8.2 Empirical Evidence; Firms, Polarisation and Displacement Flashcards
What does the wage inequality curve suggest? - what was Katz and Murphy’s estimation
Suggests that the wage premium should go down if relative supply of skilled workers increases
reminder: x axis is p = Ls/Lu
y axis is r = Ws/Wu
Estimation of graph:
ln (Ws/Wu) = B0 + B1t + B2ln(St/Ut)
B2 is negative - an increase in relative supply lowers relative wages
B1 is positive: demand shifting each year in favour of S
The model does not explicitly explain why B1 is positive, but tells us that the demand for S relative to U is increasing
What has been the common trend in relative share of skilled workers
The log supply index has been increasing over past few decades
How effective is the wage inequlaity model
- Assumes rate of technological progress constant - technology not directly measured; it is assumed
Estimated model does quite well in predicting the wage premium
This is consistent with the skill based technological change theory (also consistent with others though)
What is evidence at the industry level
What is the source of increase in demand for skilled workers?
- Employment of non production workers increased in 70s, accelerated in 1980s
- Employment of production workers in manufacturing fell
Found most of this increase in demand was within the industry
- Skill upgrading is positively associated with investment in computers and research and development expenditure
What ist he effect of technology on firms?
Direct measures of technology adoption at the firm level:
- Surveyed manufacturing plants asked about use of automation equipment such as computer automated design and numerically controlled machines.
- Firms use new technology employ larger fraction of workers in scientific, engineering, managerial and precision craft occupations
Correlation or causation - time series evidence suggest investing in new tech does not increase employment of high skill workers
- In contrast to findings on the effect of investment in IT technology
What has been the effect of job polarisation?
Job polarisation - shows change in occupational shares:
- Middle have reduced shares - between 15 and 5%
- Low have mainly positive around 3% change, however UK and Finland negative share
- High paying have positive change, between 3 and 12%
What is the effect of job displacement?
Automation does displace workers - even if aggregate effect is positive
But how long does it take dissolved workers to find another job - does it ahve a long term effect:
Jacobson et al investigated long term effect of job loss:
Large job lossses even after 5 years
- Graph at bottom of 8.4
Which 3 ways is technological change modelled?
- Change in factor prices (computers cheaper)
- Factor augmentation (computers make skilled workers more produtive)
- Task replacement (computers replace workers who do routine tasks)
All tell similar story - technology increases output and benefits some workers, but harms others
What are some policy issues?
- Education
- Worker mobility
- Income redistribution and compensation
- Universal basic income
What should we expect looking forward
Range of tasks machines do will be vastly greater - need environmental control and machine learning
Doesnt mean end of work - principle of comparative advantage important here
- Creation of tasks; do not know what these tasks are
Keynes believed we can all work a few hours a week - but distribution of work may be very unequal - distribution may occur.