Dr Jacki ONeill DIODE Workshop 25 th May 2017 Ondemand economies and social good New ways of organizing work show promise for emerging markets But little evidence that they positively impact on disadvantaged communities ID: 736556
Download Presentation The PPT/PDF document "From Crowdwork to Ola Auto: Can Platform..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site 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.
Slide1
From Crowdwork to Ola Auto: Can Platform Economies Improve Livelihoods in Emerging Markets?
Dr.
Jacki
O’Neill
DIODE Workshop, 25
th
May 2017Slide2
On-demand economies and social good
New ways of organizing work show promise for ‘emerging markets’
But little evidence that they positively impact on disadvantaged communities
Examine whether and how these platforms impact livelihoodsSlide3
Ethnographic studies
Digital work: Amazon
Mechnical
Turk (AMT) in India and USA
Physical work
:
Ola Auto; Uber and Ola cab drivers in BengaluruSlide4
Themes
Independence
and
flexibility
Algorithmic management
Motivation and Job Satisfaction
Communication and Collaboration
Hidden Work
Market Transparency and Opacity
[…]Slide5
1. Workers as independent ‘freelancers’
What’s it like being a non-contracted worker? Slide6
Independence, flexibility and digital middlemen
Crowdwork as ultimate flexible workingThe crowdworker has to be flexible to the rhythms of work on the platform
Already independent workersOla acts as a digital middleman, eroding independence of drivers, whist doing little to reduce uncertainty
Amazon Mechanical Turk
Ketan
:
“I also try and look for work on
MTurk
when I have some time
but mostly I work at night because that’s when there are some jobs available”
Ola Auto
Mr. L
“if you wait for Ola rides to come in when you are waiting at a particular location […] you end up waiting all day without getting either normal or Ola passengers”
Workers value their independence and flexibilitySlide7
Current platform design
limits both the independence and flexibility of workers in these labour
markets
Raises issues of transparency and controlSlide8
3. Algorithmic management
What’s it like when your manager is an algorithm?Slide9
Implications of algorithmic management
Examples: workflow coordination and worker evaluationCoordination workThe work to organize people and groups so they work harmoniously togetherOften hidden workE.g. two models of task assignmentLeaving it to the user Automating itSlide10
Automating task assignment
Automation is not simple
Ride-assignment in ridesharing rarely takes into account all the contingencies
For example, cabs vs. and auto-rickshaws:Journey time, passenger opportunity, driver choiceAlgorithmically specifying all the contingencies is complex, even in ‘simple’ casesSlide11
Implications of algorithmic management
Examples: workflow coordination and worker evaluationCoordination workThe work to organize people and groups so they work harmoniously togetherOften hidden work
E.g. two models of task assignmentLeaving it to the user Automating itWorker evaluation Interchangeable workforce
Distinguished only by ratings (often stars) or reputationSlide12
Worker evaluation:
OlaAutoDrivers are evaluated by Olacabs and passengersBut the implementation is poor
The current system it is not actionable, leading us to ask, who is it for?
Mr N. “Sometimes in the afternoon, I have lesser stars and it increases in the evening. Some days when I have accepted lots of rides, I get lesser stars and some days when I barely have any passengers, I get rated high. I am not entirely sure why this happens.”Slide13
Algorithms will never be able to
take into account all the situated complexities of the world
Currently consequences of failure fall largely on workerSlide14
Can Platform Economies Improve Livelihoods in Emerging Markets?Slide15
On-demand economies for social good?
On-demand economies are here to stay
Provide valuable, flexible
employment opportunitiesBut current design limits effectivenessModelled as technologies not labor markets and limits of technologies not recognizedIgnore work practices and
design out rather than design for human factorsBalance of power
is skewed too far towards the platformSlide16
But it doesn’t have to be this way!
More equitable by design
Design opportunities
Promote greater independence and flexibilityFair algorithmic management Promote worker agency and individuality rather than control and standardizationSlide17
Research conducted
over many years with David Martin, Neha Gupta, Ban Hanrahan, Noopur Raval, Anupama Dhareshwar, Baneen Karachiwala, Srihari H Muralidhar, Syed Ishtiaque Ahmed
Thank you!Papers
Ahmed, Syed Ishtiaque, Nicola Bidwell, Himanshu Zade, Srihari Muralidhar, Anupama Dhareshwar, Cedrick Tandong Baneen Karachiwala, and Jacki O’Neill. Peer-to-peer in the workplace: A view from the road. In Proceedings of CHI, pp. 7-12. 2016Martin, David, O’Neill, Jacki, Hanrahan, Ben & Gupta, Neha (2016)
Turking in a Global Labor Market. Journal CSCW. Jan 2016
Neha Gupta, David Martin, Ben Hanrahan, and Jacki O'Neill,
Turk-Life in India
, in Proceedings of Group 2014,
2014
Martin, David, Hanrahan, Ben, O’Neill, Jacki, Gupta, Neha (2014),
Being a
Turker
. CSCW 2014
O’Neill
, J., Martin,
D (2013)
Relationship-based business process crowdsourcing
.
INTERACT 2013
O’Neill
, J., Roy, S. Martin, D., & Grasso, A.
(2013)
Form
Digitization in BPO: From Outsourcing to Crowdsourcing
.
CHI 2013
O’Neill, J., Roy, S. & Grasso, A
Is BPO work
crowdsourceable
?
India HCI
. Pune. India
2012Slide18
Back upSlide19
DATA
Auto-rickshaw drivers
in Bengaluru
66 auto-rickshaw drivers with and without Ola AutoObservation and in-situ interviewing of 23 drivers during 14 days in the field48 semi-structured interviews
Indian
crowdworkers
78 ‘door opener’ surveys
35 in-depth interviews and
walkthoughs
,
12 observations in 5 locations
Recordings of HITs
Ola and Uber cab drivers
in Bengaluru
48 Ola & Uber cab drivers
Observation of 35 rides
16 driver and passenger interviewsSlide20
2.
Communication and collaboration
Who do we talk to? Who do we learn from? Slide21
Platforms design collaboration out of the work
Networks of crowdworkers: physical & virtualHelp manage scarcity of good jobsWork quality (Yin et al, 2016)Provide community and support‘Employers’ work around system to form loose relationships with workers
US-based Uber drivers use forums to organize, make sense of the app and algorithms, etc. (Lee et al 2015)OlaAuto drivers do not have this option
Discuss locally, troubleshoot together, work things out on their ownLess powerful
But work remains collaborative!
Crowdwork
Uber, Ola, Ola Auto