Towards a Framework for Time Use, Welfare and
Household-centric Economic Measurement
Sixth International Monetary Fund Statistical Forum
November 19, 2019
Washington DC
Diane Coyle, University of Cambridge
Leonard Nakamura, Federal Reserve Bank of
Philadelphia
* The views expressed today are my own and not necessarily those of the Federal Reserve Bank of Philadelphia or the Federal Reserve System.
Why We need a New
Measure
• Rapid changes in consumer behavior as measured in
time use
– Is not seen in GDP or PCE growth
• Consumers face zero marginal prices for many
Internet products
– Because their marginal cost of reproduction has fallen to
zero
• With two-way mass communication, boundary
between paid work and household work has become
porous
• Widely viewed disconnect between welfare and GDP
www.philadelphiafed.org
Can Studying Time Use Place
a Money Metric on Welfare?
• If GDP doesn’t measure welfare, what can economists and
national statisticians do?
• Can we measure true economic progress with a money
metric?
– Can we answer quantitatively: is an economy performing well for its
participants?
• We argue that perhaps we can only answer this by
incorporating time use
• Time is:
– Fundamental to all human experience
– A required input to all consumption activity
– The costly input when consumption goods have zero marginal price
– Divided into paid work, household production, and leisure
www.philadelphiafed.org
“The real price of every thing, what every thing really costs to the man who wants to
acquire it, is the toil and trouble of acquiring it.” Adam Smith
What shall I buy
today?
What shall I do
today?
Time vs Money
• All face same budget
constraint
• Save more, more to
spend
• Unequal budget
constraints
• Save more, less to
spend
But they are inter-related: spend time to earn
more money, spend money to gain more time
The Internet ate my home work:
Adult hours online in UK
Activities move across
boundaries
Paid Work
Home
Production
Leisure
Routine
Teller
Driving
Jogging
Non-routine
Travel Agent
Caring
Cooking
Creativity:
taking photos
and videos
Life cycle issues: time can be
scarcest commodity
• Young educated households
– Work long hours developing their careers
– Spend quality time with kids
– Scrimp to afford a house in a good school district
– Commute long hours
– Don’t sleep much
– Often report unhappiness
• Shows the limits to intertemporal time trade
www.philadelphiafed.org
GDP ignores unpaid time
inputs
• The welfare benefit under GDP is basically personal
consumption expenditures as seen in transactions
– Conceived as U(C), where C is an long vector of purchased goods and
services
– Home production mainly outside the boundary of GDP
– While this takes place in time, time is usually not incorporated as an
input
– Nor does GDP ask how consumers feel while they consume or work
www.philadelphiafed.org
Expanding consumption to
utility
• Work on time use has taken broadly two directions:
– Household economics, with a shadow value of time, as proposed
initially by Becker
– Well-being studies based on self-reports
• For example, time use surveys that ask how participants feel in
given activities
• In household production (Becker, 1967)
– Count unpaid household work as labor
– Leisure time is consumption time
– Opportunity cost is the wage (in simplest version)
– If all time is measured by the wage, full income is wT
• Where T is time at paid work+ unpaid work + leisure
www.philadelphiafed.org
Survey measures of well-
being and value
• More recently, economists have been looking
to surveys to understand well-being
• Time use surveys increasingly include feelings
while engaged in activities (stated feelings)
– Direct reports of well-being
• Economists are also asking how feelings or
activities can be placed on a money metric
(stated preferences)
www.philadelphiafed.org
Lots of recent studies on time
use and/or broader wellbeing
measures
• Diewert & Fox 2018
• Alpman et al 2018
• Cassar & Meier, and Kaplan & Schulhofer-Wohl, JEcPerspectives 2018
• Hulten & Nakamura 2018
• Jones & Klenow, AER 2016
• Dotsey et al, Int Economic Review 2014
• Gershuny & Fisher 2014
• Benjamin et al, AER, 2012, 2014
• Deaton, 2018
• Bridgman, 2016
• Maestas et al, 2018
• Aguiar and Hurst, Handbook of Macro, 2016
+
• Goolsbee & Klenow, AER 2009
• Krueger et al, 2009
• Brynjolfsson et al 2018a, b
• Coyle, Economica 2018
• Coyle & Rogers in progress
Shadow value of time
• First approximation: wage rate
• But work may be pleasant or unpleasant
– Enjoyment means there is a consumption value to work, which raises
the shadow value of time relative to the wage
• And work may involve learning
– Learning by doing further raises the shadow value of time
• And work may be meaningful
– Studies show that people will accept lower wages to do work they
consider meaningful
• Can we find the shadow value of time by asking people?
– E.g., what wage would you have to be paid to shelve books at a
library?
– Or, what would you pay to have a shorter commute?
www.philadelphiafed.org
Three ways to estimate
shadow value of time
• “Revealed” preference: how much will
someone spend to save time?
– Trade-off between commute and rent may be
captured in measured rent gradients
• “Stated” preference: how much would you
pay to have a shorter commute?
• “Stated Feelings” : how do you feel as you
commute (direct report on welfare)
– Miserable commute raises the shadow cost of
commute relative to value of time
www.philadelphiafed.org
Many, many questions
• Can we reconcile the answers we get from these different
approaches?
– Can econometric encompassing techniques help us out?
• How will these money-metric utility measures align with real
consumption measures?
• We need time use data with stated feelings (discontinued in
US)
• We need more surveys—both private and official.
www.philadelphiafed.org
Summary
• Without a credible measure of aggregate welfare, economists’
ability to make macro policy recommendations will be
increasingly attenuated.
• To recapture welfare in the age of digitalization, we need
shadow prices, particularly of time.
• We are a long ways from a complete new picture, but a
tremendous amount of research has been launched.
• Coordinating this research, and maintaining it statistically over
time so that we can make time series, is the big task ahead.
www.philadelphiafed.org
Thanks!
• This is very much work in progress
• Email us with comments
– Leonard.nakamura@phil.frb.org
www.philadelphiafed.org
