Notes for Wednesday Podcast on Future of Work In
Description: Notes for Wednesday Podcast on Future of Work In general, why automation creates jobs Historically, waves of innovation drive demand Someone has to make, program hone the machines Someone has to design the products sell them That means
Related Topics
Download Presentation
"Notes for Wednesday Podcast on Future of Work In" is the property of its rightful owner. Permission is granted to download and print the materials on this website for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Presentation Transcript
slide1. Notes for Wednesday Podcast on Future of Work<br>
slide2. In general, why automation creates jobs Historically, waves of innovation drive demand
Someone has to make, program & hone the machines
Someone has to design the products & sell them
That means advertising & marketing
It might mean harnessing big data/machine learning
There also several indirect effects.<br>
slide3. Lessons from history Partial equilibrium view is just that: partial. The idea is that automation sometimes reduces jobs in labor intensive manufacturing.
But what about the general equilibrium? Since Industrial Revolution, industries have become automated. Textiles were the first, first with water power and then with steam.
Another germane example is agriculture. 1870: North of 80% employed in agriculture; 1900: 40%. Today only 1%. Combine harvesters/economies of scale/computers.
Displaced workers found new jobs—urbanization.<br>
slide4. New innovations lead to new markets Digital platforms: gig economy and independent contractors due to reduction in transaction costs. Ability to match & reduce information & coordination costs.
New financial engineers associated with algorithms & big data. New opportunities to save & invest. Better risk assessment. Microloans with higher yields.
IOT: device to device exchanges. Devices might be optimize to exchange with each other on behalf of individuals – so new payment services & ecosystem around these exchanges.<br>
slide5. Where will the new jobs come from? Lower costs for firms due to process innovation ïƒ shift supply curve out.
Direct effect
Lower prices means more disposable income for consumers: new demand for old stuff and new stuff: entertainment, VR, healthcare
More leisure time: new leisure industries.
Healthcare: cosmetic & optional (enhancements).<br>
slide6. Product Innovation Effect: shifting demand curves out New products are associated with R&D, design, consumer surveys and marketing: someone has to make and sell products that folks willing to pay for.
In other words, demand curves shifted out.
Who gets employed:
Scientists
Engineers
managers<br>
slide7. Indirect effect of shifting supply curve out Lower costs for firms due to process innovation ïƒ shift supply curve out.
Indirect effect: skills to complement automation related to new process innovation, such as monitoring and finetuning. Many skilled labor/technicians for last mile (textile automation).
Automation can sometimes lead to MORE jobs: ATM machines made it cheaper to operate a bank & thus more branches. More tellers needed, but their skillset changed: softer.<br>
slide8. What is the context we are in today? Factoids from my book in progress What is our point of departure?
What are the facts on the ground and salient patterns.
The takeaway point here: there are two economies. A high tech one that is quite healthy and a non tech sector that has not kept pace.<br>
slide9. TFP (growth rate, level & distribution within sectors) & profits (growth rate, level & distribution w/in sectors). Firm level & inventor level panel datasets for USA. Firm level d.s. has ~45K observations (6,500 firms) btwn. 2000 & 2015. Inventor level d.s. has ~59K observations btwn. 1976 & 2003; plus firm level patents btwn. 1973 & 2006.
TFP: how efficient firms are @ converting capital & labor into goods & services; Why should we care about it?
Most important source of differences in country living standards.
TFP grew tons during post WWII; explains higher real wages, lower inequality, bigger middle class & upward mobility.
When TFP high & growth faster, fewer zero sum battles centered on redistributing stagnant or shrinking pie.
Bottom line upfront:
Growth rate of TFP has been negative since 2000. High-tech firms’ TFP has grown, non-high tech firms have not, despite fact that high tech firms have had higher baseline levels of TFP.
Yet, high-tech firms’ profit margins have largely stayed same; non-high tech firms’ profit growth rate that’s increased.
Variance of TFP & profits has increased for non high-tech firms, not for high tech firms—instead, compression.
Less technological diffusion = superstar firms in non high tech sectors with systematically lower costs & higher TFP earn Ricardian Rents; more marginal firms struggle just to stay solvent. In high tech sectors, firms share technological knowhow via standardization & tighter networks.
Why more diffusion in some sectors, less in others & different rates btwn. sectors? I look to IP & tacit knowledge.<br>
slide10. Key to understanding productivity slowdown & heterogeneity in productivity & profits btwn. & within sectors Consider that 59% of your level of productivity btwn. 2007 & ’15 explained by your sector’s level of productivity.
Suggests technological diffusion story to make sense of productivity slowdown & heterogeneity in productivity & profits btwn. & within sectors.
In high tech sectors, firms all speaking same language around GPTs that are widely used, such as semiconductors & wireless infrastructure. [info. on patents citing high tech patents come from many technology classes.]
Standardization: these firms simply speak to each other more often & there are stronger connections btwn. [info. on distribution of backward citations by firms—high tech sectors have less citation inequality.]<br>
slide11. Standardiztion via SEPs of more general technologies Modems technology in wireless devices.
Data transmission = achieves higher data rates & system throughput
Carrier aggregation: create high capacity networks from fragmented radio spectrum.
Battery solutions: enhanced sleep modes that enhance battery life.
Resource allocation: coordinating transmission to prevent interference such as when multiple phones attempt to access same cell tower to increase user data rates & higher system throughput.<br>
slide12. Why tech clusters? Look to sociology & social network analysis (Granovetter) Social networks link senior managers (Saxenian; Casper):
culture of decentralized social ties linking scientists & engineers across local companies helps diffuse innovation—information sharing btwn. firms.
Skilled individuals help manage career risks of working in failure-prone firms. Create flexible labor pools with high mobility btwn. firms:
dense social networks across key personnel supporting career mobility
this incentivizes risktaking because if your startup fails, there’s a safety net: high density of other tech firms.
Examples:
Hybritech played dominant role in seeding development of two generations of successor biotech companies in San Diego: startup & successors embraced & commercialized UCSD tech.
Cleveland during development of electric machinery: Brush Electric Co.
Detroit: Olds Motor Works; Cadillac Motor; Ford; Buick.<br>
slide13. Importance of venture capital Technologies become focal points of conversation as managers & employees exchange scientific & technical information.
Venture capital acts as receiver & conduit of information—they can screen good ideas & develop reputation for doing so effectively.
Serving as hub to do demos & venture capital gets to know inventors & entrepreneurs.
Seed money from VC is Good Housekeeping Seal of Approval; once you get your start you can tap other funding sources, including loans.
Demonstration & information effects/can help create buzz.<br>
slide14. Facts on TFP Growth, TFP levels & TFP variance Avg. TFP growth for firms btwn. 2000 & 2015: -2.7%; median: -1.5%
Yet, things have slightly improved over time; less negative by .44% a year: pre-2006, -3.7%; post 2006 -1.89% per yr.
High-tech firms (semiconductors, computer manufacturing, software publishing, wireless telecom, data processing & hosting & computers system design) are exception: TFP growth: 3.3% per yr (3,201 obs.), with slight deterioration over time (4.4% btwn. 2000 & 2005; 2.4% btwn. 2007 & 2015), despite its higher baseline levels.
Due to high tech, TFP level has been stagnant instead of negative.
Increased variance over time in TFP driven by non high tech firms.<br>
slide15. Facts on Profit Growth, Profit levels & Profit variance (gross profits/costs of goods & services). Avg. profit growth for firms btwn. 2000 & 2015: 22% yr.
Things have improved 4% a yr.: pre-2006, 15%; post 2006, 32% per yr.
High-tech firms (semiconductors, computer manufacturing, software publishing, wireless telecom, data processing & hosting & computers system design) are exception: TFP growth: 3.3% per yr (3,201 obs.), with slight deterioration over time (4.4% btwn. 2000 & 2005; 2.4% btwn. 2007 & 2015), despite its higher baseline levels (145% more profits during sample period).
Modest convergence over time btwn. tech & non-tech in profit margins.
Increased variance over time in profits driven by non high tech firms.<br>
slide2. In general, why automation creates jobs Historically, waves of innovation drive demand
Someone has to make, program & hone the machines
Someone has to design the products & sell them
That means advertising & marketing
It might mean harnessing big data/machine learning
There also several indirect effects.<br>
slide3. Lessons from history Partial equilibrium view is just that: partial. The idea is that automation sometimes reduces jobs in labor intensive manufacturing.
But what about the general equilibrium? Since Industrial Revolution, industries have become automated. Textiles were the first, first with water power and then with steam.
Another germane example is agriculture. 1870: North of 80% employed in agriculture; 1900: 40%. Today only 1%. Combine harvesters/economies of scale/computers.
Displaced workers found new jobs—urbanization.<br>
slide4. New innovations lead to new markets Digital platforms: gig economy and independent contractors due to reduction in transaction costs. Ability to match & reduce information & coordination costs.
New financial engineers associated with algorithms & big data. New opportunities to save & invest. Better risk assessment. Microloans with higher yields.
IOT: device to device exchanges. Devices might be optimize to exchange with each other on behalf of individuals – so new payment services & ecosystem around these exchanges.<br>
slide5. Where will the new jobs come from? Lower costs for firms due to process innovation ïƒ shift supply curve out.
Direct effect
Lower prices means more disposable income for consumers: new demand for old stuff and new stuff: entertainment, VR, healthcare
More leisure time: new leisure industries.
Healthcare: cosmetic & optional (enhancements).<br>
slide6. Product Innovation Effect: shifting demand curves out New products are associated with R&D, design, consumer surveys and marketing: someone has to make and sell products that folks willing to pay for.
In other words, demand curves shifted out.
Who gets employed:
Scientists
Engineers
managers<br>
slide7. Indirect effect of shifting supply curve out Lower costs for firms due to process innovation ïƒ shift supply curve out.
Indirect effect: skills to complement automation related to new process innovation, such as monitoring and finetuning. Many skilled labor/technicians for last mile (textile automation).
Automation can sometimes lead to MORE jobs: ATM machines made it cheaper to operate a bank & thus more branches. More tellers needed, but their skillset changed: softer.<br>
slide8. What is the context we are in today? Factoids from my book in progress What is our point of departure?
What are the facts on the ground and salient patterns.
The takeaway point here: there are two economies. A high tech one that is quite healthy and a non tech sector that has not kept pace.<br>
slide9. TFP (growth rate, level & distribution within sectors) & profits (growth rate, level & distribution w/in sectors). Firm level & inventor level panel datasets for USA. Firm level d.s. has ~45K observations (6,500 firms) btwn. 2000 & 2015. Inventor level d.s. has ~59K observations btwn. 1976 & 2003; plus firm level patents btwn. 1973 & 2006.
TFP: how efficient firms are @ converting capital & labor into goods & services; Why should we care about it?
Most important source of differences in country living standards.
TFP grew tons during post WWII; explains higher real wages, lower inequality, bigger middle class & upward mobility.
When TFP high & growth faster, fewer zero sum battles centered on redistributing stagnant or shrinking pie.
Bottom line upfront:
Growth rate of TFP has been negative since 2000. High-tech firms’ TFP has grown, non-high tech firms have not, despite fact that high tech firms have had higher baseline levels of TFP.
Yet, high-tech firms’ profit margins have largely stayed same; non-high tech firms’ profit growth rate that’s increased.
Variance of TFP & profits has increased for non high-tech firms, not for high tech firms—instead, compression.
Less technological diffusion = superstar firms in non high tech sectors with systematically lower costs & higher TFP earn Ricardian Rents; more marginal firms struggle just to stay solvent. In high tech sectors, firms share technological knowhow via standardization & tighter networks.
Why more diffusion in some sectors, less in others & different rates btwn. sectors? I look to IP & tacit knowledge.<br>
slide10. Key to understanding productivity slowdown & heterogeneity in productivity & profits btwn. & within sectors Consider that 59% of your level of productivity btwn. 2007 & ’15 explained by your sector’s level of productivity.
Suggests technological diffusion story to make sense of productivity slowdown & heterogeneity in productivity & profits btwn. & within sectors.
In high tech sectors, firms all speaking same language around GPTs that are widely used, such as semiconductors & wireless infrastructure. [info. on patents citing high tech patents come from many technology classes.]
Standardization: these firms simply speak to each other more often & there are stronger connections btwn. [info. on distribution of backward citations by firms—high tech sectors have less citation inequality.]<br>
slide11. Standardiztion via SEPs of more general technologies Modems technology in wireless devices.
Data transmission = achieves higher data rates & system throughput
Carrier aggregation: create high capacity networks from fragmented radio spectrum.
Battery solutions: enhanced sleep modes that enhance battery life.
Resource allocation: coordinating transmission to prevent interference such as when multiple phones attempt to access same cell tower to increase user data rates & higher system throughput.<br>
slide12. Why tech clusters? Look to sociology & social network analysis (Granovetter) Social networks link senior managers (Saxenian; Casper):
culture of decentralized social ties linking scientists & engineers across local companies helps diffuse innovation—information sharing btwn. firms.
Skilled individuals help manage career risks of working in failure-prone firms. Create flexible labor pools with high mobility btwn. firms:
dense social networks across key personnel supporting career mobility
this incentivizes risktaking because if your startup fails, there’s a safety net: high density of other tech firms.
Examples:
Hybritech played dominant role in seeding development of two generations of successor biotech companies in San Diego: startup & successors embraced & commercialized UCSD tech.
Cleveland during development of electric machinery: Brush Electric Co.
Detroit: Olds Motor Works; Cadillac Motor; Ford; Buick.<br>
slide13. Importance of venture capital Technologies become focal points of conversation as managers & employees exchange scientific & technical information.
Venture capital acts as receiver & conduit of information—they can screen good ideas & develop reputation for doing so effectively.
Serving as hub to do demos & venture capital gets to know inventors & entrepreneurs.
Seed money from VC is Good Housekeeping Seal of Approval; once you get your start you can tap other funding sources, including loans.
Demonstration & information effects/can help create buzz.<br>
slide14. Facts on TFP Growth, TFP levels & TFP variance Avg. TFP growth for firms btwn. 2000 & 2015: -2.7%; median: -1.5%
Yet, things have slightly improved over time; less negative by .44% a year: pre-2006, -3.7%; post 2006 -1.89% per yr.
High-tech firms (semiconductors, computer manufacturing, software publishing, wireless telecom, data processing & hosting & computers system design) are exception: TFP growth: 3.3% per yr (3,201 obs.), with slight deterioration over time (4.4% btwn. 2000 & 2005; 2.4% btwn. 2007 & 2015), despite its higher baseline levels.
Due to high tech, TFP level has been stagnant instead of negative.
Increased variance over time in TFP driven by non high tech firms.<br>
slide15. Facts on Profit Growth, Profit levels & Profit variance (gross profits/costs of goods & services). Avg. profit growth for firms btwn. 2000 & 2015: 22% yr.
Things have improved 4% a yr.: pre-2006, 15%; post 2006, 32% per yr.
High-tech firms (semiconductors, computer manufacturing, software publishing, wireless telecom, data processing & hosting & computers system design) are exception: TFP growth: 3.3% per yr (3,201 obs.), with slight deterioration over time (4.4% btwn. 2000 & 2005; 2.4% btwn. 2007 & 2015), despite its higher baseline levels (145% more profits during sample period).
Modest convergence over time btwn. tech & non-tech in profit margins.
Increased variance over time in profits driven by non high tech firms.<br>