Tuesday, May 20, 2025

It's Easy to Criticize Tax Policy, Hard to Manage Well

There always is plenty to disagree about whenever a change in U.S. taxation policy is proposed, and one frequent criticism is that the benefits of proposed tax reductions disproportionately benefit the wealthier taxpayers.


It is a charge one might always agree with, but policymakers always face a balancing act. 


On one hand, legislators must ensure that the federal government has adequate financial resources to cover its expenditures. The rest of us do not have that primary concern. Policymakers have to worry about the economic impact of any changes. The rest of us likely are more concerned about perceived fairness. 


But tax policy profoundly influences economic behavior. High marginal tax rates can discourage incentives to work, save, and invest. So the trick is to create structures that minimize disincentives while still generating revenue. And since capital “has no nationality,” tax rates often must be adjusted to make the U.S. more attractive for businesses compared to other countries.


The vast majority of federal government revenue comes from individual taxpayers, in the form of Individual income taxes and payroll taxes (social security). 


And that’s where we confront complicating issues, namely the actual incidence (who bears the burden of payment). In 2022, for example, the top one percent of taxpayers contributed 40 percent of all federal individual income taxes.


The top 10 percent of taxpayers  paid around 72 percent of income taxes.


Conversely, the bottom 50 percent of taxpayers contributed about three percent of revenue.


The point is that it typically is possible to argue that proposed changes are unfair in some major way: either because wealthier taxpayers get more of the benefits or must bear more of the burden; or because lower-income taxpayers or middle-income taxpayers do not receive enough of the benefits. 


Policymakers face a different problem: the only place changes are meaningful are at the top: The top 25 percent of filers contribute 87 percent of total revenue. 


Income Group

Share of Total Income Taxes Paid (2022)

Bottom 50%

3.00%

50% to 25%

10.00%



25% to 10%

15.00%

10% to 5%

11.00%

5% to 1%

21.00%

Top 1%

40.40%

Source: Tax Foundation, "Summary of the Latest Federal Income Tax Data, 2025 Update”


If one wishes to raise more revenue, it has to affect taxpayers at the top. If one wishes to cut taxes, it will disproportionately affect those at the top. 


That might never set too well with many citizens, but is a fundamental policymaker constraint. Also, though “soak the rich” might be a nice bumper sticker for some, investment effects matter. 


Looking only at public markets, the top 10 percent of U.S. households own about 93 percent of U.S. households' stock market wealth, with the richest one percent owning 54 percent of public equity markets, according to the Institute for Policy Studies, based on Federal Reserve data. The bottom 50 percent of U.S. households own less than one percent of stock market wealth.


The point is that private investment hinges on a small percentage of U.S. households. So policymakers have to balance revenue needs with fairness with maintaining investment incentives. “Soak the rich” might be popular in some circles. It is not so easy for policymakers.


Hot Buzzwords are Part of Computing Supplier Marketing War

Hot buzzwords always seem to have been part of the sales and marketing spin for computing-related companies and their products. Today’s effort is to label everything “AI.” But “com” was the rage around the turn of the century and “personal” was key in the 1980s. “Cloud” and “virtualized” are of more-recent vintage. “Zero trust” and “low code/no code” have been heard over the last few years. 


Era/Period

Buzzword(s)

Context / Typical Usage

1980s–Early 1990s

Personal Computer (PC), GUI, Multimedia

Launch of consumer computing, graphical interfaces

Late 1990s

.COM, E-commerce, Web Applications, ASP

Internet boom, online business, dynamic web development

Early 2000s

Distributed Computing, XML, COM, CORBA

System integration, interoperability, enterprise IT

Mid-2000s

SOA, Web Services, Agile, Best Practices

Modular software, service integration, project mgmt

Late 2000s

Cloud Computing, Virtualization, SaaS

Move to hosted services, scalable infrastructure

Early 2010s

Big Data, Mobile-First, BYOD, DevOps

Data analytics, mobile devices, agile IT operations

Mid-2010s

IoT (Internet of Things), Machine Learning, Blockchain, Chatbots

Connected devices, automation, decentralized ledgers

Late 2010s

AI (Artificial Intelligence), Digital Transformation, Edge Computing, AR/VR

Automation, business change, real-time processing, immersive tech

Early 2020s

Metaverse, Quantum Computing, Zero Trust, Multi-Cloud, Serverless, Low Code/No Code

Virtual worlds, advanced computing, security, cloud strategies, simplified dev

Mid-2020s (2024–2025)

Generative AI, Cloud-Native, FinOps, Hybrid Cloud, Cloud Security Posture Management (CSPM), Anywhere Operations

AI content creation, optimized cloud, financial ops, security, remote work


Monday, May 19, 2025

Nvidia Unveils NVLink Fusion, Enabling Use of Third Party CPUs, GPUs

Among other announcements Jensen Huang, Nvidia CEO, made in a speech at the Computex trade show, he unveiled NVLink Fusion, a platform that allows customers to integrate non-NVIDIA CPUs or accelerators with NVIDIA's GPUs in custom rack-scale AI systems. 

The move allows chip designers, such as Marvell Technology and MediaTek, to use NVLink for high-speed chip-to-chip communication, facilitating the creation of powerful, custom AI systems. 

Customers can pair NVIDIA Blackwell GPUs with custom CPUs or AI accelerators, such as Qualcomm’s Snapdragon CPU cores, within NVIDIA’s ecosystem, for example. 

Sunday, May 18, 2025

NBER Study Suggests Limited AI Chatbot Impact on Earnings, Productivity

A study of artificial intelligence chatbot impact on labor markets in Denmark suggests the economic impact is “minimal.” Indeed, the study authors say “AI chatbots have had no significant impact on earnings or recorded hours in any occupation.” 


The study published by the U.S. National Bureau of Economic Research involved two large-scale adoption surveys conducted in late 2023 and 2024 covering 11 occupations; 25,000 workers and 7,000 workplaces.


Productivity gains were said to be modest, with an average time savings of three percent. But the study notes that AI chatbots have created new job tasks for 8.4 percent of workers, including some who do not use the tools themselves.


Nor has there been any impact on worker earnings. “Workers overwhelmingly report no impact on earnings as of November 2024,” the study says. 


Nor do productivity gains seem to have much impact on earnings. “We estimate that only three to seven percent of workers’ productivity gains are passed through to higher earnings,” say authors Anders Humlum and Emilie Vestergaard.


“Comparing workplaces with high versus low rates of chatbot usage, we find no evidence that firms with greater adoption have experienced differential changes in total employment, wage bills, or retention of

incumbent workers,” the authors say. 


The authors also note that Denmark has institutional characteristics similar to those of the United States, with similar uptake of generative AI; how hiring and firing costs; decentralized wage bargaining and annual wage negotiations. 


The 11 occupations studied included accountants, customer support specialists, financial advisors, HR professionals, IT support specialists, journalists, legal professionals, marketing professionals, office clerks, software developers, and teachers.


The findings should not come as a surprise. The “productivity J-curve" suggests that initial investments in new technologies may temporarily suppress productivity before delivering long-term benefits.


Study

Technology Examined

Lag Time Observed

Key Findings

McKinsey Global Institute 1,5,7

Digital technologies, AI

Years to decades

Benefits emerge after business process redesign and "creative destruction." Historical parallels (e.g., electric power) show lags of decades. Generative AI may shorten lags to months or years.

CEPR Study on French Industrialization 3

General-purpose technologies

5–10 years

Firms delayed adoption due to uncertainty, and early adopters operated technologies inefficiently. Aggregate productivity gains materialized slowly as organizational practices evolved.

Stanford CS Analysis 4,5

IT investments

2–5 years

Executives reported 5-year lags for IT payoffs. Complementary investments and learning curves delayed measurable productivity growth.

Productivity Paradox Research 5

IT, automation

2–5 years

"Productivity J-curve" observed: short-term costs offset gains until workflows adapted. Measurable aggregate gains emerged in the 2000s from 1990s IT investments.

Brynjolfsson et al. (McKinsey) 7

Generative AI

Months to a few years

Shorter lag due to existing digital infrastructure, but still requires process redesign. Early adopters see inefficiencies before optimization.


Saturday, May 17, 2025

Declining Robotics Costs Drive Substitution for Human Labor

Robots, as a form of embodied artificial intelligence, are declining in cost so much that it is virtually inevitable they will become functional substitutes for human labor in many instances, as wage costs continue to grow at a relatively consistent rate. 


On the other hand, sometimes there is no substitute for humans.

Year

U.S. Labor Cost Index (Normalized)

Estimated Robot Cost per Working Hour (USD)

2010

80

40

2011

82

38

2012

85

36

2013

87

34

2014

90

32

2015

92

30

2016

95

28

2017

97

26

2018

100

24

2019

103

22

2020

106

20

2021

110

18

2022

113

16

2023

116

14

2024

119

12

2025

121

10

In many cases, at some point, there is a total cost of ownership crossover: it becomes more reasonable to use robots instead of humans for some tasks. 


Labor Rate Trends vs. Cost of Using Robots in the U.S. (2010-2025)


Using an employment cost index, which compares wage costs, inflation and benefits, the cost of human labor has increased about 40 percent since 2010. The cost of using robots, per hour, has fallen 400 percent over the same period. 


Year

U.S. Labor Cost Index (Normalized)

Estimated Robot Cost per Working Hour (USD)

2010

80

40

2011

82

38

2012

85

36

2013

87

34

2014

90

32

2015

92

30

2016

95

28

2017

97

26

2018

100

24

2019

103

22

2020

106

20

2021

110

18

2022

113

16

2023

116

14

2024

119

12

2025

121

10


It's Easy to Criticize Tax Policy, Hard to Manage Well

There always is plenty to disagree about whenever a change in U.S. taxation policy is proposed, and one frequent criticism is that the benef...