta' X_{dst} +& \\ \delta_{ds} + \gamma_{dt} + \nu_{st} + \epsilon_{dst}& \end{align*}$$

Step 5: Give your boss some intermediate results

Step 5: Give your boss some intermediate results

Step 6: Scratch your “Shoe leather”, again, to go and talk to NIH admins

Grant assessment at a study section level

Step 6: Scratch your “Shoe leather”, again, to go and talk to NIH admins

Grant assessment at a disease institute level

Compute Windfall funding (Instrument)

Step 7: Solving the endogeneity challenge

Step 8: Give your boss the final results

Back to the point - Claim 1

CBO Director: "For every $10M spent on NIH funding, how many useful outputs to society?"

CBO Employee (You):

2.7 patents ✅

From an economic perspective, we can observe that innovation policy is cost-effective.

Claim 2

There are clear political-economic reasons to believe that politicians will NOT be incentivized to allocate resources to innovation.

In this section:

• "Is this idea clear?" ✅
• "Is this idea correct?" ❌

Players, Utilities, and Strategies

• Players: Citizens, Incumbent ruler, New ruler
• Discrete time: $t$
• Everyone's utility depends on a single good: $y_t=A_t$
• Strategies:
  • Incumbent ruler: $x_t \in \{0 ,1 \}; T$
  • New ruler: $\widehat{x_t} \in \{0,1\}; \widehat{T} $
  • Citizens: $p_t \in \{0,1\}$

Timing

$x_t$	$p_t$	$\widehat{x_t}$	$y_t=A_t=...$
0	0	-	$A_{t-1}$
1	0	-	$A_{t-1}\left[ \alpha \right]$
0	1	0	$A_{t-1}\left[ 1-z \right]$
0	1	1	$A_{t-1}\left[ \alpha - z \right]$
1	1	0	$A_{t-1}\left[ 1 - z' \right]$
1	1	1	$A_{t-1}\left[ \alpha - z' \right]$

Cost of removing the incumbent ($z, z’$) and political competition ($\mu$)

$$ \begin{align*} &z' \sim F^I \equiv \mathcal{U}\left( \mu - \frac{1}{2}, \mu + \frac{1}{2}\right) \\ &z \sim F^N \equiv\mathcal{U}\left( \gamma\mu - \frac{1}{2}, \gamma\mu + \frac{1}{2}\right) \\ &\text{where } \gamma \geq 1 \end{align*}$$

Incentives to block innovation and political competition ($\mu$)

Modifications to the model

• Political rents and human capital
• External threats (i.e.) a war with a more/less developed nation

More pessimism!

• Delayed payoffs to innovation
• Opportunity cost of innovation policy

Claim 3

Yet we observe politicians allocating resources to innovation for political reasons.

3 reasons why politicians invest in R&D

1. Reason 1: Politicians can signal competence with risky innovation projects.
2. Reason 2: Politicians can channel benefits of innovation to connections.
3. Reason 3: Politicians can extract rents by creating a "lemons" market of patents.

Reason 1: Politicians can signal competence with risky innovation projects.

Competence signaling game

The competence signaling game - Sequence

The competence signaling game - Making sense of the utilities

The competence signaling game - Separating equilibrium

$$\delta \tau R \geq W \geq V \implies$$

The competence signaling game - Pooling equilibrium

$$V > W \,\, \& \,\, \delta \tau R > W \implies$$

Back to statistical “Shoe leather”

Vignette survey experiment

Vignette survey experiment

Results of the survey

Reason 2: Politicians can channel benefits of innovation to connections.

Reason 3: Politicians can extract rents by creating a “lemons” market of patents.

• What happens to the market when the low competence incumbent invests in R\&D in the pooling equilibrium of the competence signaling game?
• We get lemons (i.e) a market filled with bad quality patents!
  • Reference to George Akerlof

Conclusion - Same story everywhere!

Key takeaways from today’s class