Appendix 1
Digital labour platforms: Estimates of workers,
investments and revenues
Table A1.1 List of country codes
Country
ISO
Alpha 3
Albania ALB
Algeria DZA
Argentina ARG
Armenia ARM
Australia AUS
Bangladesh BGD
Belarus BLR
Benin BEN
Bolivia, Plurinational
State of
BOL
Bosnia and Herzegovina BIH
Brazil BRA
Bulgaria BGR
Cameroon CMR
Canada CAN
Chile CHL
China CHN
Colombia COL
Costa Rica CRI
Croatia HRV
Cyprus CYP
Denmark DNK
Dominican Republic DOM
Ecuador ECU
Egypt EGY
El Salvador SLV
Ethiopia ETH
Finland FIN
Country
ISO
Alpha 3
France FRA
Georgia GEO
Germany DEU
Ghana GHA
Greece GRC
India IND
Indonesia IDN
Ireland IRL
Israel ISR
Italy ITA
Jamaica JAM
Japan JPN
Kazakhstan KAZ
Kenya KEN
Madagascar MDG
Malaysia MYS
Mauritius MUS
Mexico MEX
Morocco MAR
Nepal NPL
Netherlands NLD
New Zealand NZL
Nicaragua NIC
Nigeria NGA
North Macedonia MKD
Norway NOR
Pakistan PAK
Country
ISO
Alpha 3
Peru PER
Philippines PHL
Poland POL
Portugal PRT
Republic of Moldova MDA
Romania ROU
Russian Federation RUS
Saint Lucia LCA
Senegal SEN
Serbia SRB
Singapore SGP
Slovakia SVK
South Africa ZAF
Spain ESP
Sri Lanka LKA
Sweden SWE
Thailand THA
Tunisia TUN
Turkey TUR
Uganda UGA
Ukraine UKR
United Arab Emirates ARE
United Kingdom GBR
United States USA
Uruguay URY
Venezuela, Bolivarian
Republic of
VEN
Viet Nam VNM
The role of digital labour platforms in transforming the world of work
2
Table A1.2 Estimates of workers performing tasks on digital platforms
Reference Estimate Countries and years Time period/proportion of income Denition
Urzì Brancati,
Pesole and
Fernández Macías
(2020)
9.5–11% of adult population (aged
between 16 and 74 years)
16 EU Member
States,* 2018
Ever gained income from providing services via online
platforms.
Providing labour services via online platforms; payment
is conducted digitally via the platform, and tasks are
performed either online web-based or on-location.
1.9–2.4% of adult population
Provided labour services via platforms but less than once
a month over the last year.
3.1% of adult population
At least monthly, but for less than 10 hours a week and
earned less than 25% of their income via platforms.
4.1% of adult population
At least monthly, for between 10 and 19 hours or earned
between 25% and 50% of their income via platforms.
1.4% of adult population
At least monthly, and worked on platforms at least 20
hours a week or earned at least 50% of their income via
platforms.
Pesole et al. (2018)
9.7% on average
6–12% of adult population
14 EU Member States,
2017
Provided labour services at any time in the past.
Providing services via online platforms (location-based
and web-based).
8% on average
4–10% of adult population
Provided services regularly at least once a month in the
past year.
Alsos et al. (2017) 1% of working-age population Norway, 2016–17 Earned money through labour platforms in the past year.
CIPD (2017) 4% of working adults (18–70 years)
United Kingdom,
2016
Engaged in paid platform work at least once in the
previous 12 months.
Platform work includes performing tasks online, providing
transport or physically delivering food or other goods.
Huws et al. (2017)
9–12% in Germany, Netherlands,
Sweden, United Kingdom
18–22% in Austria, Italy, Switzerland
Austria, Germany,
Italy, Netherlands,
Sweden, Switzerland,
United Kingdom,
2016–17
Performed crowdwork at any time in the past.
Crowdwork is paid work via an online platform, such as
freelance platforms or outside one’s home on location-
based platforms.
6–8% in Germany, Netherlands,
Sweden, United Kingdom
13–15% in Austria, Italy, Switzerland
Performed crowdwork at least monthly.
5–6% in Germany, Netherlands,
Sweden, United Kingdom
9–12% in Austria, Italy, Switzerland
Performed crowdwork at least weekly.
Farrell, Greig and
Hamoudi (2018)
1.6% on all platforms
1.1% on labour platforms, 0.2% on
capital platforms, 0.4% selling (28
million US bank accounts)
United States, 2016
Earned income from platform work over the past month.
Labour platforms are those on which participants perform
discrete tasks, and capital platforms are those whose
participants sell goods or rent assets.
4.5% on all platforms Earned income from all platform work over the past year.
Appendix 1. Digital labour platforms: Estimates of workers, investments and revenues
3
Reference Estimate Countries and years Time period/proportion of income Denition
Burson-Marsteller,
Aspen Institute
and Time (2016)
42% of adult population
United States, 2015
Have purchased or used one of the services.
Services in the on-demand economy include: ride-sharing,
accommodation sharing, task services, short-term car
rental, or food or goods delivery.
22% of adult population Have oered at least one of the services in the past.
7% of adult population
Earn in a typical month at least 40% through on-demand
economy.
Katz and Krueger
(2016)
0.5% of labour force United States, 2015 Reference period – one week. Working through an online intermediary.
Surveys conducted by national statistical oces
Switzerland, FSO
(2020)
0.4% of the population
1.6% of the population
Switzerland, 2019 In the past 12 months.
Carried out work via internet-mediated platforms.
Provided internet-mediated platform services.
United States, BLS
(2018)
1% of total employment United States, 2017 In the last week.
Electronically mediated workers, doing short jobs or tasks
through websites or mobile apps that both connect them
with customers and arrange payment for the tasks.
Ilsøe and Madsen
(2017)
2.4% of working-age population
Denmark, 2017
In the past year.
Earned money via digital platforms, both labour and
capital platforms.
1% of working-age population In the past year.
Earned money via a labour platform such as Upwork,
Happy Helper.
1.5% of working-age population In the past year.
Earned money via a capital platform such as Airbnb,
GoMore.
Sweden, SOU
(2017)
4.5% of working-age population
Sweden, 2016
In the past year. Tried to get an assignment via a digital platform.
2.5% of working-age population In the past year. Performed work via a digital platform.
Canada, Statcan
(2017)
9.5% of adult population
(≥ 18 years) (7% ride services;
4.2% accommodation)
Canada, 2015–16
In the past 12 months.
Used either peer-to-peer ride services or private
accommodation services.
0.3% of adult population (≥ 18 years) In the past 12 months. Oered peer-to-peer ride services.
0.2% of adult population (≥ 18 years) In the past 12 months. Oered private accommodation services.
Statistics Finland
(2018)
7% of adult population Finland, 2017 In the past 12 months.
Worked or earned income from the following platforms:
Airbnb, Uber, Tori./Huuto.net, Solved, and others.
* These 16 EU Member States are Czechia, Croatia, Finland, France, Germany, Hungary, Ireland, Italy, Lithuania, the Netherlands, Portugal, Romania, Slovakia, Spain, Sweden and the United Kingdom.
Source: ILO compilation.
Table A1.2 (cont’d)
The role of digital labour platforms in transforming the world of work
4
Table A1.3 Total funding from venture capital and other
investors, selected categories of digital labour platforms,
by region and type of platform, 1999–2020
Funding
(US$ million)
Number
of platforms
Number
of countries
Delivery 37 495 164 47
Africa 13 5 4
Arab States 48 6 5
Central and Western Asia 51 6 4
East Asia 8 915 16 3
Eastern Europe 110 10 3
Latin America and the Caribbean 3 019 15 9
North America 11 116 44 2
South Asia 4 19 9 21 2
South-East Asia and the Pacic 222 8 3
Western Europe 9 803 33 12
Taxi 62 784 61 30
Africa 45 8 6
Arab States 772 1 1
Central and Western Asia 929 2 2
East Asia 21 581 4 2
Eastern Europe 1 001 2 2
Latin America and the Caribbean 337 6 3
North America 33 032 19 1
South Asia 3 850 5 4
South-East Asia and the Pacic 26 4 2
Western Europe 1 211 10 7
Online web-based 2 69 0 142 31
Arab States 0.3 1 1
Central and Western Asia 113 4 2
East Asia 579 11 3
Eastern Europe 12 5 3
Latin America and the Caribbean 2 4 3
North America 1 601 66 2
South Asia 7 6 2
South-East Asia and the Pacic 77 9 5
Western Europe 299 36 10
Hybrid 16 999 5 4
Africa 908 1 1
South-East Asia and the Pacic 15 10 0 2 2
East Asia 991 2 1
Source: Crunchbase database.
Appendix 1. Digital labour platforms: Estimates of workers, investments and revenues
5
Table A1.4 Estimated annual revenue of digital labour
platforms, by region and type of platform, 2019–20
Revenue
(US$ million)
Number
of platforms
Number
of countries
Delivery 25 063 191 36
Africa 10 3 4
Arab States 113 7 3
Central and Western Asia 231 1 1
East Asia 9 107 101 4
Eastern Europe 63 7 5
Latin America and the Caribbean 934 6 4
North America 9 10 4 34 1
South Asia 690 10 1
South-East Asia and the Pacic 90 6 5
Western Europe 4 772 16 8
Transportation 17 343 31 18
Africa 7 2 2
Arab States 119 1 1
Central and Western Asia 1 000 1 1
East Asia 401 1 1
Eastern Europe 501 1 1
Latin America and the Caribbean 17 2 1
North America 14 521 9 1
South Asia 460 4 3
South-East Asia and the Pacic 17 3 2
Western Europe 300 7 5
Online web-based 2 509 107 22
Africa 2 1 1
Central and Western Asia 107 1 1
East Asia 127 6 3
Eastern Europe 24 3 2
Latin America and the Caribbean 1 1 1
North America 1 572 61 2
South Asia 26 7 1
South-East Asia and the Pacic 494 7 4
Western Europe 155 20 7
Hybrid 6 273 5 4
Africa 180 1 1
South-East Asia and the Pacic 3 60 0 2 2
East Asia 2 493 2 1
Source: Owler database, annual reports and lings by platform companies to the Securities
and Exchange Commission of the United States.
The role of digital labour platforms in transforming the world of work
6
Table A1.5 Mergers and acquisitions in delivery platforms
Name of platform
Merger/
acquisition
Name of platform/company
(merged with/acquired by)
Date
of merger/
acquisition
Appetito24 Acquisition PedidosYa (acquired by Delivery Hero) 14.08.2017
Baedaltong Acquisition Delivery Hero 09.12. 2014
BGMENU.com Acquisition Takeaway.com (now Just Eat Takeaway.com) 23.02.2018
Canary Flash Acquisition Just Eat (now Just Eat Takeaway.com) 01.09.2019
Carriage Acquisition Delivery Hero 29.05.2017
Caviar Acquisition DoorDash 01.08.2019
Chef Shuttle Acquisition Bitesquad 23.06.2017
CitySprint Acquisition LDC 19.02.2016
Dáme Jídlo Acquisition Delivery Hero 09.01.2015
Daojia Acquisition Yum! China 17.05.2017
Delicious Deliveries Acquisition Bitesquad 10.10.2017
Deliveras Acquisition Delivery Hero 12.02.2018
Delivery.com Acquisition Uber 11.10.2019
Delyver Acquisition Big Basket 12.06.2015
Domicilios.com Acquisition iFood 08.04.2020
Doorstep Delivery Acquisition Bitesquad 28.08.2017
Eat24 Acquisition Grubhub 03.08.2017
Eats Media Acquisition delivery.com 26.08.2009
Eda.ua Acquisition Menu Group (UK) Limited 05.08.2019
Favor Acquisition HE Butt Grocery 15.02.2018
Feedr Acquisition Compass Group PLC 26.05.2020
Foodarena.ch Acquisition Takeaway.com (now Just Eat Takeaway.com) 22.06.2018
Foody Acquisition Delivery Hero 20.09.2017
Foodfox Acquisition Yandex 28.11.2017
Foodie Call Acquisition Bitesquad 10.10.2017
FoodNinjas Acquisition Velonto 04.2020
Foodonclick.com Acquisition Delivery Hero 05.2015
Foodora Acquisition Delivery Hero 09.2015
Foodpanda Acquisition Delivery Hero 10.12.2016
Foodpanda India Acquisition Ola 19.12.2017
FoodTime Acquisition Fave 24.05.2019
Appendix 1. Digital labour platforms: Estimates of workers, investments and revenues
7
Name of platform
Merger/
acquisition
Name of platform/company
(merged with/acquired by)
Date
of merger/
acquisition
Freshgora
Minority stake
investment
Meal Temple Group 2019
Gainesville2Go Acquisition Bitesquad 01.10.2017
HipMenu Acquisition Delivery Hero 08.2018
Honest Food Acquisition Delivery Hero 20.12.2019
Hungerstation.com Acquisition Foodpanda 09.08.2016
Lieferando Acquisition Takeaway.com (now Just Eat Takeaway.com) 10.04.2014
Menulog Acquisition Just Eat (now Just Eat Takeaway.com) 08.05.2015
Mjam Acquisition Delivery Hero 2012
MyDelivery Acquisition Meal Temple Group 26.02.2019
NetPincér hu Acquisition Foodpanda, then by Delivery Hero
12.2014
and 12.2016
respectively
PedidosYa Acquisition Delivery Hero 26.06.2014
Pyszne.pl Acquisition Lieferando, then by Just Eat Takeaway.com
23.03.2012
and
10.04.2014
respectively
Rickshaw Acquisition DoorDash 14.09.2017
SberMarket Acquisition Sberbank 30.11.2020
Seamless Acquisition Grubhub 01.05.2013
SkipTheDishes Acquisition Just Eat (now Just Eat Takeaway.com) 15.12.2016
Stuart Acquisition Geopost 07.05.2017
Takeaway.com and Just Eat Merger Just Eat Takeaway.com 23.04.2020
Talabat Acquisition Internet Rocked, then by Delivery Hero
02.2015
and 12.2016
respectively
Tapingo Acquisition Grubhub 25.09.2018
Uber Eats (India) Acquisition Zomato 21.01.2020
Waitr Acquisition Landcadia Holdings 16.05.2018
Woowa Bros Acquisition Delivery Hero 12.2020
Yemeksepeti Acquisition Delivery Hero 05.05.2015
YoGiYo Acquisition Delivery Hero 2014
Zakazaka Acquisition Mail.Ru Group 02.05.2017
Source: Crunchbase database, annual reports and platform websites.
Table A1.5 (cont’d)
The role of digital labour platforms in transforming the world of work
8
Table A1.6 Mergers and acquisitions in taxi platforms
Name of platform
Merger/
acquisition
Name of platform/company
(merged with/acquired by)
Date
of merger/
acquisition
99 Acquisition DiDi 03.01.2018
Beat Acquisition Intelligent Apps 16.02.2017
Careem Acquisition Uber 26.03.2019
Citybird Acquisition Felix 12.06.2018
Curb Acquisition Verifone 13.10.2015
Easy Taxi Acquisition Cabify 01.01.2017
Fasten Acquisition Vezet Group, then by MLU BV
02.03.2018
and 15.07. 2019
respectively
Flinc Acquisition Diamler 28.09.2017
FREE NOW Acquisition Intelligent Apps 26.07.2016
Savaree Acquisition Careem, then by Uber
30.03.2016
and 26.03.2019
respectively
Vezet Group Acquisition MLU BV 15.07.2019
Yandex.Taxi and Uber (Russia, CIS) Merger MLU BV 02.2018
Source: Crunchbase database, annual reports and platform websites.
Appendix 1. Digital labour platforms: Estimates of workers, investments and revenues
9
Table A1.7 Mergers and acquisitions in online web-based platforms
Name of platform
Merger/
acquisition
Name of platform/company
(merged with/acquired by)
Date
of merger/
acquisition
99designs Acquisition VistaPrint 05.10.2020
Applause Acquisition Vista Equity Partners 23.08.2017
AudioKite Acquisition ReverbNation 04.11.2016
Brandstack Acquisition DesignCrowd 20.12.2011
ClearVoice Acquisition Fiverr 13.02.2019
Codechef Acquisition Unacademy 18.06.2020
DesignCrowd
Acquisition
and merger
DesignBay
(since renamed DesignCrowd)
23.11.2009
Freelancer Technology Acquisition Music Freelancer.net 02.01.2019
Gengo Acquisition Lionbridge 16.01.2019
Guru Acquisition Emoonlighter 01.07. 2003
Indiez Acquisition GoScale 26.02.2020
Iwriter Acquisition Templafy 07.05.2019
Kaggle Acquisition Alphabet (includes Google) 07- 03.2017
Liveops Acquisition Marlin Equity Partners 01.12.2015
Mila Acquisition Swisscom 02.01.2013
MOFILM Acquisition You & Mr Jones 11.06.2015
Streetbee Acquisition BeeMyEye 16.01.2019
Test IO Acquisition EPAN Systems 21.05.2019
Topcoder Acquisition Appirio, then by Wipro Technologies
17.09. 2013
and
20.10.2016
respectively
Twago Acquisition Randstad 14.06.2016
VerbalizeIt Acquisition Smartling 19.05.2016
WeGoLook Acquisition Crawford & Company 06.12.2016
Xtra Global Acquisition Rozetta Corp 09.08.2016
Zooppa Acquisition TLNT Holdings SA 07.2019
Source: Crunchbase database, annual reports and platform websites.
Appendix 2
ILO interviews with digital platform companies
and analysis of terms of service agreements
2A. ILO interviews with digital platform companies
To understand the functioning of digital platform companies, interviews with representatives of
both location-based platforms and online web-based platforms were conducted. With regard to
location-based platforms, interviews with representatives of taxi and delivery platforms were
conducted, in collaboration with consultants, using a semi-structured questionnaire prepared
by the ILO. The consultants approached taxi and delivery platforms in their cities of operations,
requesting them to participate on the basis of a letter provided by the ILO. The interviews
collected information on the platforms’ business proles, operations and marketing strategies,
business model, recruitment practices and future strategies. However, only a few taxi platforms
(in Chile, Ghana, India and Kenya) and one delivery platform (in Ghana) agreed to the interviews,
which were conducted in person by the consultants or using video call by the ILO.
With regard to online web-based platforms, the ILO contacted about 30 platform companies
with signicant or growing presence at the country or regional levels, requesting them to
participate in the study. The ILO conducted interviews with eight such platform companies and
with one open-source platform (Apache Software Foundation). The interviews used semi-struc-
tured questionnaire, which were quite similar to those for the taxi and delivery platforms but
platform specic. In addition, the interviews sought information related to tasks, matching
process, algorithmic management, work evaluation and the platforms’ global operations. All
these interviews were conducted using video call, and follow-up meetings were held with
some platforms.
Table A2.1 lists the platform companies whose representatives were interviewed. The inter-
views were conducted between March 2019 and March 2020, and took between approximately
30 minutes and two hours.
The role of digital labour platforms in transforming the world of work
2
X
Table A2.1 Interviews conducted with digital platform companies
Platform company Person interviewed Coverage
A. Online web-based platforms
1. Clickworker CEO Berlin, Germany
2. Upwork Human Resources Manager Santa Clara, California, United States
3. Hsoub CEO London, United Kingdom
4. Worknasi CEO United Republic of Tanzania
5. Nabeesh CEO United Arab Emirates
6. Playment CEO Bengaluru, India
7. Crowd Analytix CEO Bengaluru, India
8. GoWorkABit
CEO (and member of the Sharing
Economy Association)
Estonia
9. Apache Foundation Board member (and Treasurer) Berlin, Germany
B. Location-based platforms
Taxi platforms
1. Uber Employee, operations department Accra, Ghana
2. Maramoja Employee, operations department Nairobi, Kenya
3. Uber
Employee, responsible for public policy
in East Africa
Nairobi, Kenya
4. Bolt Employee Nairobi, Kenya
5. Ola Employee, operations department New Delhi, India
6. Beat CEO Santiago, Chile
7. DiDi Director, corporate aairs Santiago, Chile
Delivery platforms
1. Okada Employee, operations department Accra, Ghana
Appendix 2. ILO interviews with digital platform companies and analysis of terms of service agreements
3
2B. Analysis of terms of service agreements
The terms of service agreements and other related documents of 31 platforms have been analysed
for this report. Chapters 2 and 5 draw on this analysis to understand the functioning of the platform
business model. Of these, 16 are online web-based platforms (4 freelance, 3 contest-based, 5 competitive
programming and 4 microtask) and 15 are location-based platforms, of which 7 are in the taxi sector and
8 are in the delivery sector, operating in a number of countries.
The online web-based platforms were chosen because of their coverage in the global microtask, freelance
and competitive programming surveys for this report, and some additional platforms were analysed
because of their prominence. All the location-based platforms analysed for the business model were iden-
tied in the country-specic worker surveys that were conducted in Africa (Ghana, Kenya and Morocco),
Asia (China, India and Indonesia), Central and Eastern Europe (Ukraine), Latin America (Argentina, Chile
and Mexico), and the Middle East (Lebanon). The exception is Deliveroo, which was considered for ana-
lysis because of its distinct characteristics compared to other delivery platforms, in order to enable a
comparison to be made with these other platforms. In addition, with respect to Grab and Gojek, the
terms of service agreements for Singapore were also analysed, as both these platforms are based there;
some key aspects of the agreements in Singapore may dier from those in other countries where these
platforms operate. The platform websites provide information related to the agreements and other
related documents (see table A2.2). Where it was not possible to obtain the information required, the
information from the country surveys and interviews conducted for the purposes of this report were
used.
1
The analysis focuses on the following aspects:
X
Contractual relationship: The terms of service agreements of both online web-based and loca-
tion-based platforms provide information on the contractual relationship. They all use terminology
which seeks to deny any relationship of employment between themselves and the platform users (see
tables A2.2 and A2.3 for more details).
X
Types of services: The websites of both online web-based and location-based platforms provide
information on the types of services available. Though the terms of service agreements also provide
such information, this is usually very brief compared to the details posted on the main websites. For
online web-based platforms in particular, information included in the main text of this report is also
based on the interviews conducted with representatives of the platforms.
X
Revenue model: The websites of online web-based platforms provide information on the dierent
types of fees charged to the various users (clients, workers and so on). These include fees for on-
boarding, commission fees or service charges for performing the tasks, transaction/withdrawal fees,
maintenance fees and cancellation charges. Some of the platforms also have optional fees, which
include fees for clients to mark projects as urgent or to highlight them so as to attract higher-quality
submissions, and fees for workers to obtain access to more job proposals and better listings. Some
also have a subscription model, and the amounts payable for the various subscriptions along with the
dierent services and benets they provide are available on the respective platform websites.
In the case of location-based platforms, the terms of service agreements of both transportation and
delivery platforms provide information on the types of fees charged, which almost invariably include
commission fees, cancellation fees and waiting-time fees, as well as various other surcharges, such as
for airport trips and tolls or for cleaning and maintenance services. The terms and conditions of loca-
tion-based platforms also provide information on surge pricing, specifying that the prices of services
vary according to demand and supply. Nevertheless, the agreements do not include information on
the exact amount of these fees. For some platforms, such as Bolt and Cornershop in Mexico, and Grab
or GrabFood and Gojek in Singapore, more precise information on commission fees can be found on
their websites (usually in the FAQ or support sections), but where such information was not available, it
was collected from the surveys of taxi drivers and delivery workers in the various countries, and from
the interviews conducted with restaurant and grocery shop owners.
1 The text indicates when that is the case.
The role of digital labour platforms in transforming the world of work
4
X
Recruitment and matching: Information on onboarding requirements and procedures was collected
from a number of sources. In some cases, privacy policies stipulate that users can access platforms
via third parties, such as social networking services, while for other platforms this information can be
deduced from the registration sections on their websites, which clearly give users the option to sign
up via third parties such as Google, Facebook or LinkedIn.
For instance, the privacy policies of both online web-based and location-based platforms provide
information on the documents needed to create an account. In the case of location-based platforms,
in particular, information on both the personal and technical requirements needed for joining as either
a driver or a courier (depending on whether the platform provides transportation or delivery services)
was collected from the registration sections on the platforms’ websites. Moreover, the support or
frequently asked question (FAQ) sections of online web-based platforms’ websites contain information
on verication and vetting procedures, which can include anything from ID verication via camera to
registration of user proles based on standards set by the platform. Information was also collected
from the country-specic surveys and the interviews conducted with companies for this report. Finally,
much of the information on the indicators used in assignment of work is based on an analysis of the
websites of 117 online labour platforms listed on Crunchbase.
X
Work processes and performance management: The websites of online web-based platforms con-
tain various sections relevant to work processes and performance management. There are sections
analysing platforms’ rating systems and the various levels assigned to workers based on such ratings,
and others referring to tools that the platforms make available to facilitate communication between the
parties and that enable them to track projects in real time (for example, in-app messaging systems, live
chat features and remote desktop apps). There are specic sections outlining the testing methods that
determine workers’ continued access to tasks and to the platforms. Information on the ratings systems
used by location-based platforms was more dicult to obtain. Their terms of service agreements are
usually silent on the matter and only a few platforms outline their rating systems on their websites.
Though most of the information concerning both online web-based and location-based platforms
was collected from their websites, the terms of service agreements were also relevant; they often
include clauses prohibiting activities such as communication between parties and payment being
made outside the platform, the use of automated methods (such as Google Translate in the case of
Appen) or the use of subcontractors. In the case of location-based platforms in particular, terms and
conditions include provisions on codes of conduct, customer service etiquette, and cancellation and
communication time frames.
Rules of platform governance
X
Account access/deactivation: Information on who can access the platforms and under what condi-
tions was mostly collected from terms of service agreements. In general, both online web-based and
location-based platforms deactivate user accounts when the users are considered to have breached
the terms of service agreements. That said, the power of platforms to deactivate accounts is often
broadly formulated. Many agreements contain clauses on platforms’ discretionary power to refuse
registration and deactivate accounts, often without the need to provide a reason or prior notice. In
the case of online web-based platforms, in particular, their websites often include sections with add-
itional information on deactivation and the reasons that might lead to it, which can include low ratings,
plagiarism or simply unoriginality of work, breach of codes of conduct (for instance, abuse of other
users), non-performance or submission of work which does not meet the platform’s or the client’s
specications or quality standards.
Appendix 2. ILO interviews with digital platform companies and analysis of terms of service agreements
5
X
Dispute resolution: Most information on the dispute resolution processes of both online web-based
and location-based platforms was collected from the terms of service agreements, which usually con-
tain entire sections dedicated to dispute resolution in which the governing law and jurisdiction are
clearly specied. In the case of online web-based platforms, such sections tend to be lengthier, given
that dispute resolution procedures usually take the form of arbitration proceedings, the conditions
of which are dened in detail by the platforms. In addition, online web-based platforms often include
dierent dispute resolution policies depending on the issue in question, and information on these
dierent policies is usually located on their websites. For instance, Upwork has dierent dispute reso-
lution procedures for hourly and xed-price contracts. Online web-based platforms also tend to have
separate dispute resolution processes for disputes concerning intellectual property. For location-based
platforms, the governing law and jurisdiction is usually that of the country where the services are being
provided, though in some cases it is that of another country, as is the case with certain countries where
Uber, Bolt and Glovo operate.
X
Data collection and usage: Obtaining information on the data that platforms collect and how they
process it was fairly straightforward, since such information is provided through privacy policies which
are uniformly structured. These policies, for both online web-based and location-based platforms,
clearly specify the kind of data collected, how it is collected, when and from where, as well as how they
use it and when and with whom they share it. Data can be collected directly (i.e. when users provide
it) or indirectly (i.e. by technological means such as cookies). Data collected directly from users varies
across platforms and can include a user’s contact and nancial details, specic identity documents,
criminal records, vehicle registration and insurance documents, or even more sensitive information
such as race, religion and marital status (the latter observed only for Grab).
Data collected indirectly also varies, and can include anything from usage data (such as browsing and
searching history, areas within the platform visited, duration of visits and number of clicks) and device
information (such as IP address, device identier and browser type), to data on communication between
users and other data stored in the user’s device (information from address books and calendars, or
even the names of other applications installed in the device). Such automatically collected information
also includes data relating to worker performance, such as their ratings and participation statistics,
while location-based platforms may even collect driving-related data such as real-time geolocation
and acceleration or braking data (as the privacy policies for Uber and Grab specify).
Apart from describing the kinds of data they collect, platforms’ privacy policies also outline the
various ways in which they use such data. For instance, they process user data to provide, enhance
and personalize their services, to understand how users use their services, to comply with the law, and
for automated decision-making (for instance the privacy policy of Uber species that it uses data to
match workers with clients, determine prices based on demand and suspend or deactivate accounts).
Although platforms may describe in detail the kinds of data they collect and the ways in which they
process it, they do not, however, clearly link data collection to data processing; in other words, it is not
always clear how a particular kind of data, such as location data, is used. Moreover, platforms share
user data with their business partners, with other users of the platform, and with an array of third-
party service providers including payment processors, insurance and nancial partners, advertising
companies, social networking services, cloud storage providers, research and marketing providers,
and law enforcement agencies. Privacy policies provide information on data protection, usually by
asserting that they abide by certain data protection laws, such as the European Union’s General Data
Protection Regulation, or that they ensure that any party with access to the platform’s data abides by
its privacy policy.
The role of digital labour platforms in transforming the world of work
6
X
Intellectual property rights (IPR): The terms of service agreements of both online web-based and
location-based platforms clearly state that any IPR rest with the platform. In the case of online web-
based platforms, however, it is not always clear in the terms and conditions which party has IPR over
the creative work produced via the platform. In most cases, IPR are transferred from the worker to the
client upon payment, though in some cases (such as Toptal) workers contractually assign any rights
in their work to the platform, which then transfers such rights to the clients upon payment. Certain
online web-based platforms also require that users sign non-disclosure agreements – as is the case
with private contests in 99designs and Designhill – while other platforms give clients the option to sign
such an agreement in return for a fee (such as Freelancer, PeoplePerHour). This information, wherever
possible, was collected from the platform websites, though often such information was not available.
X
Taxation: All the online web-based and location-based platforms under analysis specify that any prices
quoted on the platform are inclusive of taxes, and emphasize that the responsibility to determine and
pay taxes falls on the users (workers and clients). Nevertheless, there are some platforms that mention
in their terms of service that they deduct taxes from workers’ earnings. For instance, both Ola and
Zomato in India make deductions from proceeds as per the Income Tax Act, 1961. Freelancer recently
updated its “Fees and Charges” policy by adding a section on taxation, specifying that taxes will be
applied based on a user’s country of residence/registration. Similarly, Uber’s updated terms for Chile
state that Uber will transfer and collect the applicable taxes.
Appendix 2. ILO interviews with digital platform companies and analysis of terms of service agreements
7
X
Table A2.2 Online sources of platforms’ terms of service agreements
A) Online web-based platforms
Freelance platforms
Freelancer
User agreement: https://www.freelancer.com/about/terms
For data collection and usage see the privacy policy: https://www.freelancer.com/about/privacy
Revenue model
Fees and charges: https://www.freelancer.com/feesandcharges/
Membership: https://www.freelancer.com/membership/
Enterprise: https://www.freelancer.com/enterprise
Project management: https://www.freelancer.com/project-management/
Also see link under user agreement.
Ranking/ratings
Freelancer ratings: https://www.freelancer.com/support/General/freelancer-ratings
Freelancer rewards: https://www.freelancer.com/faq/topic.php?id=42
The preferred freelancer program: https://www.freelancer.com/support/freelancer/general/
the-preferred-freelancer-program?keyword=preferred
What is the preferred freelancer program?: https://www.freelancer.com/community/articles/
what-is-the-preferred-freelancer-program
Recruitment and matching
Sign up: https://www.freelancer.com/signup
Restrictions in some countries: https://www.freelancer.com/support/freelancer/General/
restrictions-in-some-countries
Know your customer and identity verication policy: https://www.freelancer.com/page.php?p=info%2Fkyc_policy
Also see links under user agreement and revenue model.
Work processes and performance management
Code of conduct: https://www.freelancer.com/info/codeofconduct
Communicating or paying outside Freelancer.com: https://www.freelancer.com/support/freelancer/General/
communicating-or-paying-outside-freelancer-com
Messaging my employers: https://www.freelancer.com/support/project/messaging-on-projects
Using the desktop app: https://www.freelancer.com/support/freelancer/project/
using-the-desktop-app?keyword=desktop%20a
Also see link under user agreement.
Rules of platform governance
Violations that lead to account closure: https://www.freelancer.com/support/freelancer/General/
violations-that-lead-to-account-closure
Reopening closed account: https://www.freelancer.com/support/Prole/can-i-reopen-my-closed-account
Milestone dispute resolution policy: https://www.freelancer.com/page.php?p=info%2Fdispute_policy
Also see links under user agreement and code of conduct.
The role of digital labour platforms in transforming the world of work
8
A) Online web-based platforms (cont’d)
Freelance platforms (cont’d)
PeoplePerHour
Terms and conditions: https://www.peopleperhour.com/static/terms
For data collection and usage see the privacy and cookies statement: https://www.peopleperhour.com/static/
privacy-policy
Revenue model
Loyalty programs for premium buyers: https://www.peopleperhour.com/premium-programme
What’s the dierence between PeoplePerHour and TalentDesk.io?: https://www.peopleperhour.com/blog/
product-platform/dierence-between-peopleperhour-and-talentdesk-io/
Also see link under terms and conditions.
Ranking/ratings
Understanding CERT: https://support.peopleperhour.com/hc/en-us/articles/205218587-Understanding-CERT
Recruitment and matching
Sign up: https://www.peopleperhour.com/site/register
Your freelancer application: https://support.peopleperhour.com/hc/en-us/
articles/205217827-Your-Freelancer-Application
Freelancer application got declined: https://support.peopleperhour.com/hc/en-us/
articles/360039120094-Freelancer-Application-got-declined?mobile_site=false
Verify your account: https://support.peopleperhour.com/hc/en-us/
articles/360001764608-Verify-your-Account?mobile_site=false
Prole policies: https://support.peopleperhour.com/hc/en-us/articles/205218177-Prole-policies
PeoplePerHour academy: https://www.peopleperhour.com/academy
Also see links under terms and conditions, and revenue model.
Work processes and performance management
WorkStream policies: https://support.peopleperhour.com/hc/en-us/articles/205218197-WorkStream-Policies
Also see links under terms and conditions, and prole policies.
Rules of platform governance
See links under terms and conditions, prole policies, and WorkStream policies.
Toptal
Terms and conditions: https://www.toptal.com/tos
For data collection and usage see the privacy policy: https://www.toptal.com/privacy
Revenue model
Enterprise: https://www.toptal.com/enterprise
The Toptal referral partners program: https://www.toptal.com/referral_partners
Frequently asked questions: https://www.toptal.com/faq
Recruitment and matching
See links under terms and conditions, privacy policy, and frequently asked questions.
Rules of platform governance
See links under terms and conditions, and frequently asked questions.
X
Table A2.2 (cont’d)
Appendix 2. ILO interviews with digital platform companies and analysis of terms of service agreements
9
A) Online web-based platforms (cont’d)
Freelance platforms (cont’d)
Upwork
User agreement: https://www.upwork.com/legal#useragreement
For data collection and usage see the privacy policy: https://www.upwork.com/legal#privacy
Revenue model
Pricing: https://www.upwork.com/i/pricing/
Freelancer plus: https://support.upwork.com/hc/en-us/articles/211062888-Freelancer-Plus
Enterprise: https://www.upwork.com/enterprise/
Featured jobs: https://support.upwork.com/hc/en-us/articles/115010712348-Featured-Jobs
Use connects: https://support.upwork.com/hc/en-us/articles/211062898-Use-Connects; https://support.upwork.
com/hc/en-us/articles/360057604814-11-24-FREE-Connects-to-Do-More-on-Upwork-
How to bring your own talent to Upwork: https://support.upwork.com/hc/en-us/
articles/360051696934-How-to-Bring-Your-Own-Talent-to-Upwork
Fee and ACH authorization agreement: https://www.upwork.com/legal#fees
Hourly, bonus, and expense payment agreement with escrow instructions: https://www.upwork.com/
legal#escrow-hourly
Fixed-price escrow instructions: https://www.upwork.com/legal#fp
Milestones for xed-price jobs: https://support.upwork.com/hc/en-us/
articles/211068218-Milestones-for-Fixed-Price-Jobs
PayPal fees and timing: https://support.upwork.com/hc/en-us/articles/211063978-PayPal-Fees-and-Timing
Payoneer fees and timing: https://support.upwork.com/hc/en-us/articles/211064008-Payoneer-Fees-and-Timing
M-Pesa fees and timing: https://support.upwork.com/hc/en-us/articles/115001615787-M-Pesa-Fees-and-Timing-
Wire transfer fees and timing: https://support.upwork.com/hc/en-us/
articles/211063898-Wire-Transfer-Fees-and-Timing
Direct to local bank fees and timing: https://support.upwork.com/hc/en-us/
articles/211060578-Direct-to-Local-Bank-Fees-and-Timing-
Direct to US bank (ACH) fees and timing: https://support.upwork.com/hc/en-us/
articles/227022468-Direct-to-US-Bank-ACH-Fees-and-Timing
Also see link under user agreement.
Ranking/ratings
Job success score: https://support.upwork.com/hc/en-us/articles/211068358-Job-Success-Score
Upwork’s talent badges: https://support.upwork.com/hc/en-us/articles/360049702614
Expert-vetted talent: https://support.upwork.com/hc/en-us/articles/360049625454-Expert-Vetted-Talent
Recruitment and matching
Sign up: https://www.upwork.com/signup/?dest=home
Eligibility to join and use Upwork: https://support.upwork.com/hc/en-us/
articles/211067778-Eligibility-to-Join-Upwork
Create a 100% complete freelancer prole: https://support.upwork.com/hc/en-us/
articles/211063188-Create-a-100-Complete-Freelancer-Prole
Application to join Upwork declined: https://support.upwork.com/hc/en-us/
articles/214180797-Application-to-Join-Upwork-Declined
Multiple account types: https://support.upwork.com/hc/en-us/articles/360001171768-Multiple-Account-Types
ID verication badge: https://support.upwork.com/hc/en-us/articles/360010609234-ID-Verication-Badge
Types of ID verication: https://support.upwork.com/hc/en-us/articles/360001176427-Types-of-ID-Verication
Sele ID review process: https://support.upwork.com/hc/en-us/articles/360001706047-Sele-ID-Review-Process
Also see links under user agreement, privacy policy, pricing, Freelancer plus, enterprise, featured jobs, and use connects.
Work processes and performance management
Upwork’s work diary: what it is and why use it: https://www.upwork.com/hiring/community/upworks-work-diary/
About the desktop app: https://support.upwork.com/hc/en-us/articles/211064038-About-the-Desktop-App
Upwork for clients app: https://support.upwork.com/hc/en-us/articles/211064028-Upwork-for-Clients-App
Upwork for freelancers app: https://support.upwork.com/hc/en-us/
articles/360015504093-Upwork-for-Freelancers-App
Use messages: https://support.upwork.com/hc/en-us/articles/211067768-Use-Messages
Video and voice calls: https://support.upwork.com/hc/en-us/articles/217698348-Video-and-Voice-Messaging
Freelancer education hub: https://www.upwork.com/hiring/education/getting-started-for-freelancers/
Readiness test: https://support.upwork.com/hc/en-us/articles/360047551134-Upwork-Readiness-Test
Also see link under user agreement.
Rules of platform governance
Freelancer violations and account holds: https://support.upwork.com/hc/en-us/
articles/211067618-Freelancer-Violations-and-Account-Holds
Non-disclosure agreements: https://support.upwork.com/hc/en-us/articles/211063608-Non-Disclosure-Agreements
Also see links under user agreement; hourly, bonus, and expense payment agreement with escrow instructions;
xed-price escrow instructions; and multiple account types.
X
Table A2.2 (cont’d)
The role of digital labour platforms in transforming the world of work
10
A) Online web-based platforms (cont’d)
Contest-based platforms
99designs
Terms of use: https://99designs.com/legal/terms-of-use
For data collection and usage see the privacy policy: https://99designs.com/legal/privacy
Revenue model
Pricing: https://99designs.com/pricing
What is a platform fee?: https://support.99designs.com/hc/en-us/articles/360022206031
What is a client introduction fee?: https://support.99designs.com/hc/en-us/articles/360022018152
Can I choose how much I pay for a contest?: https://support.99designs.com/hc/en-us/
articles/204760735-Can-I-choose-how-much-I-pay-for-a-contest-
What is a payout and how do I request one?: https://support.99designs.com/hc/en-us/
articles/204108819-What-is-a-payout-and-how-do-I-request-one-
100% money-back guarantee? for real?!: https://support.99designs.com/hc/en-us/
articles/204108729-100-Money-back-guarantee-For-real-
Also see link under terms of use.
Ranking/ratings
What are designer levels?: https://support.99designs.com/hc/en-us/articles/115002951643-What-are-designer-levels-
What are the benets for each designer level?: https://support.99designs.com/hc/en-us/articles/360022097311
What is top level status?: https://support.99designs.com/hc/en-us/articles/360001153443
Availability status and responsiveness score: https://support.99designs.com/hc/en-us/
articles/360000537386-Availability-Status-and-Responsiveness-Score
Recruitment and matching
How does 99designs’ application process work?: https://support.99designs.com/hc/en-us/
articles/360036552311-How-does-99designs-application-process-work-
What are 99designs’ quality standards?: https://support.99designs.com/hc/en-us/
articles/204862935-What-are-99designs-quality-standards-
What is identity verication?: https://support.99designs.com/hc/en-us/
articles/205460145-What-is-identity-verication-
Can I have more than one account?: https://support.99designs.com/hc/en-us/
articles/204761325-Can-I-have-more-than-one-account-?mobile_site=false
Best design awards: https://99designs.com/best-design-awards/
Also see links under terms of use, privacy policy, pricing, and ranking/ratings.
Work processes and performance management
Designer code of conduct: https://support.99designs.com/hc/en-us/articles/204109559-Designer-Code-of-Conduct
Designer resource center: https://99designs.com/designer-resource-center
Also see links under terms of use, pricing, and what are 99designs’ quality standards?
Rules of platform governance
Non-circumvention policy: https://support.99designs.com/hc/en-us/
articles/360022405192-Non-Circumvention-Policy
What’s a non-disclosure agreement (NDA)?: https://support.99designs.com/hc/en-us/
articles/204760785-What-s-a-non-disclosure-agreement-NDA-
Who owns what and when?: https://support.99designs.com/hc/en-us/articles/204761115-Who-owns-what-and-when-
Also see links under terms of use; what are 99designs’ quality standards?; can I have more than one account?; and
designer code of conduct.
X
Table A2.2 (cont’d)
Appendix 2. ILO interviews with digital platform companies and analysis of terms of service agreements
11
A) Online web-based platforms (cont’d)
Contest-based platforms (cont’d)
Designhill
Terms and conditions: https://www.designhill.com/terms-conditions
For data collection and usage see the privacy policy: https://www.designhill.com/privacy
Revenue model
Pricing guide: https://www.designhill.com/pricing/logo-design?services=contest
What is included in the enterprise package?: https://support.designhill.com/hc/en-us/
articles/360013633753-What-is-included-in-the-Enterprise-package-
Here is what you get when you go for subscription upgradation: https://www.designhill.com/design-blog/
here-is-what-you-get-when-you-go-for-subscription-upgradation/
Why should you upgrade your designer membership subscription?: https://www.designhill.com/design-blog/
why-should-you-upgrade-your-designer-membership-subscription/
What is a payout and how do I request one?: https://support.designhill.com/hc/en-us/
articles/115001380229-What-is-a-payout-and-how-do-I-request-one-
Can I choose how much I pay for a contest?: https://support.designhill.com/hc/en-us/
articles/115001213765-Can-I-choose-how-much-I-pay-for-a-contest-
How much do 1-to-1 projects cost to customers?: https://support.designhill.com/hc/en-us/
articles/115001517009-How-much-do-1-to-1-Projects-cost-to-customers-
Also see link under terms and conditions.
Recruitment and matching
Sign up: https://www.designhill.com/signup
How can I create an account?: https://support.designhill.com/hc/en-us/
articles/115001187805-How-can-I-create-an-account-
Can I have multiple accounts?: https://support.designhill.com/hc/en-us/
articles/115001186685-Can-I-have-multiple-accounts-
Also see links under terms and conditions; pricing guide; what is included in the enterprise package?; here is what
you get when you go for subscription upgradation; and why should you upgrade your designer membership
subscription?
Work processes and performance management
Designer code of conduct: https://support.designhill.com/hc/en-us/articles/115004513989
Free small business tools online: https://www.designhill.com/tools/
Also see links under pricing guide.
Rules of platform governance
Suspension policy: https://support.designhill.com/hc/en-us/articles/115004544629-Suspension-Policy
Concept originality policy: https://support.designhill.com/hc/en-us/articles/115004544729-
What if someone breaches my NDA?: https://support.designhill.com/hc/en-us/
articles/360013262574-What-if-someone-breaches-my-NDA-
Also see links under terms and conditions; can I have multiple accounts?; and designer code of conduct.
Hatchwise
Terms and conditions: https://www.hatchwise.com/terms-and-conditions
For data collection and usage see the privacy policy: https://www.hatchwise.com/privacy-policy
Revenue model
Contest pricing: https://www.hatchwise.com/contest-pricing
Our money back guarantee: https://www.hatchwise.com/guarantee
Also see link under terms and conditions.
Recruitment and matching
See link under contest pricing.
Work processes and performance management
The Hatchwise learning center: https://www.hatchwise.com/resources
Frequently asked questions: https://www.hatchwise.com/frequently-asked-questions
Rules of platform governance
See links under terms and conditions, and frequently asked questions.
X
Table A2.2 (cont’d)
The role of digital labour platforms in transforming the world of work
12
A) Online web-based platforms (cont’d)
Competitive programming platforms
CodeChef
Terms of service: https://www.codechef.com/terms
For data collection and usage see the privacy policy: https://www.codechef.com/privacy-policy
Revenue model
Refund policy: https://www.codechef.com/refund-policy
Guidelines: https://www.codechef.com/problemsetting
Setting: https://www.codechef.com/problemsetting/setting
Testing: https://www.codechef.com/problemsetting/testing
CodeChef business: https://business.codechef.com
Ranking/ratings
Rating mechanism: https://www.codechef.com/ratings
Recruitment and matching
Create your CodeChef account: https://www.codechef.com/signup
Code of conduct: https://www.codechef.com/codeofconduct
Also see links under guidelines, setting, testing, and ranking/ratings.
Work processes and performance management
How does CodeChef test whether my solution is correct or not?: https://discuss.codechef.com/t/
how-does-codechef-test-whether-my-solution-is-correct-or-not/332
Also see links under guidelines, setting, testing, and code of conduct.
Rules of platform governance
See links under terms of service, setting, testing, and code of conduct.
HackerEarth
Terms of service: https://www.hackerearth.com/terms-of-service/
For data collection and usage see the privacy policy: https://www.hackerearth.com/privacy/
Revenue model
Pricing: https://www.hackerearth.com/recruit/pricing/
Recruitment and matching
Sign up: https://www.hackerearth.com
Also see link under pricing.
Rules of platform governance
What is HackerEarth’s plagiarism policy?: https://help.hackerearth.com/hc/en-us/
articles/360002921714-What-is-HackerEarth-s-plagiarism-policy-
Also see link under terms of service.
HackerRank
Terms of service: https://www.hackerrank.com/terms-of-service
For data collection and usage see the privacy policy: https://www.hackerrank.com/privacy
Revenue model
Pricing: https://www.hackerrank.com/products/pricing/?h_r=pricing&h_l=header
Ranking/ratings
Scoring documentation: https://www.hackerrank.com/scoring
Recruitment and matching
Sign up: https://www.hackerrank.com/auth/signup?h_l=body_middle_left_button&h_r=sign_up
Also see links under terms of service and revenue model.
X
Table A2.2 (cont’d)
Appendix 2. ILO interviews with digital platform companies and analysis of terms of service agreements
13
A) Online web-based platforms (cont’d)
Competitive programming platforms (cont’d)
Kaggle
Terms of use: https://www.kaggle.com/terms
For data collection and usage see the privacy policy: https://www.kaggle.com/privacy
Revenue model
Meet Kaggle: https://www.kaggle.com/static/slides/meetkaggle.pdf?Host_Business
Also see link under terms of use.
Ranking/ratings
Kaggle progression system: https://www.kaggle.com/progression
Recruitment and matching
Sign in: https://www.kaggle.com/account/login?phase=startRegisterTab&returnUrl=%2Fterms
Also see link under terms of use.
Work processes and performance management
Community guidelines: https://www.kaggle.com/community-guidelines
Courses: https://www.kaggle.com/learn/overview
Rules of platform governance
Why has my account been blocked: https://www.kaggle.com/contact
Also see links under terms of use, meet Kaggle, and community guidelines.
Topcoder
Terms and conditions: https://www.topcoder.com/community/how-it-works/terms/
For data collection and usage see the privacy policy: https://www.topcoder.com/policy/privacy-policy
Revenue model
Enterprise: https://www.topcoder.com/enterprise-oerings/
Talent as a service: https://www.topcoder.com/enterprise-oerings/talent-as-a-service/
Ranking/ratings
Algorithm competition rating system: https://www.topcoder.com/community/competitive-programming/
how-to-compete/ratings
Development reliability ratings and bonuses: https://help.topcoder.com/hc/en-us/
articles/219240797-Development-Reliability-Ratings-and-Bonuses
Recruitment and matching
Log in to Topcoder: https://accounts.topcoder.com/member
Also see links under terms and conditions, revenue model, and algorithm competition rating system.
Work processes and performance management
Community code of conduct: https://www.topcoder.com/community/topcoder-forums-code-of-conduct/
Account policies: https://www.topcoder.com/thrive/articles/Topcoder%20Account%20Policies
Rules of platform governance
Cheating infractions and process: https://www.topcoder.com/thrive/articles/Cheating%20Infractions%20&%20
Process
Non-Disclosure Agreement (NDA): https://www.topcoder.com/thrive/articles/Non%20Disclosure%20Agreement%20
(NDA)
Also see links under terms and conditions, account policies, and community code of conduct.
X
Table A2.2 (cont’d)
The role of digital labour platforms in transforming the world of work
14
A) Online web-based platforms (cont’d)
Microtask platforms
Amazon
Mechanical
Turk
Participation agreement: https://www.mturk.com/participation-agreement
For data collection and usage see the privacy notice: https://www.amazon.com/gp/help/customer/display.html/
ref=footer_privacy?ie=UTF8&nodeId=468496
Revenue model
Pricing: https://www.mturk.com/pricing
Amazon Mechanical Turk pricing: https://requester.mturk.com/pricing
Amazon pay fees: https://pay.amazon.com/help/201212280
FAQs: https://www.mturk.com/worker/help
Also see link under participation agreement.
Ranking/ratings
Qualications and worker task quality: https://blog.mturk.com/
qualications-and-worker-task-quality-best-practices-886f1f4e03fc
New feature for the MTurk marketplace: https://blog.mturk.com/
new-feature-for-the-mturk-marketplace-aaa0bd520e5b
Also see link under FAQs.
Recruitment and matching
See links under participation agreement, privacy notice, revenue model, FAQs, pricing, and qualications and
worker task quality.
Work processes and performance management, and rules of platform governance
See links under participation agreement and FAQs.
Clickworker
For the terms and conditions and privacy policy see: https://www.clickworker.com/terms-privacy-policy/
Revenue model
Pricing: https://www.clickworker.com/pricing/
Clickworker FAQ: https://www.clickworker.com/faq/
Customer FAQ: https://www.clickworker.com/customer-faq/
Survey participants for online surveys: https://www.clickworker.com/
survey-participants-for-online-surveys/#fee-recommendations
Recruitment and matching
Qualications at Clickworker: https://www.clickworker.com/crowdsourcing-glossary/qualications-at-clickworker/
What does a Clickworker do?: https://www.clickworker.com/clickworker-job/#distribution
Clickworker starts new SMS account verication system: https://www.clickworker.com/2014/05/08/sms_verication/
Also see links under terms and conditions and privacy policy, Clickworker FAQ, and customer FAQ.
Work processes and performance management, and rules of platform governance
See links under terms and conditions and privacy policy, Clickworker FAQ, and customer FAQ.
Appen
Legal terms: http://f8-federal.com/legal/
Revenue model
Frequently asked questions: https://success.appen.com/hc/en-us/
articles/115000832063-Frequently-Asked-Questions
Also see link under legal terms.
Recruitment and matching
Glossary of terms: https://success.appen.com/hc/en-us/
articles/202703305-Getting-Started-Glossary-of-Terms#tainted_judgment
Guide to: test question settings (quality control): https://success.appen.com/hc/en-us/
articles/202702975-Test-Questions-Settings
Also see link under frequently asked questions.
Work processes and performance management
Guide to quality control page: https://success.appen.com/hc/en-us/articles/201855709
Also see links under legal terms and recruitment and matching.
Rules of platform governance
See links under legal terms, frequently asked questions, glossary of terms, guide to: test question settings (quality
control), and guide to: quality control page.
X
Table A2.2 (cont’d)
Appendix 2. ILO interviews with digital platform companies and analysis of terms of service agreements
15
A) Online web-based platforms (cont’d)
Microtask platforms (cont’d)
Microworkers
Terms of use: https://www.microworkers.com/terms.php
For data collection and usage see the privacy policy: https://www.microworkers.com/privacy.php
Revenue model
FAQ: https://www.microworkers.com/faq.php
Guidelines FAQ: https://www.microworkers.com/faq-guidelines.php
Also see link under terms of use.
Recruitment and matching, and work processes and performance management
See links under FAQ and guidelines FAQ.
Rules of platform governance
See links under terms of use, FAQ, and guidelines FAQ.
B) Location-based platforms
Taxi platforms
Bolt (Taxify) Ghana
Kenya
General terms for drivers: https://bolt.eu/en/legal/terms-for-drivers/
Terms and conditions for passengers: https://bolt.eu/en/legal/terms-for-riders/
For data collection and usage see:
Privacy policy for drivers: https://bolt.eu/en/legal/privacy-for-drivers/
Privacy policy for passengers: https://bolt.eu/en/legal/privacy-for-riders/
Revenue model
Commission fee: https://support.taxify.eu/hc/en-us/articles/115002946374-Commission-Fee
Driver paid wait time fees: https://support.taxify.eu/hc/en-us/
articles/360009458774-Driver-Paid-Wait-Time-Fees
Issue with a cancellation fee: https://support.taxify.eu/hc/en-us/
articles/360009457274?ash_digest=7dcc15def68f2cf475d9152c23ca169b44e11f2f
Damage or cleaning fee: https://support.taxify.eu/hc/en-us/
articles/360003640779-Damage-or-Cleaning-Fee
Also links under general terms for drivers and terms and conditions for passengers.
For Ghana: Driver payouts and commission: https://support.taxify.eu/hc/en-us/
articles/360001892993-Driver-Payouts-and-Commission
For Kenya: Driver balance and commission: https://support.taxify.eu/hc/en-us/
articles/360010650180-Driver-Balance-and-Commission
Ranking/ratings
Activity score calculation: https://support.taxify.eu/hc/en-us/
articles/115002946174-Activity-Score-Calculation
Acceptance rate calculation: https://support.taxify.eu/hc/en-us/
articles/360007690199-Acceptance-Rate-Calculation
Rating a passenger: https://support.taxify.eu/hc/en-us/
articles/115002907553-Rating-a-Passenger
How to leave a rating: https://support.taxify.eu/hc/en-us/articles/115002918034-Rating-a-Ride
Also see link under general terms for drivers.
Recruitment and matching
Becoming a Bolt driver: https://support.taxify.eu/hc/en-us/
articles/115003390894-Becoming-a-Bolt-Driver
Also see link under general terms for drivers.
Work processes and performance management, and rules of platform governance
See links under general terms for drivers, terms and conditions for passengers, activity score
calculation, and acceptance rate calculation.
X
Table A2.2 (cont’d)
The role of digital labour platforms in transforming the world of work
16
B) Location-based platforms (cont’d)
Taxi platforms (cont’d)
Careem Morocco
Terms of service: https://www.careem.com/en-ma/terms/
For data collection and usage see the privacy policy: https://www.careem.com/en-ma/privacy/
Revenue model
How do I refer a friend?: https://help.careem.com/hc/en-us/
articles/360001609527-How-do-I-refer-a-friend-
What do starting, time, distance, minimum and promised
fare mean?: https://help.careem.com/hc/en-us/
articles/360001400007-What-do-Starting-Time-Distance-Minimum-and-Promised-fare-mean-
Cancelling a ride: https://help.careem.com/hc/en-us/articles/360001600367-Cancelling-a-ride
Also see link under terms of service.
Recruitment and matching
Drive with Careem: https://drive.careem.com
How do I create a Careem account?: https://help.careem.com/hc/en-us/
articles/360001609507-How-do-I-create-a-Careem-account-
What is in ride insurance? https://help.careem.com/hc/en-us/
articles/115010884527-What-is-in-ride-insurance-
Also see link under terms of service.
Work processes and performance management
In-ride standards: https://help.careem.com/hc/en-us/articles/360001609427-In-ride-Standards
Also see link under terms of service.
Rules of platform governance
How does an account get blocked or suspended?: https://help.careem.com/hc/en-us/
articles/360001609447-How-does-an-account-get-blocked-or-suspended-
Also see link under terms of service.
Gojek Indonesia
Terms of use: https://www.gojek.com/terms-and-condition/
For data collection and usage see the privacy policy: https://www.gojek.com/privacy-policies/
Recruitment and matching
Join as our GoRide driver: https://www.gojek.com/help/mitra/
bergabung-menjadi-mitra-go-ride/
Also see links under terms of use and privacy policy.
Singapore
User terms of use: https://www.gojek.com/sg/terms-and-conditions/
Driver services agreement: https://www.gojek.com/sg/driver/agreement/
For data collection and usage see the privacy policy: https://www.gojek.com/sg/privacy-policy/
Revenue model
What is the Gojek Service Fee?: https://www.gojek.com/sg/help/?q=service+fee
Also see links under user terms of use and driver services agreement.
Ranking/ratings
How do ratings work?: https://www.gojek.com/sg/help/driver/service/#how-do-ratings-work
Recruitment and matching
What documents will I need to upload?: https://www.gojek.com/sg/help/driver/
account/#what-documents-will-i-need-to-upload
What can I drive with on Gojek: https://www.gojek.com/sg/help/driver/
account/#what-can-i-drive-with-on-gojek
GoFleet: https://www.gojek.com/sg/driver/goeet/
Also see links under user terms of use, driver services agreement, and privacy policy.
Work processes and performance management
Driver code of conduct: https://www.gojek.com/sg/help/driver/driver-code-of-conduct
Also see links under user terms of use and driver services agreement.
Rules of platform governance
Can I share my Gojek account with Others?: https://www.gojek.com/sg/help/driver/
account/#can-i-share-my-gojek-account-with-others
I was suspended due to inactivity: https://www.gojek.com/sg/help/driver/
account/#i-was-suspended-due-to-inactivity
Also see links under user terms of use and driver services agreement.
X
Table A2.2 (cont’d)
Appendix 2. ILO interviews with digital platform companies and analysis of terms of service agreements
17
B) Location-based platforms (cont’d)
Taxi platforms (cont’d)
Grab Indonesia
Terms of service: transport, delivery, and logistics: https://www.grab.com/id/en/terms-policies/
transport-delivery-logistics/
For data collection and usage see the privacy policy: https://www.grab.com/id/en/terms-policies/
privacy-policy/
Revenue model
Grab referral programme terms and conditions: https://www.grab.com/id/en/pax-refer-friend/
privacy/
GrabFood – partner with us: https://www.grab.com/id/en/merchant/food/
Also see link under terms of service.
Recruitment and matching
Register now: https://www.grab.com/id/en/driver/transport/car/
Also see links under terms of service, privacy policy and GrabFood partner with us.
Singapore
Terms of service - transport, delivery and logistics: https://www.grab.com/sg/terms-policies/
transport-delivery-logistics/
For data collection and usage see the privacy policy: https://www.grab.com/sg/terms-policies/
privacy-policy/
Revenue model
FAQs: https://www.grab.com/sg/driver/transport/car/faq/
Updated cancellation policy from 25 Mar 2019: https://help.grab.com/passenger/en-
sg/115008318688; https://www.grab.com/sg/passenger-cancellation-fees/
I was charged a cancellation fee: https://help.grab.com/passenger/
en-sg/115005276987-I-was-charged-a-cancellation-fee
What are grace waiting periods and waiting fees: https://help.grab.com/passenger/
en-sg/360035841031-What-are-grace-waiting-periods-and-waiting-fees
#AskGrab: where does the Merchant commission go?: https://www.grab.com/sg/blog/
askgrab-where-does-the-merchant-commission-go/
How do I refer?: https://www.grab.com/sg/gfm-referral/
Also see link under terms of service.
Ranking/ratings
Acceptance and cancellation rating: https://help.grab.com/driver/
en-sg/115013368427-Acceptance-and-Cancellation-rating
Recruitment and matching
Drive: https://www.grab.com/sg/driver/drive/
Deliver: https://www.grab.com/sg/driver/deliver/
Drive with Grab using your own car in 4 Steps: https://www.grab.com/sg/
drive-with-grab-using-your-own-car/
Also see links under terms of service and privacy policy.
Work processes and performance management
How to improve my star rating: https://help.grab.com/driver/
en-sg/115015441428-Driver-Rating-How-is-this-calculated
Also see link under terms of service.
Little Kenya
Terms and conditions: https://www.little.bz/ke/tnc.php
X
Table A2.2 (cont’d)
The role of digital labour platforms in transforming the world of work
18
B) Location-based platforms (cont’d)
Taxi platforms (cont’d)
Ola India
Subscription agreement: https://partners.olacabs.com/public/terms_conditions
Terms and conditions: https://www.olacabs.com/tnc?doc=india-tnc-website
For data collection and usage see the privacy policy: https://www.olacabs.com/
tnc?doc=india-privacy-policy
Revenue model
Why is cancellation fee charged: https://help.olacabs.com/support/dreport/208298769
Also see links under subscription agreement and terms and conditions.
Ranking/ratings
How can I rate a ride?: https://help.olacabs.com/support/dreport/205098571
Recruitment and matching
Drive with Ola: https://partners.olacabs.com/drive
Lease a car: https://partners.olacabs.com/lease
Ola rolls out ‘Chalo Bekar’ comprehensive insurance program for
its driver partners: https://www.olacabs.com/media/in/press/
ola-rolls-out-chalo-bekar-comprehensive-insurance-program-for-its-driver-partners
Ola oers coverage of up to Rs. 30,000 for driver-partners and their spouses aected by COVID-19;
also brings free medical help for their families: https://www.olacabs.com/media/in/press/
ola-oers-coverage-of-up-to-rs-30000-for-driver-partners-and-their-spouses-aected-by-
covid-19-also-brings-free-medical-help-for-their-families
Also see link under subscription agreement.
Uber Argentina
Chile
Ghana
India
Kenya
Lebanon
Mexico
Morocco
United States
For the general terms of use, privacy notice, and general community guidelines see: https://
www.uber.com/legal/en/ (“Uber legal”) – select the relevant policy in the link and then the
relevant country.
Revenue model
For Ghana: tracking your earnings: https://www.uber.com/gh/en/drive/basics/
tracking-your-earnings/
Wait time fees: https://help.uber.com/riders/article/
wait-time-fees?nodeId=5960f72c-802a-4b61-a51c-2c9498c3b041
Am I charged for cancelling an Uber ride?: https://help.uber.com/riders/article/
am-i-charged-for-cancelling-an-uber-ride-?nodeId=5f6415dc-dfdb-4d64-927a-66bb06bc4f82
Also see links under Uber legal.
Recruitment and matching
Vehicle requirements: https://help.uber.com/driving-and-delivering/article/
vehicle-requirements?nodeId=2ddf30ca-64bd-4143-9ef2-e3bc6b929948
What does the background check look for: https://help.uber.com/driving-and-delivering/article/
what-does-the-background-check-look-for?nodeId=ee210269-89bf-4bd9-87f6-43471300ebf2
Why am I being asked to take a photo of myself?: https://help.uber.com/driving-and-delivering/
article/why-am-i-being-asked-to-take-a-photo-of-myself--?nodeId=7fa8a60d-cf6f-49ac-9a50-
b4bf6a3978ef
Getting a trip request: https://help.uber.com/driving-and-delivering/article/
getting-a-trip-request?nodeId=e7228ac8-7c7f-4ad6-b120-086d39f2c94c
When and where are the most riders?: https://help.uber.com/driving-and-delivering/article/
when-and-where-are-the-most-riders?nodeId=456fcc51-39ad-4b7d-999d-6c78c3a388bf
Insurance: https://help.uber.com/driving-and-delivering/article/
insurance-?nodeId=a4afb2ed-75af-4db6-8fdb-dccecfcc3fd7
Also see link under Uber legal.
Work processes and performance management
Can I use other apps or receive personal calls while online?: https://help.uber.com/
driving-and-delivering/article/can-i-use-other-apps-or-receive-personal-calls-while-online---
?nodeId=a5a7c0c7-da4b-46af-a180-7ad1d2590234
Also see link under Uber legal.
X
Table A2.2 (cont’d)
Appendix 2. ILO interviews with digital platform companies and analysis of terms of service agreements
19
B) Location-based platforms (cont’d)
Delivery platforms
Cornershop Mexico
Terms of use: https://cornershopapp.com/en/terms
For data collection and usage see the privacy policy: https://cornershopapp.com/es-mx/privacy
Revenue model
Cornershop for stores: https://cornershopapp.com/en/stores?adref=customer-landing
Cornershop Pop, la membresía de envíos gratis ilimitados: https://blog.cornershop.mx/
cornershop-pop-la-membresia-de-envios-gratis-ilimitados-mx/
Also see link under terms of use.
Deliveroo France
Conditions generales de prestation de service de Deliveroo: https://deliveroo.fr/en/legal
For data collection and usage see politique de condentialité de Deliveroo France: https://deliveroo.
fr/en/privacy
Revenue model
Comment suis-je payé ?: https://riders.deliveroo.fr/fr/support/nouveaux-livreurs-partenaires/
vous-etes-payes-pour-chaque-livraison-eectuee.-les
Also see links under conditions generales de prestation de service de Deliveroo, and politique
de condentialité de Deliveroo France.
Recruitment and matching
Ride with us: https://deliveroo.fr/en/apply
Nouveaux livreurs partenaires: https://riders.deliveroo.fr/fr/support/
nouveaux-livreurs-partenaires
Gérer votre entreprise: https://riders.deliveroo.fr/fr/support/gerer-votre-entreprise
Assurances Deliveroo: https://riders.deliveroo.fr/fr/support/toutes-vos-assurances-deliveroo
Also see links under conditions generales de prestation de service de Deliveroo, and politique
de condentialité de Deliveroo France.
United Kingdom
Terms of service: https://deliveroo.co.uk/legal
Scooter supplier agreement: https://old.parliament.uk/documents/commons-committees/work-
and-pensions/Written_Evidence/Deliveroo-scooter-contract.pdf
For data collection and usage see:
Privacy policy: https://deliveroo.co.uk/privacy
UK rider privacy policy: https://rider.deliveroo.co.uk/rider-privacy#information-collected
Revenue model
Refer a friend: https://riders.deliveroo.co.uk/en/refer
Fees: https://riders.deliveroo.co.uk/en/support/fees
Invoices, refunds and payments (Deliveroo restaurants): https://help.deliveroo.com/en/
collections/2612291-5-invoices-refunds-and-payments
FAQ: https://deliveroo.co.uk/faq
Also see links under terms of service, privacy policy, and UK rider privacy policy.
Recruitment and matching
Ride with us: https://deliveroo.co.uk/
apply?utm-campaign=ridewithus&utm-medium=organic&utm-source=landingpage
New riders: https://riders.deliveroo.co.uk/en/support/new-riders
Become a Deliveroo partner: https://restaurants.deliveroo.com/en-gb/
Orders: https://riders.deliveroo.co.uk/en/support/orders
Kit: https://riders.deliveroo.co.uk/en/support/kits
Insurance: https://riders.deliveroo.co.uk/en/support/insurance
Also see links under terms of service, privacy policy, and UK rider privacy policy.
Work processes and performance management
Can someone else work on my behalf (substitute)?: https://riders.deliveroo.co.uk/en/support/
account/substitute
Also see links under terms of service, privacy policy, and UK rider privacy policy.
Rules of platform governance
My supplier agreement was terminated. Can I dispute Deliveroo’s decision?: https://riders.deliveroo.
co.uk/en/support/account/request-sa-review
Also see links under terms of service, privacy policy, and UK rider privacy policy.
X
Table A2.2 (cont’d)
The role of digital labour platforms in transforming the world of work
20
B) Location-based platforms (cont’d)
Delivery platforms (cont’d)
Glovo Argentina
Chile
Kenya
General terms of use and contracting: https://glovoapp.com/es-ar/legal/terms/ –
Select the
relevant country in the link.
For data collection and usage see the privacy and data protection policy: https://glovoapp.com/
es-ar/legal/privacy/
–
Select the relevant country in the link.
Revenue model
Glovo business: https://business.glovoapp.com
Help and support: https://glovoapp.com/en/faq/
Also see link under general terms of use and contracting.
Ranking/ratings
For Kenya:
Peak slots: https://glovers.glovoapp.com/ke/tips/peak-slots
Excellence score: https://glovers.glovoapp.com/ke/faq/excellence-score
Recruitment and matching
For Kenya:
How to book slots: https://glovers.glovoapp.com/ke/basics/how-to-book-slots
About insurance: https://glovers.glovoapp.com/ke/safety/about-insurance
Jumia Food Ghana
Kenya
Morocco
Terms and conditions:
Ghana: https://food.jumia.com.gh/contents/terms-and-conditions.htm
Kenya: https://food.jumia.co.ke/contents/terms-and-conditions.htm
Morocco: https://food.jumia.ma/contents/terms-and-conditions.htm
For data collection and usage see the privacy policy:
Ghana: https://food.jumia.com.gh/contents/privacy.htm
Kenya: https://food.jumia.co.ke/contents/privacy.htm
Morocco: https://food.jumia.ma/contents/privacy.htm
Revenue model
Prime:
Kenya: https://food.jumia.co.ke/prime
Morocco: https://food.jumia.ma/prime
Also see links under terms and conditions.
Rappi Argentina
Chile
Mexico
For the terms and conditions, and the privacy policy see: https://legal.rappi.com/colombia/
terminos-y-condiciones-de-uso-de-plataforma-rappi-2/ – Select the relevant country in the link
and then the relevant policy.
Revenue model
Prime:
Argentina: https://www.rappi.com.ar/prime
Chile: https://www.rappi.cl/prime
Mexico: https://www.rappi.com.mx/prime
Also see link under terms and conditions.
Swiggy India
Terms and conditions: https://www.swiggy.com/terms-and-conditions
For data collection and usage see the privacy policy: https://www.swiggy.com/privacy-policy
Revenue model
Cancellation and refund policy: https://www.swiggy.com/refund-policy
Also see link under terms and conditions.
Recruitment and matching
Benets of being a Swiggy pick-up and delivery partner: https://ride.swiggy.com/en/
tiny-start-up-to-number-one-swiggys-growth-story-1
X
Table A2.2 (cont’d)
Appendix 2. ILO interviews with digital platform companies and analysis of terms of service agreements
21
B) Location-based platforms (cont’d)
Delivery platforms (cont’d)
Uber Eats Argentina
Chile
Mexico
Kenya
Uber Eats community guidelines: https://www.uber.com/legal/en/ – Select the relevant policy in
the link and then the relevant country.
Zomato India
Terms of service: https://www.zomato.com/conditions
Delivery partner terms and conditions: https://zomato.runnr.in/delivery-partner-tandc.html
For data collection and usage see the privacy policy: https://www.zomato.com/privacy
Recruitment and matching
Summary of all of Zomato’s COVID-19 related initiatives: https://www.zomato.com/blog/
covid-19-initiatives
Also see link under delivery partner terms and conditions.
Work processes and performance management
Guidelines and policies: https://www.zomato.com/policies
Also see link under delivery partner terms and conditions.
Rules of platform governance
See links under terms of service, delivery partner terms and conditions, and guidelines and
policies.
X
Table A2.2 (cont’d)
The role of digital labour platforms in transforming the world of work
22
X
Table A2.3 Terminology used to identify platform users in terms of service agreements
Platform
Country
of registration
Worker Client/customer General
A) Online web-based platforms
Freelance platforms
Freelancer United States Seller
Entrant
Buyer User
PeoplePerHour United Kingdom Freelancer Buyer User
Toptal United States Freelancer – User
Upwork United States Freelancer Client User
Contest-based platforms
99designs United States Designers Customer Users
Designhill India Designer Customer User
Hatchwise United States Creative
Designer
Writer
Contest holder
Project holder
Session holder
Client
User
Competitive programming platforms
CodeChef India – – User
HackerEarth United States Candidate User
Recruiter
–
HackerRank United States Hacker – –
Kaggle United States Participant user Host user User
Topcoder United States Contestant Competition sponsor User
Microtask platforms
Amazon Mechanical Turk United States Workers Requesters –
Clickworker Germany Clickworker Service requester User
Appen Australia Contributor Task author
Customer
User
Microworkers United States Workers Employers User
Appendix 2. ILO interviews with digital platform companies and analysis of terms of service agreements
23
Platform Survey country Worker Client/Customer
Businesses listing their products
on the delivery platform
General
B) Location-based platforms
Taxi platforms
Bolt (Taxify) Ghana Driver Passenger n/a –
Kenya
Careem Morocco Captains Users n/a –
Gojek Indonesia Service provider – n/a Users
Singapore Transportation provider Passenger n/a User
Grab Indonesia Third party provider
(driver/ delivery partner)
Passenger n/a User
Singapore Third party provider
(driver/ delivery partner)
Passenger n/a User
Little Kenya Service provider Customer n/a User
Ola India Driver
Transport service provider
Customer n/a User
Uber Argentina Tercero proveedor – n/a Usuario
Chile Tercero proveedor – n/a Usuario
Ghana Third party provider – n/a User
India Third party provider
Driver partner
Rider n/a User
Kenya Third party provider – n/a User
Lebanon Third party provider – n/a User
Mexico Tercero proveedor – n/a Usuario
Morocco Prestataires tiers – n/a Utilisateur
United States Third party provider – n/a User
X
Table A2.3 (cont’d)
The role of digital labour platforms in transforming the world of work
24
Platform Survey country Worker Client/Customer
Businesses listing their products
on the delivery platform
General
B) Location-based platforms (cont’d)
Delivery platforms
Cornershop Mexico
Contractors: shoppers,
deliverers
– Retailers User
Deliveroo
France Livreur partenaire Client Restaurants partenaires –
United Kingdom
Rider
Suppliers
Customer Partners –
Glovo
Argentina
Glovers Clientes Comercios Usuarios
Chile
Kenya Mandataries Users Merchants Users
Jumia Food
Ghana –
Eligible user (for Jumia
Prime)
Partner restaurant –
Kenya –
Eligible User (for Jumia
Prime)
Partner restaurant –
Morocco –
Utilisateurs éligibles
(pour Jumia Prime)
Restaurant partenaire –
Rappi
Argentina Rappitenderos Usuarios Comercios aliados –
Chile RappiRepartidor(es) Consumidor(es) Comercios aliados Usuario
Mexico Comisionista Consumidor(es) – Usuario
Swiggy India
Third party service providers i.e.
pick–up and delivery partners
Buyer/s Merchant/s User
Uber Eats
Argentina Socio repartidor
Comensal Socio restaurantero
Usuario
Chile Socio repartidor Comensal
Socio restaurantero
Usuario
Mexico Socio repartidor
Comensal Socio restaurantero
Usuario
Kenya Delivery partner Consumer Restaurant partner User
Zomato India Delivery partner - Restaurant partner User
n/a = not applicable.
X
Table A2.3 (cont’d)
Appendix 3
ILO interviews with businesses and clients
To understand the opportunities and challenges arising as a result of the digital transform-
ations in the world of work, the ILO conducted interviews with various types of businesses and
clients. These included IT companies, start-up companies, business clients who use delivery or
taxi platforms, and business process outsourcing (BPO) companies that provide digital services.
Table A3.1 lists the companies and individuals interviewed. The processes followed to identify
businesses for interviews are elaborated below. The interviews were conducted between March
2019 and March 2020 and lasted between 30 minutes and two hours.
3A. IT companies
A number of IT companies in India were contacted for interviews to understand whether they
were using digital labour platforms and which strategies they were adopting to integrate into
the digital economy. Despite intense eorts, only two IT companies agreed to an interview,
with the support of the ILO Country Oce for India. The interviews were conducted through
in-person meetings and using video call with company executives. The semi-structured inter-
views covered a range of issues, including how the digital transformations are aecting the
IT sector, strategies used by the companies to adapt to changing digital technologies, their
operations, recruitment strategies, performance management, productivity and innovation,
and how digitalization is shaping their strategic thinking and future business strategy.
3B. Start-up companies providing tools
and/or complementary products and AI services
Digital labour platforms use various applications, tools and complementary products to provide
services to businesses and cater to their needs. Based on an analysis of the digital labour
platforms covered in the report, such as Glovo, PeoplePerHour, Upwork and 99designs, and
using a website proling tool (BuiltWith), the dierent applications, tools and products used
by platforms could be identied. Many of these were related primarily to productivity, commu-
nication and collaboration (e.g. Slack, Zoom, Skype, Dropbox), payments (e.g. PayPal, Venmo),
video and audio (e.g. YouTube), and translation (e.g. Google Translate). The embedded apps
identied related to data analytics and customer outreach (e.g. CrazyEgg, Notice Board) and
advertising (Twitter analytics and Google analytics). Based on this, some 35 start-up businesses
were contacted, most of them based in India and the United States, 12 of which agreed to the
interviews.
In addition, the Indian Institute of Information Technology, Bengaluru’s incubation centre, also
helped to identify digital technology start-ups which were developing applications and tools
for digital labour platforms or traditional companies; ve start-ups were identied through this
process. In all, 17 start-up companies were interviewed, but only 10 of them are included in
the analysis as the remaining seven either did not provide in-depth insights that could be used
for this report or did not want their companies to be used for analysis. The questionnaire was
semi-structured and included questions on motivations behind the launch of such companies,
how they grew, their regional or global focus, their opportunities and challenges, and their
future growth strategies. Some of the interviews were conducted in person during a mission
to India, while others were carried out over Skype or Zoom.
The role of digital labour platforms in transforming the world of work
2
X
Table A3.1 ILO interviews conducted on the subject of digital labour platform experience
Interviewee category Person(s) interviewed Location
A. IT companies
1. Wipro – Head of Technovation Centre
– Head of Public Policy
and Corporate Aairs
Bengaluru, India
2. Infosys Senior Manager,
Corporate Strategy Planning
Bengaluru, India
B. Start-up companies providing tools and/or complementary products
1. Cloudinary Head of Marketing San Francisco, United States
2. Crazyegg Public Relations Manager San Francisco, United States
3. Rytangle CEO Bengaluru, India
4. Krittur Technology CEO Bengaluru, India
5. Notice Board – CEO
– Chief Technical Ocer
Bengaluru, India
6. Bionic Yantra CEO Bengaluru, India
7. Vision Empower CEO Bengaluru, India
8. Jordan
1
CEO San Francisco, United States
9. Ever Labs CEO Cherkasy, Ukraine
10. 300 Brains CTO Warsaw, Poland
C. Clients or small businesses using delivery platforms (number of interviews conducted per country)
1. Restaurants Owner Ghana (4); Kenya (3); Indonesia (3);
Lebanon (6); Morocco (6); Ukraine (5)
2. Retail businesses
(small shops, grocery stores)
Owner Ghana (6); Indonesia (9); Kenya (1)
3. Corporate companies Owner Kenya (4)
D. BPO companies
1
1. HN, AT, CF, CCI, SS, IN (6) – CEO
– CTO
– Head of Operations
– Chief Strategic Ocer
Nairobi and Mombasa, Kenya
2. TR, CO, FS, ASAP, GIIP (5) – CEO
– Co-founder
Bengaluru and New Delhi, India
E. Customers using taxi and delivery platforms (number of interviews conducted per country)
1. Individual customers using
delivery platforms
Chile (4)
2. Individual customers using taxi
and delivery platforms
Ghana (10)
India (14)
3. Interviews with individual
customers using taxi platforms
Kenya (5)
1
The names of the start-up and BPO companies cited have been changed to preserve their anonymity.
Appendix 3. ILO interviews with businesses and clients
3
3C. Clients or small businesses using delivery platforms
Interviews with clients and small businesses were conducted in collaboration with consultants based on
a semi-structured questionnaire prepared by the ILO. Based on the survey of app-based delivery workers
conducted in each country, restaurants or small businesses were identied that were using delivery
platforms for their activities. The potential interviewees were approached by the consultants in all the
countries. However, only in some countries (Ghana, Indonesia, Kenya, Lebanon, Morocco and Ukraine),
were the clients and small businesses willing to be interviewed. In the other countries considered –
Argentina, Chile, China, India and Mexico – contacting and engaging with these businesses proved more
dicult, and the interviews could not be carried out. The interviews largely focused on the interviewees’
motivations for using digital platforms to conduct their business, and the opportunities and challenges
encountered. All the interviews were conducted in person by the consultants in the respective countries.
3D. Business process outsourcing (BPO) companies
The approach adopted for BPO companies in India and Kenya was dierent. In India, BPO companies
were identied with the support of the ILO Country Oce for India and through contacts provided by
researchers at the Indian Institute of Information Technology in Bengaluru, one of the collaborating part-
ners for the study. The companies concerned were either in the process of transforming their businesses
to cater to new digital needs or setting up new BPO companies to service big technology companies.
In Kenya, the ILO collaborated with a consultant who helped in establishing contact with the selected
BPO companies. The consultant had earlier undertaken a study on BPO companies for researchers at the
Oxford Internet Institute and was therefore familiar with the sector. This was important and instrumental
in establishing contacts with these companies and conducting the interviews. All the interviews were
carried out in person by the consultant in collaboration with the ILO team (either in person during the
mission or through Skype).
The semi-structured interviews in both countries focused on the business strategies of the companies
as they transition towards providing digital services; the nature of the services provided and their dier-
ences in comparison to those provided previously; the implications of the transition on the company’s
human resources and skill sets; how the company improves its productivity; and their strategies for
future business development.
3E. Customers using taxi and delivery platforms
The interviews with customers using taxi and delivery platforms were conducted in collaboration with
consultants in the respective countries and were based on a semi-structured questionnaire prepared by
the ILO. The consultants in Chile, Ghana, India and Kenya identied individuals who were willing to share
their experiences and motivations for using these platforms. The interviews focused on the motivations
for using digital platforms and their benets. All interviews were conducted in person by the consultants
in the respective countries.
Appendix 4
ILO surveys, interviews and statistical analysis
4A. ILO surveys and interviews
4A.1 Selected country surveys of taxi drivers
and delivery workers (2019–20)
The survey of delivery workers was conducted in 11 countries, that of taxi drivers in nine coun-
tries. While in some countries the survey was restricted to one city, in others several cities were
targeted (see table A4.1), based on the spread of platform companies across cities and the
feasibility of conducting surveys in multiple cities.
The surveys of app-based and traditional taxi drivers and delivery workers were based on
four questionnaires developed by the ILO (app-based taxi, traditional taxi, app-based delivery,
traditional delivery). The structure of the questionnaires was similar, with some adaptations
depending on the sector. Each questionnaire included questions on the respondent’s socio-
demographic background; work history and information about other work they are engaged
in; and detailed questions related to their work as taxi drivers or delivery workers, including
working time, income, work-related expenditures, social security coverage, income security,
autonomy and control, perceptions of work and workplace solidarities. While the majority
of questions were quantitative, some were qualitative and allowed for open-ended textual
answers. In addition, the questionnaire provided space for additional notes and comments,
and enumerators were encouraged to note any information or statements that they considered
valuable for the analysis.
The surveys were implemented in collaboration with consultants from the respective coun-
tries. In Argentina, the survey was coordinated by the ILO Country Oce for Argentina and
conducted by FLASCO, Buenos Aires. In all other countries, the surveys were conducted in col-
laboration with researchers or research institutes. The questionnaires were adapted to the local
context in consultation with the consultants and translated into the local language(s) where
necessary. In each country, a pilot test of the questionnaires was carried out to identify potential
issues and to further rene the questionnaire where necessary before the nal data collection
started. The interviews were conducted using computer-assisted personal interviewing (CAPI)
with inbuilt validation rules using mobile devices (cell phones, tablets).
1
Prior to commencing the pilot tests and eldwork, the consultants and enumerators involved
in this study were instructed on the relevance of the project and trained in understanding
each question, capturing each variable with detailed querying techniques to gain extensive
in-depth information related to the respondent’s situation, and using the survey form on their
device. This training was provided using video call in all countries, except for Kenya, where it
was provided in-person during a mission of the ILO research team.
1 The surveys were implemented on KoBoToolbox, an open-source survey tool: https://www.kobotoolbox.org/.
The role of digital labour platforms in transforming the world of work
2
As there are no ocial statistics on these types of platform workers, including their number and charac-
teristics, there was no sampling base from which a random sample could be drawn. In this context, the
primary objective was to achieve a sample that would be as representative of the target population as
possible. The target population consisted of any worker aged 18 years or older who had been working
in the sector for at least three months. The criterion of working in the sector for three months was used
to ensure that the worker could provide meaningful responses.
2
To ensure heterogeneity within the sample, the interviews were conducted in dierent neighbourhoods
of the city (see gure A4.1), on dierent days of the week (including weekends) and during dierent
times of the day. Prior to starting the eldwork, the main platform companies operating in each city
were identied. Enumerators were advised to capture respondents who were registered with dierent
platform companies. However, in some countries, such as Lebanon, it proved dicult to nd workers
who were registered with dierent platform companies as one platform company was clearly dominating
the market (see table A4.2).
2 In Argentina, the sample was limited to workers aged 16 or above, with at least one month of experience in the app-based
delivery sector.
X
Table A4.1 Number of observations for country surveys of taxi drivers
and delivery workers, by interview city
Country
Taxi Delivery
App-based Traditional App-based Traditional
Argentina – – Buenos Aires (300) –
Chile Santiago (126) Santiago (147) Santiago (251) Santiago (50)
China – – Beijing (514) –
Ghana Accra (198) Accra (196) Accra (226) –
India
Delhi (169) Delhi (170) Bengaluru (283) Mumbai (55)
Mumbai (155) Mumbai (158) Delhi (269)
Indonesia Jakarta Metropolitan Area
(344)
Jakarta Metropolitan Area
(148)
Jakarta Metropolitan Area
(112)
–
Kenya
Kisumu (45) Kisumu (43) Kisumu (17) Kisumu (27)
Mombasa (43) Mombasa (62) Mombasa (24) Mombasa (29)
Nairobi (151) Nairobi (185) Nairobi (130) Nairobi (94)
Lebanon
Beirut (130) Beirut (100) Beirut (65) Beirut (47)
Jounieh (70) Jounieh (40) Jounieh (35) Jounieh (20)
Tripoli (60) Tripoli (25)
Mexico Mexico City (200) Mexico City (200) Mexico City (249) –
Morocco
Rabat (192) Casablanca (38) Casablanca (78)
–
Sale (2) Rabat (118) Rabat (158)
Sale (48) Sale (9)
Skhirate–Temara (1) Skhirate–Temara (1)
Ukraine Kiev (252) Kiev (150) Kiev (244) –
Source: ILO selected country surveys of taxi drivers and delivery workers (2019–20).
Appendix 4. ILO surveys, interviews and statistical analysis
3
X
Figure A4.1 Distribution of interviews across cities, selected surveys and countries
Source: ILO selected country surveys of taxi drivers and delivery workers (2019–20).
Ukraine (Kiev), app-based taxi drivers
Mexico (Mexico City), app-based delivery workers
Ghana (Accra), traditional taxi drivers
Kenya (Nairobi), traditional delivery workers
X
Table A4.2 Number of observations for country surveys of app-based taxi drivers
and delivery workers, by platform company
Taxi Delivery
Argentina – Glovo (109), Rappi (105), PedidosYa (86)
Chile Uber (89), Beat (21), Cabify (12), Other (4) Rappi (76), Cornershop (60), PedidosYa (59), Uber Eats (56)
China – Meituan (259), Ele.me (140), Flashex (67), SF Express (25), Other (23)
Ghana Uber (124), Bolt (61), Yango (13) Jumia (85), Papa’s Pizza (22), Other (119)
India Uber (195), Ola (129) Zomato (141), Swiggy (133), Uber Eats (115), Dunzo (46), Amazon (33),
Big Basket (30), Flipkart (30), Grofers (14), Other (10)
Indonesia Grab (197), Gojek (146), Lainnya (1) Gojek (68), Grab (44)
Kenya Uber (98), Bolt (88), Safe Boda (23),
Little (22), Other (8)
Jumia Foods (54), Glovo (33), Uber Eats (23), Sendy (18), Bolt (9), Other (34)
Lebanon Uber (167), Careem (33) Toters (96), Fastpax (4)
Mexico Uber (91), Didi (84), Other (25) Rappi (132), Uber Eats (95), SinDelantal (10), Other (12)
Morocco Careem (188), Blinc (5), VTCG0 (1) Glovo (172), Jumia (74)
Ukraine Uber (124), Uklon (69), Bolt (48), Other (11) Glovo (189), Uber Eats (22), Nova Poshta (22), Other (11)
Source: ILO selected country surveys of taxi drivers and delivery workers (2019–20).
The role of digital labour platforms in transforming the world of work
4
The primary approach used to identify workers was to locate them on the street. To locate taxi drivers,
enumerators would target places such as gas stations, clusters of oce complexes, shopping malls, air-
ports, railway stations, platform company support oces and taxi stands, among others. While traditional
taxis could easily be identied through the physical appearance of the vehicle, identifying app-based
drivers proved quite dicult in some countries. For example, in Chile app-based workers operate in a
legal grey area; they try to reduce their visibility. In some countries many respondents, in particular those
whose work was mediated by a platform company, were worried that the survey was being conducted by
the platform company. Their concern was usually mitigated by showing them an ocial letter conrming
that this study was being conducted on behalf of the ILO.
Delivery workers were located mainly near restaurants, shopping malls, or waiting points where they
would gather. They could often be identied by their branded vehicles, transport boxes or uniforms
(jackets, helmets). In addition, in countries where it proved dicult to reach the target sample size,
snowball sampling was also used.
3
In case workers were willing to respond to the survey but were unable to do so when they were rst
approached (for example, because they just received an order or could not/did not want to log out of
their app), enumerators would set up appointments at a time and place that was convenient for the
respondent. Similarly, if an interview was interrupted, later appointments were set up to complete the
interview. Respondents were compensated for the time they spent doing the interview with a xed,
country-specic amount paid at the end of the completed interview.
4
On average, it took workers around
40 minutes to complete the survey.
The target sample sizes were 200–250 app-based taxi drivers, 200 traditional taxi drivers, 200–250 app-
based delivery workers, and 50–150 traditional delivery workers. Surveys of traditional delivery workers
were undertaken only in cities where the background study showed that delivery work had been prevalent
prior to platform companies entering the market. As in some cases it proved very dicult to reach the
target sample size for each category, the sample size or distribution had to be revised in some countries
while the survey was being conducted. The nal sample size as presented in table A4.2 reects these
diculties and also that a few observations were removed from the dataset because they did not full
the selection criteria in terms of age or experience or because of quality concerns.
The rst surveys of workers were conducted in April 2019 and the last ones in February 2020. The period
of implementation for each country and survey is listed in table A4.3. As it was impossible to coordinate
the surveys in all countries at the same time, the starting and end dates varied. In addition, the duration
largely depended on the number of enumerators in each country and/or city and the ease or diculty in
locating workers for each category.
3 In Argentina, the sampling design included a rst stage that consisted of identifying potential interviewees through social net-
work groups (Facebook, WhatsApp and so on). From this rst sample, workers were asked to designate other peers who might be
interested in being part of the study, limiting the number of new participants each participant could provide.
4 The amounts were set in consideration of the country’s minimum wage and also based on the peak earnings of taxi drivers or
delivery workers.
Appendix 4. ILO surveys, interviews and statistical analysis
5
4A.2 Rapid-assessment survey on the impact of COVID-19
on workers in the taxi and delivery sectors (2020)
To assess the impact of the COVID-19 pandemic on workers in the delivery and taxi sectors, a rapid-as-
sessment survey was conducted in four countries: Chile, India, Kenya and Mexico. These countries were
chosen in an eort to understand the implications of COVID-19 on workers across the dierent regions.
The questionnaire was developed by the ILO and implemented with the help of the same consultants
who had been responsible for the 2019 survey in their respective countries. Before the nal survey was
implemented, pilot tests were conducted to rene the questionnaire and the enumerators were trained
in understanding the content and relevance of each question. The interviews were conducted using
computer-assisted telephonic interviewing (CATI) in August 2020.
X
Table A4.3 Data collection periods for country surveys of taxi drivers and delivery workers
Taxi Delivery
App-based Traditional App-based Traditional
Argentina – – 1–31 July 2019 –
Chile
11 August–
8 September 2019
5 July–
16 August 2019
10 June–
8 August 2019
18 August–
1 October 2019
China – – 6–24 July 2019 –
Ghana
6 October–
12 December 2019
19 October –
12 December 2019
4 October–
12 December 2019
–
India
22 May–
6 August 2019
21 June–
22 July 2019
9 August–
3 December 2019
28 January–
8 February 2020
9–30 August 2019
Indonesia 9–30 August 2019 10–31 August 2019
9 August–
4 September 2019
–
Kenya
30 October–
5 December 2019
31 October–
5 December 2019
31 October–
4 December 2019
Lebanon
19 September–
12 October 2019
18 September–
7 October 2019
20 September–
11 November 2019
26 September–
10 October 2019
Mexico
28 August–
21 November 2019
26 August–
2 November 2019
12 April–
12 August 2019
–
Morocco
14 December 2019–
9 January 2020
13 December 2019–
5 January 2020
14 December 2019–
15 January 2020
–
Ukraine
23 October–
3 December 2019
25 October–
15 December 2019
23 October–
1 December 2019
–
Source: ILO selected country surveys of taxi drivers and delivery workers (2019–20).
The role of digital labour platforms in transforming the world of work
6
The target sample consisted of delivery workers and taxi drivers who had participated in the 2019 survey,
and who were either still working in their respective sector or not currently working but planning to return
to this work once the situation in their city had improved. The sample size was 10 per cent of the 2019
sample of each category (app-based taxi, traditional taxi, app-based delivery, traditional delivery), while
aiming at a similar distribution in terms of gender, platform or company they work for, and migrant status
(Chile). In Kenya, it was decided to redistribute the sample from Kisumu to Nairobi and Mombasa, as the
Kisumu area was less adversely aected by restrictions put in place due to COVID-19. The nal sample
distribution is displayed in table A4.4.
X
Table A4.4 Number of observations for the rapid-assessment surveys on the impact
of COVID-19 on taxi drivers and delivery workers, by interview city
Taxi Delivery
App-based Traditional App-based Traditional
Chile Santiago (16) Santiago (16) Santiago (26) Santiago (5)
India Delhi (19) Delhi (18) Bengaluru (27) Mumbai (6)
Mumbai (16) Mumbai (16) Delhi (29) –
Kenya Mombasa (8) Mombasa (9) Mombasa (2) Mombasa (6)
Nairobi (19) Nairobi (20) Nairobi (16) Nairobi (9)
Mexico Mexico City (20) Mexico City (20) Mexico City (25) –
Source: ILO rapid-assessment surveys of taxi drivers and delivery workers (2020).
The questionnaire contained both qualitative and quantitative questions related to work arrangements,
household composition, social protection, work and income security, occupational safety and health,
collective action, and stigma and discrimination. On average, it took workers around 30 to 40 minutes
to complete the survey.
Workers who at the start of the questionnaire indicated that they had permanently stopped working as a
taxi driver or delivery worker were asked questions related to why and when they had left these sectors
and their current employment status. For these workers, it took around 10 to 15 minutes to complete the
survey. All workers who completed the survey were compensated for their time spent doing the interview.
5
To reach the target samples, in total 996 respondents were contacted using the phone number they had
provided in the 2019 survey. Many calls were unsuccessful because the number was no longer in use or
had changed owner, or individuals were not reachable because phones were switched o or because
they did not pick up. Finally, despite getting through to the respondents, some also said they were too
busy or otherwise refused to participate (see table A4.5).
5 The amounts paid to each worker who continued to work or planned to return to work in the taxi or delivery sector were twice
the amounts of the original survey conducted in 2019–20. The amounts paid to those who had stopped working in the taxi or
delivery sector were the same as in the 2019–20 survey.
Appendix 4. ILO surveys, interviews and statistical analysis
7
X
Table A4.5 Number of participants contacted for the rapid-assessment surveys on the impact
of COVID-19 on taxi drivers and delivery workers, by country
Taxi Delivery
Total
App-based Traditional App-based Traditional
Chile
Successful
Currently working 9 11 25 4 49
Not currently working but planning to return 7 5 1 1 14
Permanently quit 4 0 11 1 16
Subtotal 20 16 37 6 79
Unsuccessful
Number no longer operational or changed owner 6 0 25 1 32
Did not pick up/phone o 1 4 8 0 13
Not willing to participate 3 8 3 0 14
Subtotal 10 12 36 1 59
Total attempted contacts 30 28 73 7 138
India
Successful
Currently working 21 18 46 0 85
Not currently working but planning to return 14 16 10 6 46
Permanently quit 2 1 10 0 13
Subtotal 37 35 66 6 144
Unsuccessful
Number no longer operational or changed owner 5 2 9 3 19
Did not pick up/phone o 6 3 18 2 29
Not willing to participate 2 0 3 0 5
Subtotal 13 5 30 5 53
Total attempted contacts 50 40 96 11 197
Kenya
Successful
Currently working 23 16 13 13 65
Not currently working but planning to return 4 13 5 2 24
Permanently quit 8 10 8 1 27
Subtotal 35 39 26 16 116
Unsuccessful
Number no longer operational or changed owner 18 8 10 1 37
Did not pick up/phone o 57 79 38 20 194
Not willing to participate 8 13 5 3 29
Subtotal 83 100 53 24 260
Total attempted contacts 118 139 79 40 376
Mexico
Successful
Currently working 19 17 24 – 61
Not currently working but planning to return 1 3 1 – 5
Permanently quit 0 0 0 – 0
Subtotal 20 20 25 – 65
Unsuccessful
Number no longer operational or changed owner 56 39 33 – 128
Did not pick up/phone o 9 27 19 – 55
Not willing to participate 1 4 2 – 7
Subtotal 66 70 54 – 190
Total attempted contacts 86 90 80 – 255
Source: ILO rapid-assessment surveys of taxi drivers and delivery workers (2020).
The role of digital labour platforms in transforming the world of work
8
Background information.
6
The restrictions and economic situation impacting workers in the delivery
and taxi sectors during the COVID-19 pandemic diered among countries, cities and time periods. The
events and measures aecting these two sectors until the time of the survey (August 2020) are briey
mentioned below.
In Santiago de Chile, there had been a curfew (10 p.m. to 5 a.m.) from 22 March, and a dynamic lockdown
(that is, only certain areas of the city) from 25 March, while on 15 May a total lockdown in all municipalities
of the metropolitan area of Santiago de Chile was issued. The total lockdown was progressively relaxed
from 15 August onwards for dierent municipalities of the metropolitan area of Santiago, and the curfew
was shortened on 21 August (11 p.m. to 5 a.m.). While delivery workers (both app-based and traditional)
and traditional taxi drivers were classied as essential workers and granted a permit to move within
lockdown territories while performing their work, app-based taxi drivers were not classied as essential
workers and thereby were not eligible for such a permit. Only taxi platforms that also had delivery options
or options to request a traditional taxi were allowed to operate.
Apart from the pandemic, the economy of Santiago was also aected by a large civil protest that started
on 18 October 2019. The survey also tried to capture the implications of the protest on the respondents’
work and income security.
In India, there was a nationwide lockdown eective from 24 March until 3 May. In Bengaluru, partial
opening started on 4 May, while in Delhi and Mumbai some services started operating again on 18 May.
App-based food delivery services were operating throughout the lockdown, and parcel delivery started
again on 18 May. Taxi platforms were shut down during the lockdown. In Mumbai, traditional taxis were
allowed to operate again on 2 June, and app-based taxi platforms on 5 June. In Bengaluru and Delhi, taxi
services (both traditional and app-based) started operating on 18 May, with a restriction on the number
of passengers (1 for autorickshaws, 2 for cabs). After a surge in cases in Bengaluru, there was another
lockdown from 14 July to 22 July, where taxi services were allowed to operate only in emergencies or
towards the airport or railway stations.
In addition, with the passing of the Citizen’s Bill in early December 2019, there were a number of protests
in many cities in India, including in New Delhi, Mumbai and Bengaluru, from December 2019 to March
2020. This aected the work and earnings of the taxi drivers and delivery workers, and the survey also
tried to capture the impact thereof.
In Kenya, on 23 March all hotels and restaurants were closed for business, with only take-aways still
operating until 4 p.m. From 27 March there was a nation-wide curfew (7 p.m. to 5 a.m.). From 6 April,
movement in and out of Nairobi and Mombasa, among other cities, was restricted. Restaurants were
allowed to open (from 5 a.m. to 4 p.m.) on 27 April. On 6 June the curfew was relaxed (9 p.m. to 4 a.m.), but
the restrictions of movement in and out of Nairobi and Mombasa continued. Taxi and delivery services
(both app-based and traditional) were allowed to operate during the entire period, respecting the curfew.
In Mexico City, restrictions on economic activities started on 26 March. On 21 April, the situation was
declared to have reached the status of a pandemic in the country. Starting from 1 June, a trac-light
colour system was introduced, denoting the situation in each state and updated weekly. Mexico City
was classied as “red” for June, July and August, implying that only essential activities would operate.
Nevertheless, life on the streets slowly started to pick up again during these months. The taxi and delivery
sectors (both app-based and traditional) were both considered as essential during the entire period, and
therefore allowed to operate.
6 Based on information provided by consultants in each country.
Appendix 4. ILO surveys, interviews and statistical analysis
9
4A.3 Surveys of workers on online web-based platforms
To better understand working conditions on online web-based platforms, several surveys were under-
taken by the ILO: a global survey of crowdworkers (2017); a global survey of workers on freelance and
competitive programming platforms (2019–20); and country-level surveys of platform workers in China
and Ukraine (2019). All these surveys contained a section on worker demographics, work experience
and work history, as well as detailed information on types of tasks completed and working conditions
such as hours worked, income, benets received, and nancial and social security. The questionnaires
included both quantitative questions as well as some open-ended questions requiring textual answers,
which provided ndings of a more qualitative nature. The survey language was English for the global
surveys and Chinese and Ukrainian respectively for the surveys of platform workers in those countries.
As there is no database of online web-based platform workers, it was not possible to draw a random
sample. Depending on the target group, dierent sampling methodologies were chosen. Independent
of the sampling methodology, workers self-selected to participate in the surveys. The dierent surveys
and the sampling methodologies used are described in more detail below.
4A.3.1 Global survey of crowdworkers (2017)
The survey was conducted between February and May 2017 on ve major microtask platforms operating
across the globe: AMT, CrowdFlower (now Appen), Clickworker, Microworkers and Prolic (formerly Prolic
Academic). It was a follow-up to and extension of a survey undertaken on AMT and CrowdFlower in 2015
(see Berg 2016). The 2015 questionnaire was modied by the ILO with assistance from SoundRocket, a
survey research company specialized in the social sciences. The survey was listed as a paid task on the
ve platforms. There were no restrictions as to who could participate except in the case of AMT, where
workers from India and the United States were targeted. The survey was posted in small batches at
dierent times of the day and the workers self-selected to participate in the survey. This is common
practice among empirical studies of crowdwork and is considered to be the best way of reaching out to
a wide range of workers engaged on the platforms. On average, it took respondents about 30 minutes
to complete the survey. Out of a total of 3,345 respondents who participated in the survey, almost 30
per cent had to be excluded from the analysis because they only partially completed the survey, or did
not pay sucient attention, or used algorithms to complete it, or used multiple accounts or platforms to
complete it (see Berg et al. 2018). This resulted in a nal sample of 2,350 workers from 75 countries (see
tables A4.6 and A4.7).
In addition to the survey, semi-structured interviews were conducted on Skype with 21 workers in
August 2017 in order to have a better understanding of their motivations, the tasks they performed, their
(dis)satisfaction with microtask work and how it aected their personal and professional life.
4A.3.2 Global survey of freelance and contest-based workers,
and competitive programmers, 2019–20
The questionnaires for the global surveys of freelancers and contest-based workers, and competitive
programmers were developed by the ILO with assistance from SoundRocket, the survey research com-
pany that assisted the ILO in conducting the global survey of crowdworkers in 2017. The target population
included freelance, free competition and competitive programmers engaged in work or training activities
on any of 12 predetermined digital platforms (99designs, CodeChef, Codeforces, Designhill, Freelancer,
HackerEarth, HackerRank, Hatchwise, Iceberg, PeoplePerHour, Topcoder, Upwork). These platforms had
been selected because they are some of the major ones in their respective elds and it seemed possible
to verify that workers were undertaking tasks on them.
The role of digital labour platforms in transforming the world of work
10
Dierent models of recruitment were assessed for their feasibility and the nal strategy included:
X
Recruitment directly on the platform: This method consisted in listing the survey as a paid task on the
platform. Around 90 per cent of the freelancers and 8 per cent of the competitive programmers in
the sample were recruited in this way. Recruitment postings were made on Upwork, Freelancer and
PeoplePerHour, across a range of job categories and task types, in an attempt to recruit a diverse range
of survey participants. This method worked well on Upwork, whereas on Freelancer the quality of pro-
posals was very low, with many workers submitting bids unrelated to the task, and on PeoplePerHour
the job postings were agged and removed by the moderators shortly after being published. During
the recruitment on Upwork, it was ensured that there was variety in terms of tasks, geographic location
and worker experiences. While some respondents recruited through Upwork had earned less than
US$100 on the platform (4 per cent), the majority had earned over US$1,000 (78 per cent), with some
exceeding US$10,000 (35 per cent) or even US$100,000 (2 per cent).
X
Identication of workers through other digital platforms (AMT): AMT was used to recruit previous partici-
pants from the 2017 survey of crowdworkers who were identied as potentially eligible for this survey
and had indicated that they would be willing to participate in future surveys. The survey was set up as
two tasks. The rst asked basic demographic questions and identied whether the worker participated
in freelance, contest-based or competitive programming platforms. The second task was oered to
those deemed eligible and consisted of the detailed questionnaire for freelance workers or competitive
programmers, respectively. About 60 and 29 respondents on AMT successfully completed the freelance
or competitive questionnaire, respectively. However, a detailed analysis of answers provided by re-
spondents, including to open-ended textual questions, made clear that a substantive number of these
entries were of low quality. Some respondents probably had no experience of working on freelance or
competitive programming platforms, but indicated that they did in order to be able to complete the
survey and receive the nancial reward – sometimes multiple times with dierent accounts, as was
obvious from striking similarities in textual answers. After exclusion of these low-quality observations,
about 8 per cent of the freelancers (36 respondents) and 3 per cent of competitive programmers
(2 respondents) in the nal sample were recruited through AMT.
X
Targeting of workers through online advertisements: Advertisements were developed to target the three
worker populations, and were posted on Facebook. These were relatively successful during the pilot
study. However, a change in Facebook’s advertising policy resulted in a drastic reduction in the number
of micro-targeting options, which signicantly reduced their eectiveness for participant recruitment
during the nal study. Of over 50,000 individuals who clicked on these advertisements, only about
250 entered the survey. While 14 respondents completed the survey successfully, only six were con-
sidered in the nal sample (1 per cent of freelancers) due to quality concerns for the remaining eight
observations.
X
Coordination with online content creators to share the survey with their audience: A short segment adver-
tising the study was included in two videos that were posted to the YouTube channel of a YouTuber
with whom there was a collaboration for this study. This YouTuber’s channel focused on competitive
programming and had over 240,000 subscribers. He also shared information about the study in one
Facebook and two Twitter posts. About 74 per cent of the competitive programmers were recruited
through this method.
X
Recruitment through online forums: A handful of competitive programmers were recruited from the
CodeChef community forum. In addition, potential respondents were also contacted through other
online forums or social media platforms (such as Quora, Meetup, LinkedIn), but without success. About
15 per cent of the competitive programmers were recruited through online forums.
Appendix 4. ILO surveys, interviews and statistical analysis
11
Another method of recruitment that had been assessed for its feasibility was snowball sampling. This
method was unsuccessful during the pilot study, as none of the initial volunteers followed up to share
the survey with others, and it was therefore deemed unfeasible.
The responses were collected between the end of August 2019 and the end of January 2020. All partici-
pants who successfully completed the questionnaire were oered compensation in appreciation of their
participation. The average completion time was about 60 minutes for the freelance survey and about 25
minutes for the survey of competitive programmers.
In total, 609 respondents completed the survey of freelance and contest-based workers and 190 respond-
ents completed the survey of competitive programmers. After data cleaning and elimination of records
that were of low quality or duplicate entries, the nal sample of the survey of freelance and contest-based
workers included 449 respondents from 80 dierent countries and three dierent platforms; and that of
the competitive programmers included 62 respondents from seven dierent countries and ve dierent
platforms (see tables A4.6 and A4.7). Due to the diculties described above of recruiting freelancers on
various platforms, 93 per cent of the respondents were performing tasks on Upwork.
X
Table A4.6 Number of observations for global surveys of online web-based platforms,
by income group and country/territory
Freelance Competitive programming Microtask
High-income Austria (2)
Belgium (1)
Canada (8)
Croatia (3)
Czechia (3)
Denmark (1)
Estonia (1)
Finland (1)
France (8)
Germany (5)
Greece (7)
Hungary (2)
Ireland (1)
Israel (2)
Italy (7)
Netherlands (3)
New Zealand (1)
Oman (1)
Poland (4)
Portugal (4)
Republic of Korea (1)
Romania (5)
Slovenia (1)
Spain (4)
Saint Kitts And Nevis (1)
Sweden (1)
Switzerland (1)
Taiwan, China (1)
United Arab Emirates (1)
United Kingdom (4)
United States (63)
Belgium (1)
Norway (1)
United States (3)
Australia (4)
Austria (9)
Belgium (4)
Brunei (1)
Canada (41)
Chile (3)
Croatia (11)
Czechia (2)
Estonia (1)
Finland (2)
France (23)
Germany (188)
Greece (6)
Hungary (6)
Ireland (4)
Israel (1)
Italy (67)
Japan (1)
Latvia (1)
Lithuania (2)
Netherlands (10)
New Zealand (3)
Poland (13)
Portugal (30)
Romania (18)
Saudi Arabia (2)
Singapore (3)
Slovakia (2)
Slovenia (1)
Spain (43)
Sweden (1)
Switzerland (4)
United Kingdom (294)
United States (697)
Uruguay (1)
The role of digital labour platforms in transforming the world of work
12
X
Table A4.6 (cont’d.)
Freelance Competitive
programming
Microtask
Upper-middle-
income
Albania (2)
Argentina (3)
Armenia (2)
Belarus (2)
Bosnia and Herzegovina (3)
Brazil (11)
Bulgaria (3)
China (5)
Colombia (5)
Dominican Republic (2)
Georgia (1)
Indonesia (3)
Jamaica (1)
Jordan (1)
Lebanon (1)
Malaysia (4)
Mexico (5)
North Macedonia (4)
Peru (1)
Russian Federation (3)
Serbia (10)
South Africa (5)
Thailand (2)
Turkey (3)
Venezuela, Bolivarian Republic of (10)
Peru (1) Albania (1)
Argentina (4)
Armenia (1)
Bosnia and Herzegovina (39)
Brazil (45)
Bulgaria (10)
China (1)
Colombia (3)
Ecuador (2)
Georgia (1)
Indonesia (28)
Jamaica (3)
Malaysia (8)
Mexico (9)
North Macedonia (10)
Peru (5)
Russian Federation (28)
Serbia (75)
South Africa (7)
Turkey (11)
Venezuela, Bolivarian Republic of (71)
Lower-middle-
income
Algeria (2)
Bangladesh (16)
Cambodia (1)
Egypt (8)
El Salvador (2)
India (41)
Kenya (20)
Myanmar (1)
Nepal (1)
Nicaragua (3)
Nigeria (8)
Occupied Palestinian Territory (9)
Pakistan (34)
Philippines (43)
Tunisia (1)
Ukraine (9)
Uzbekistan (1)
Viet Nam (2)
Bangladesh (2)
India (53)
Tunisia (1)
Algeria (6)
Bangladesh (10)
Bolivia, Plurinational State of (1)
Egypt (4)
Ghana (1)
India (343)
Kenya (7)
Kyrgyzstan (1)
Morocco (7)
Nepal (32)
Nigeria (22)
Pakistan (11)
Philippines (10)
Republic of Moldova (3)
Sri Lanka (10)
Tunisia (4)
Ukraine (14)
Viet Nam (2)
Low-income Benin (1)
Burkina Faso (1)
Ethiopia (1)
Malawi (1)
Rwanda (1)
Sources: ILO global surveys of crowdworkers (2017) and workers on freelance
and competitive programming platforms (2019–20).
Appendix 4. ILO surveys, interviews and statistical analysis
13
4A.3.3 Survey of platform workers in Ukraine (2019)
The survey of platform workers in Ukraine was conducted by the Kiev International Institute of Sociology
(KIIS) on behalf of the ILO. It took place in November 2019 and was a follow-up to a survey conducted
in November and December 2017 (see Aleksynska, Bastrakova and Kharchenko 2018). The 2017 survey
questionnaire was revised and some questions were added.
The survey targeted respondents aged 18 years and older, who reside continuously in Ukraine and who
identied themselves as workers performing work through at least one digital platform for pay in the
12 months preceding the survey. The dierent methods of participant recruitment used included:
X
spreading information about the survey on Kabanchik.ua, the leading online platform providing work
in Ukraine (74 per cent of respondents);
X
selecting participants from InPoll, an online panel that provides access to active Ukrainian internet
users (20 per cent of respondents);
X
sending invitations to the participants of the 2017 survey of online platforms workers in Ukraine who
had provided their contact details (4 per cent of respondents);
X
posting information about the survey in thematic Facebook groups (1 per cent of respondents); and
X
snowball sampling to recruit individuals otherwise difficult to reach or engage (1 per cent of
respondents).
On average, it took around 30–40 minutes to complete the survey and participants were oered a small
remuneration on completion. In total, 1,112 respondents completed the survey, out of whom 54 were ex-
cluded due to quality concerns related to their textual answers, resulting in 1,058 remaining observations.
The survey included workers who primarily found and completed tasks online, and workers who used the
platforms to nd tasks that they would complete oine, such as repair, cleaning and delivery services,
among others. The workers who were primarily completing oine tasks were excluded from the analysis
in this report, so as to allow better comparability between the dierent surveys (global survey of workers
on freelance and competitive programming platforms, and survey of platform workers in China). The
nal sample used for analysis in this report consists of 761 respondents, of whom 62 per cent named
Kabanchick.ua as the main platform they worked for, while the remainder primarily worked on other
Ukraine-based (10 per cent), Russia-based (10 per cent) or English-language platforms (18 per cent) (see
table A4.7).
X
Table A4.7. Number of observations per platform for online web-based surveys
Freelance and
contest-based
(449)
Competitive
programming
(62)
Microtask
(2 350)
Ukraine
(761)
China
(1 107)
99designs (4)
Freelancer (27)
Upwork (418)
CodeChef (13)
Codeforces (14)
HackerRank (33)
Iceberg (1)
Topcoder (1)
Amazon Mechanical Turk (489)
Clickworker (455)
CrowdFlower (now Appen; 355)
Microworkers (556)
Prolic (495)
Advego.ru (32)
Amazon Mechanical Turk (5)
.ru (13)
Free-lance.ua (7)
Freelance.ru (6)
Freelance.ua (46)
Freelancehunt.com (40)
Freelancer.com (27)
Kabanchik.ua (471)
Upwork.com (41)
Weblancer.net (7)
Others (66)
680 (293)
EPWK (232)
k68 (48)
ZBJ (534)
Sources: ILO global surveys of crowdworkers (2017) and workers on freelance and competitive programming platforms
(2019–20); ILO surveys of platform workers in China (2019) and Ukraine (2019).
The role of digital labour platforms in transforming the world of work
14
4A.3.4 Survey of platform workers in China, 2019
The survey of platform workers in China was conducted by Professor Ruixin from Harbin Institute of
Technology on behalf of the ILO. The survey questionnaire used was based on earlier survey question-
naires for crowdworkers and for platform workers in Ukraine (Berg 2016; Berg et al. 2018; Aleksynska,
Bastrakova and Kharchenko 2018), adapted to the Chinese context. It included about 85 questions (ex-
cluding follow-up questions) and took around 30 minutes to complete.
The survey was oered on China’s top four online platforms, as determined by their Alexa ranking and
Baidu weight, namely ZBJ, EPWK, 680 and k68. These platforms oer a wide range of tasks that appeal to
workers with dierent backgrounds (Chen, forthcoming). The survey was listed as a paid task, and upon
completion workers were remunerated for performing the task. Any worker aged 18 years or older who
had been doing online work for at least three months was eligible for the survey. After responses were
collected and the data was cleaned, there were a total of 1,107 respondents (see table A4.7).
4A.3.5 Regional groupings used in Chapter 4
As noted in Chapter 4, countries are grouped into “developed” and “developing” countries for the purpose
of analysis. The grouping is based on the World Bank’s 2020–21 country classication by income level,
which is based on the country’s 2019 GNI per capita in current US dollars. Countries that are classied
as high-income are considered to be “developed”, whereas all others (upper-middle-, lower-middle- and
low-income) are considered as “developing”. The number of observations for each group is displayed in
table A4.8.
Due to large dierences in sample size, the country-specic surveys of platform workers in China and
Ukraine are excluded when presenting results for developed and developing countries separately, as
otherwise the results for developing countries would be primarily driven by China and Ukraine, instead
of providing a global picture.
X
Table A4.8 Number of observations per platform for online web-based surveys
Freelance
Competitive
programming
Microtask Total
Developed countries 148 5 1 499 1 652
Developing countries 301 57 850 1 208
Sources: ILO global surveys of crowdworkers (2017) and workers on freelance
and competitive programming platforms (2019–20).
4A.4 Interviews with freelancers
Interviews were conducted with 23 freelancers in Africa, Asia, the Arab States, and Latin America and
the Caribbean between April and September 2019. They were identied using LinkedIn and other social
media platforms, where their work history was reviewed based on their proles, and selected workers
who indicated use of online platforms were invited to interviews through Skype. In some countries, after a
freelancer was identied the snowball method was used to reach out to other workers. All the freelancers
contacted expressed enthusiasm about participating in the research project. The interview took around
45 to 90 minutes, and they were compensated for their time.
Appendix 4. ILO surveys, interviews and statistical analysis
15
4B. Statistical analysis
4B.1 Earnings of taxi and delivery workers
4B.1.1 Comparing hourly earnings of workers in the app-based
and traditional taxi and delivery sectors
To compare the hourly earnings of app-based and traditional workers in the taxi and delivery sectors,
the ordinary least squares (OLS) regression method was used. The dependent variable is the workers’ log
hourly earnings in US dollars and the regressor is a binary variable that takes the value of 1 for app-based
workers and 0 for traditional workers. Several covariates were introduced to the regression, including age,
sex, education, marital status, household size, migration status, experience, ethnic group and interview
city (where applicable), and dummy variables of having another job and renting the vehicle (see tables
A4.9 and A4.10).
The regression results show that app-based taxi drivers earn more per hour on average than traditional
taxi drivers with similar characteristics. The dierence varies from 22 per cent (Ukraine) to 86 per cent
(Ghana) and is signicant at 95 per cent in Morocco and at 99 per cent in all other countries under analysis.
In the delivery sector, app-based workers earn more in Kenya (39 per cent) and Lebanon (25 per cent),
but less in Chile (24 per cent) compared to traditional delivery workers, and the dierence in earnings
is highly signicant in all three countries. India was excluded from the regression analysis for delivery
workers, as the survey of traditional delivery workers was conducted among dabbawalas (traditional
lunchbox delivery) in Mumbai, while the app-based survey was conducted in Bengaluru and Delhi, which
limits the comparability of income gures.
The role of digital labour platforms in transforming the world of work
16
X
Table A4.9 Regression results: App-based and traditional taxi drivers
(percentage changes; dependent variable: log of hourly earnings in US$)
Chile Ghana India Indonesia Kenya Lebanon Mexico Morocco Ukraine
App-based
(traditional)
72.6*** 86.1*** 78.5*** 48.1*** 34.2*** 78.1*** 72.1*** 25.8** 22.2***
Age 1.2 0.9 0.1 1.9 2.0 –0.3 0.5 1.6 2.1
Age-squared –0.0 –0.0 –0.0 –0.0 –0.0 –0.0 –0.0 –0.0 –0.0*
Education
1
Higher secondary –1.1 3.4 6.3 23.3*** 32.0*** 2.3 6.1 1.0
University degree 2.8 5.6 26.8*** 58.4*** 40.0*** –6.9 7.3 0.0 5.3
Married
(not married)
10.2 4.3 11.4* 1.2 –7. 3 0.9 5.8 –3.4 12.1*
Household size –3.6* –1.3 0.6 –3.7** 1.6 –0.9 –1.4 –3.2* –3.3
Years of experience 0.3 –1.0* 0.5* –2.1*** – 0.1 –0.0 –0.3 –0.2 0.7
Has another job 10.3 16.8** 27.3* * –18.2** 18.4* –5.5 3.7 –1.2 3.0
Rented vehicle
(own vehicle)
–15.8** 7.5 –13.2*** 31.5*** – 4.1 –4.4 –4.2 –29.7*** –31.9***
Ethnic group/nationality
2
Ewe –3.7
Ga-Adangbe –9.6
Other –5.0 –8.5
Scheduled tribes 32.6***
Scheduled castes –4.0
Other backward
castes
–7.2* *
Betawi 12.0
Sunda 6.6
Syrian 2.6
Palestinian 44.3***
City
3
Mumbai 37.1***
Kisumu 1.6
Mombasa 20.7***
Jounieh –0.5
Tripoli –69.9***
Observations 232 373 495 437 505 371 378 359 361
R-squared 0.369 0.450 0.513 0.161 0.164 0.743 0.316 0.183 0.122
Notes: Reference categories in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1. For easier interpretation of the results, the table presents
percentage change instead of regression coecients. In order to avoid approximation errors, the percentage change is calculated using
the formula 100 × [exp(coecient) – 1]. Regression results are available from the authors upon request.
1
Reference categories: Ukraine: higher secondary or below; all other countries: secondary or below.
2
Reference categories: Ghana: Akan; India: Forward castes; Indonesia: Jawa; Lebanon: Lebanese.
3
Reference categories: India: Delhi; Kenya: Nairobi; Lebanon: Beirut.
Source: ILO calculations based on ILO selected country surveys of taxi drivers (2019–20).
Appendix 4. ILO surveys, interviews and statistical analysis
17
4B.1.2 Relationship between dierent factors and hourly earnings
in the app-based delivery sector
To further investigate the relationship between dierent factors and hourly earnings of app-based
delivery workers, the OLS regression method was used. The dependent variable is workers’ log hourly
earnings in US dollars. Various covariates were introduced in order to capture factors that simultaneously
aect workers’ hourly earnings, including demographic and work-related variables (see tables A4.11 and
A4.12). Female and migration dummies were added for countries where the sample included at least
10 per cent of female or migrant workers, and the worker’s rating on the application was added where it
was available for at least 90 per cent of respondents.
The results show that the factors with a signicant correlation with earnings vary depending on the
country.
X
Table A4.10 Regression results: App-based and traditional delivery workers
(percentage changes; dependent variable: log of hourly earnings in US$)
Chile Kenya Lebanon
App-based (traditional) –23.9*** 39.0*** 25.0***
Age –0.9 4.6 1.6
Age-squared 0.0 – 0.1 –0.0
Female (male) –0.9
Education (below secondary)
Higher secondary 13.3 10.0 7.6
University degree 15.7 4.0 13.2
Married (not married) 0.1 0.6 3.2
Household size 0.9 2.3 0.4
Years of experience 3.8*** 2.1* 1.4**
Has another job 3.9 –6.1 –2.3
Nationality (Lebanese)
Syrian –19.5***
Palestinian –21.5***
City
1
Kisumu –14.3*
Mombasa 5.6
Jounieh –2.2
Tripoli –40.1***
Observations 287 307 181
R-squared 0.191 0.185 0.543
Notes: Reference categories in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1. For easier
interpretation of the results, the table presents percentage change instead of regression
coecients. In order to avoid approximation errors, the percentage change is calculated using
the formula 1000 × [exp(coecient) – 1]. Regression results are available from the authors
upon request.
1
Reference categories: Kenya: Nairobi; Lebanon: Beirut.
Source: ILO calculation based on ILO selected country surveys of delivery workers (2019–20).
The role of digital labour platforms in transforming the world of work
18
A signicant gender pay gap can be observed in Argentina and Chile, where women are expected to earn
around 14 per cent less than their male counterparts. There is no signicant gender gap in Ukraine, while
the remaining countries did not have sucient observations to assess whether a gender pay gap exists.
In most countries, hourly earnings are not correlated with education levels. However, workers with higher
secondary education earn more in India and Kenya, and those with a university degree earn more on
average in Chile and Lebanon than their counterparts with lower education levels (secondary education
or below).
In Argentina, Chile and Lebanon, a substantial proportion of the respondents are migrants. In Chile and
Lebanon, migrant app-based delivery workers tend to earn less than their non-migrant counterparts
(15 and 13 per cent respectively), while there is no signicant dierence in Argentina.
In some countries there are signicant dierences in earnings depending on the platform company. In
Argentina, workers on Glovo and PedidosYa earn about 25 per cent more than those primarily working
on Rappi. Similarly, in Chile, earnings are higher for those on PedidosYa (42 per cent) or Uber Eats (18 per
cent) than for those on Rappi. In Ghana, workers on Jumia (48 per cent) or other platforms (35 per cent)
earn less than those with similar characteristics on Papa’s Pizza. In India, those on Dunzo earn more
(11 per cent), while those on Flipkart earn less (12 per cent) than on Uber Eats. In Indonesia and Kenya,
there is no signicant dierence in earnings for workers on dierent platforms. In Mexico, workers on
Uber Eats earn about 15 per cent less than those on Rappi. In Morocco, those working on Jumia earn
16 per cent less than those on Glovo. In Ukraine, there is no signicant dierence between the hourly
earnings of Uber Eats and Glovo couriers, while those working on other platforms tend to earn 26 per
cent less than workers on Uber Eats.
Earnings are also associated with ratings. While in many countries only some respondents could provide
information on their rating, this was available for almost all respondents in Indonesia, Kenya, Lebanon and
Mexico. Higher ratings are not associated with higher hourly earnings in Lebanon and Mexico. However,
in Indonesia and Kenya workers with 1 per cent higher ratings are expected to earn around 1 per cent
more, which means that a one star dierence on a ve star scale is associated with a 20 per cent dierence
in hourly earnings.
The mode of transport also has an inuence on earnings. In Indonesia and Lebanon all respondents in
the app-based delivery sector were using motorbikes. In other countries, the mode of transport was
more varied and often also correlated with hourly earnings. In Argentina, Chile, Mexico and Ukraine
delivery workers using bicycles are expected to earn between 20 and 25 per cent less than those using
motorbikes. Similarly, in Mexico and Ukraine those mainly delivering on foot are expected to earn less
(36 and 15 per cent, respectively) than those on motorbikes.
X
Table A4.11 Regression results: App-based delivery workers
(percentage changes; dependent variable: log of hourly earnings in US$)
Argentina Chile Ghana India Indonesia Ke nya Lebanon Mexico Moro cco Ukraine
Age – 0.1 –1.7 8.1 1.3 0.1 3.0 –2.2 0.8 5.8 –1.9
Age-squared 0.0 0.0 – 0.1 –0.0 0.0 –0.1 0.0 –0.0 – 0.1 0.0
Female –14.9*** –13.4** –8.8
Education
1
Higher secondary –9.6 19.5 –20.8* 6.1* * 6.0 28.1** 6.0 2.9 –2.3
University degree –8.2 27.0** 12.8 6.3 23.9 21.3 25.8** 11.1 0.2 3.6
Married (not married) 4.7 –5.9 20.8 –3.3 –2.7 –1.0 0.6 –5.5 9.5 –3.1
Household size –0.8 1.5 –4.0* 1.5* –0.8 1.7 0.1 –1.2 3.4** –1.5
Migrant –5.9 –15.2*** –13.1**
Months of experience 0.3 3.8*** 6.8** 1.3 1.1 5.1** –0.4 0.3 1.4 0.5
Has another job –9.6* –4.1 23.0* 1.0 –10.9 –28.7*** 2.6 3.7 –2.3 13.1*
Rating 1.2** 0.6** – 0.1 0.3
Mode of transport (motorbike)
Bicycle –24.3*** –24.6*** –16.5 13.0* –19.8*** –7.6 –20.1***
Car 26.9*** 7.0 23.9* 0.7
Foot –36.4*** 19.2 –15.2**
Ethnic group
2
Ewe 10.4
Ga-Adangbe 3.4
Other –3.0 –16.1
Scheduled castes 1.0
Scheduled tribes 3.2
Other backward castes 0.1
Betawi –27.9* *
Sunda –18.3
City
3
Bengaluru 8.9***
Mombasa 3.2
Kisumu –8.1
Jounieh –13.1**
Casablanca/other 2.9
Main platform
4
Grab 8.7
Sendy –1.3
PedidosYa 23.5*** 41.5***
Glovo 26.1*** –0.1 0.7
Jumia –48.1*** –10.0 –15.5***
Uber Eats 18.0*** –15.3***
Other –34.8*** –17. 2 –10.8 –26.1**
Zomato –2.7
Swiggy –4.2
Amazon 3.7
Dunzo 11.1* *
Flipkart –11.5*
Observations 283 241 179 511 96 166 93 234 202 223
R-squared 0.258 0.418 0.220 0.065 0.129 0.183 0.265 0.233 0.10 4 0.145
Notes: Reference categories in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1. For easier interpretation of the results, the table presents percentage
change instead of regression coecients. In order to avoid approximation errors, the percentage change is calculated using the formula
100 × [exp(coecient) – 1]. Regression results are available from the authors upon request.
1
Reference categories: Ukraine: higher secondary or below; all other countries: secondary or below.
2
Reference categories: Ghana: Akan; India: Forward castes; Indonesia: Jawa.
3
Reference categories: India: Delhi; Kenya: Nairobi; Lebanon: Beirut; Morocco: Rabat.
4
Reference categories: Argentina, Chile and Mexico: Rappi; Ghana: Papa’s Pizza; India, Kenya and Ukraine: Uber Eats; Indonesia: Gojek; Morocco:
Glovo.
Source: ILO calculations based on ILO selected country surveys of delivery workers (2019–20).
The role of digital labour platforms in transforming the world of work
20
4B.2 Earnings of workers on online web-based platforms
4B.2.1 Comparing hourly earnings of workers on microtask platforms
and traditional workers in India and the United States
The purpose of this analysis is to compare the hourly earnings of microtask workers with those of
traditional workers with similar characteristics who undertake comparable activities in India and the
United States, which can give an idea about how much workers on microtask platforms could earn in
the traditional labour market. For this comparison, the ILO global survey of crowdworkers (2017), the
Periodic Labour Force Survey (PLFS) (2017–18) by the National Sample Survey Oce (NSSO) in India, and
the Current Population Survey (CPS) (2017) by the Bureau of Labor Statistics and the Census Bureau in
the United States were used.
To identify the most similar sectors, microtasks were matched with activities in the traditional labour
market. The matching process was based on the microtask description and the National Industrial
Classication (NIC) for India and the North American Industry Classication System (NAICS) for the United
States (see table A4.12). The most similar activities in the selected countries proved to be information
service activities (International Standard of Industry Classication (ISIC) 63) and oce administrative,
oce support and other business support activities (ISIC 82) due to the similarity in the nature of the
tasks as well as the skills they might require. Some tasks on microtask platforms such as content moder-
ation and transcription were matched with both activities, while others such as categorization or data
collection were matched with only one activity. Finally, content access and surveys and experiments were
not matched with any of the industry codes, and those microtask workers who performed exclusively
content access or surveys and experiments were excluded from this analysis.
To investigate the relationship between hourly earnings of workers in the traditional labour market and
those performing tasks on microtask platforms, OLS regression was used with covariates such as age, sex,
education, marital status, household size, living area and having another job. The dependent variable is
the individual’s log total hourly earnings (paid and unpaid earnings in the case of microtask workers). The
regressor of interest is the microtask binary variable which takes the value of 1 if the individual performs
work on microtask platforms and 0 otherwise. In each country, three models are specied: (i) all workers;
(ii) male; and (iii) female workers.
The OLS results suggest that, after controlling for basic characteristics, workers on microtask platforms
are associated with much lower hourly earnings than their counterparts in the traditional labour market.
This holds true for all three models in both countries, and the results are signicant at 99 per cent in each
case. Workers on microtask platforms are expected to earn 64 per cent less in India and 81 per cent less
in the United States than their counterparts undertaking similar activities in the traditional labour market
when all observations are included in the sample (see column 1 of tables A4.13 and A4.14). When only
male workers are included in the sample, microtask workers are expected to earn 63 per cent less in India
and 80 per cent less in the United States (see column 2 of tables A4.13 and A4.14). Among female workers,
microtask workers are expected to earn 69 per cent less in India and 83 less per cent in the United States
compared to their counterparts in the traditional labour market (see column 3 of tables A4.13 and A4.14).
Appendix 4. ILO surveys, interviews and statistical analysis
21
X
Table A4.12 Type of microtask and similar NIC and NAICS codes
Microtask Description NIC (India) NAICS (United States)
Articial
intelligence
and machine
learning
X
Collection of data and other
information to train machine-
learning algorithms
X
Tasks related to programming
and coding or to solving
mathematical or logical problems
X
63114 Providing data entry
services
X
63111 Data processing activities
including report writing
X
62011 Writing, modifying, testing
of computer program to meet the
needs of a client
X
518210 Data processing, hosting
and related services
X
541511 Custom computer
programming services
Categorization
X
Classication of entities into
groups (bookmarking, tagging,
classifying, pinning)
X
63114 Providing data entry
services
X
518210 Data processing, hosting
and related services
Content access
X
Promotion of a specic product,
app testing
X
Search engine optimization by
fake trac creation
– –
Content creation
and editing
X
Creating new content
X
Proofreading, editing or
translating existing materials
(mostly text)
X
Might be time-consuming
X
63111 Data processing activities
including report writing
X
518210 Data processing, hosting
and related services
Content
moderation
X
Reviewing content including text,
images and videos
X
Detecting if any of the material
posted on the website might
violate local laws, social norms or
the platform’s guidelines
X
63111 Data processing activities
including report writing
X
63999 Other information service
activities
X
82192 Document preparation,
typing, word processing and
desktop publishing
X
518210 Data processing, hosting
and related services
X
519190 All other information
services
X
561410 Document preparation
services
Data collection
X
Metadata collection
X
Finding, copying and pasting
information
X
Gathering information from
specic geographic locations
X
63114 Providing data entry
services
X
518210 Data processing, hosting
and related services
Market research
and reviews
X
Reviewing or rating of a product,
service or location (imaginary)
X
63114 Providing data entry
services
X
518210 Data processing, hosting
and related services
Surveys and
experiments
X
Completing surveys from
academic researchers
X
There may be some overlap with
market research
– –
Transcription
X
Transcription from dierent types
of media, such as audio, text,
photos or videos, into written
form
X
82192 Document preparation,
typing, word processing and
desktop publishing
X
63114 Providing data entry
services
X
561410 Document preparation
services
Verication
and validation
X
Verifying and “cleaning” existing
data or classications, or
conrming the validity of some
content
X
63111 Data processing activities
including report writing
X
518210 Data processing, hosting
and related services
Sources: ILO classication based on ILO global survey of crowdworkers (2017); U.S. Census Bureau, Current Population Survey (2017);
NSSO, Periodic Labor Force Survey (2017–18).
The role of digital labour platforms in transforming the world of work
22
X
Table A4.13 Regression results: Microtask and traditional workers in India
(dependent variable: log of hourly earnings in US$)
(1)
India, total
(2)
India, male
(3)
India, female
Coecient
Percentage
change
Coecient
Percentage
change
Coecient
Percentage
change
Microtask (traditional work)
–1.03*** –64.1*** –0.98*** –62.5*** –1.16*** –68.8***
(0.106) (0.118) (0.246)
Female (male)
–0.11*** –10.3***
(0.042)
Age
0.05*** 5.1*** 0.06*** 6.2*** 0.04 4.0
(0.010) (0.011) (0.024)
Age-squared
–0.00*** –0.0*** –0.00*** –0.1*** –0.00 –0.0
(0.000) (0.000) (0.000)
Education (no high school)
High school diploma
0.22** 25.2 ** 0.15* 15.6* 0.36 43.9
(0.091) (0.078) (0.242)
Technical degree
0.06 6.3 0.01 0.8 0.11 11.4
(0.130) (0.133) (0.294)
Bachelor’s degree
0.70*** 101.3*** 0.54*** 71.3*** 1.13*** 209.9***
(0.141) (0.136) (0.389)
Postgraduate degree
0.55*** 73.0*** 0.46*** 58.6*** 0.69*** 99.5***
(0.093) (0.079) (0.256)
Above postgraduate degree
0.64*** 89.5*** 0.50*** 65.1*** 0.88*** 140.3***
(0.103) (0.095) (0.271)
Married
0.09** 9.6** 0.02 2.5 0.21** 23.2**
(0.042) (0.049) (0.088)
Household size
–0.01 –1.4 –0.02* –2.0* 0.00 0.2
(0.009) (0.010) (0.022)
Urban
(rural)
0.13*** 14.4*** 0.13*** 13.9*** 0.17 18.5
(0.037) (0.040) (0.103)
Has another job
0.03 2.6 –0.02 –2.1 0.17 18.8
(0.114) (0.125) (0.276)
Constant
–0.46** –0.49* –0.71
(0.232) (0.257) (0.511)
Observations
1 822 1 822 1 4 45 1 445 377 377
R-squared
0.323 0.323 0.342 0.342 0.277 0.277
Notes: Reference categories and robust standard errors in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.
Percentage changes are calculated using the formula 100 × [exp(coecient) – 1].
Sources: ILO calculations based on ILO global survey of crowdworkers (2017); NSSO, Periodic Labour Force Survey (2017–18).
Appendix 4. ILO surveys, interviews and statistical analysis
23
X
Table A4.14 Regression results: Microtask and traditional workers in the United States
(dependent variable: log of hourly earnings in US$)
(1)
United States, total
(2)
United States, male
(3)
United States, female
Coecient
Percentage
change
Coecient
Percentage
change
Coecient
Percentage
change
Microtask (traditional work)
–1.67*** –81.2*** –1.56*** –79.1*** –1.78*** –83.2***
(0.063) (0.088) (0.094)
Female (male)
–0.31*** –26.7***
(0.050)
Age
0.03*** 3.4*** 0.01 0.5 0.06*** 5.7***
(0.012) (0.017) (0.016)
Age–squared
–0.00*** –0.0*** –0.00 –0.00 –0.00*** –0.1***
(0.000) (0.000) (0.000)
Education (no high school)
High school diploma
0.33 38.8 0.50 64.3 0.08 8.5
(0.204) (0.322) (0.207)
Technical degree
0.58*** 78.9*** 0.68** 97.6** 0.38* 45.8*
(0.215) (0.336) (0.222)
Bachelor’s degree
0.68*** 96.7*** 0.82** 127.4** 0.46** 57.9**
(0.208) (0.327) (0.213)
Postgraduate degree
0.41* 51.4* 0.62* 86.4* 0.15 16.4
(0.228) (0.342) (0.276)
Above postgraduate degree
0.67* 95.7* 0.80 122.8 0.58 78.8
(0.358) (0.492) (0.426)
Married
0.10* 10.6* 0.08 8.3 0.15** 16.2**
(0.052) (0.080) (0.071)
Household size
–0.04** –3.5** –0.05* –4.8* –0.03 –2.8
(0.017) (0.026) (0.023)
Urban
(rural)
–0.05 –4.4 0.02 2.3 –0.10 –9.4
(0.062) (0.081) (0.097)
Has another job
0.17* * 18.5** 0.19* 20.6* 0.15 16.5
(0.074) (0.095) (0.116)
Constant
1.88*** 2.21*** 1.42***
(0.305) (0.487) (0.363)
Observations
973 973 457 457 516 516
R–squared
0.575 0.575 0.572 0.572 0.587 0.587
Notes: Reference categories and robust standard errors in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.
Percentage changes are calculated using the formula 100 × [exp(coecient) – 1].
Sources: ILO calculations based on ILO global survey of crowdworkers (2017); U.S. Census Bureau, Current Population Survey (2017).
The role of digital labour platforms in transforming the world of work
24
As a robustness check, two additional models were specied in each country, analysing two activities
independently: one comparing workers in information service activities (ISIC 63) and workers on microtask
platforms performing similar tasks; and one comparing workers in oce administrative, oce support
and other business support activities (ISIC 82) and workers on microtask platforms performing similar
tasks (tables not presented here).
The signicant dierence in hourly earnings between workers on microtask platforms and those in the
traditional market also holds true when looking at the specic activity independently. When only those
online platform workers who engage in transcription and content moderation and workers in oce
administrative, oce support and other business support activities (ISIC 82) are included in the sample,
microtask workers are expected to earn 65 per cent less in India and 76 less per cent in the United
States. When looking at traditional workers in the information service activities sector (ISIC 63) and the
corresponding microtasks, microtask workers are expected to earn 62 per cent less in India and 87 per
cent less in the United States.
4B.2.2 Workers on freelance platforms
To estimate the dierences in hourly earnings of workers on freelance platforms by sex, education level,
experience and development status of their residence country, the OLS regression method was used.
The dependent variable is workers’ log total (paid and unpaid) hourly earnings in a typical week. Various
covariates were introduced in order to capture other factors that simultaneously aect hourly earnings,
including demographic characteristics and several online work-related variables.
In the rst model, all freelance workers were taken into consideration. In the other specications (models
(2) and (3)), respondents were divided by their country’s development status (see table A4.15).
The direction of the relationship between hourly earnings and the covariates is mostly constant across the
three models, although the signicance levels can dier depending on the model specication. Based on
the regression results of model (1), the factors that have a signicant relationship with hourly earnings are
age, having a postgraduate degree, having another paid job, having regular clients, number of platforms
used, having four to ve customers per week, and development status (see column 1 of table A4.15).
Workers in developing countries tend to earn 60 per cent less than workers in developed countries with
similar characteristics, which is signicant at the 99 per cent level.
Some covariates are signicantly correlated with hourly earnings in developed countries (model (2)), but
not in developing countries (model (3)). In developed countries, individuals with a postgraduate degree
tend to earn more compared to those without a university degree, while there is no such dierence in
the case of a bachelor’s degree. Furthermore, having regular clients is associated with higher earnings,
while undertaking tasks related to sales and marketing, as well as professional services, are associated
with lower earnings.
In developing countries, age, health status and having another paid job are correlated with hourly earn-
ings, but this is not the case in developed countries. Unlike in developed countries, there is no signicant
relationship between education and hourly earnings in developing countries.
Moreover, some variables are not associated with any signicant dierence in the hourly earnings in any
of the models. These include sex, marital status, household size, having children under six years of age,
experience, urban location, having a bachelor’s degree, migration status and main platform, as well as
undertaking certain tasks. There is no signicant dierence between the hourly earnings of male and
female workers with similar characteristics, regardless of the development status of the country they
reside in. In contrast with traditional work, where higher levels of experience are usually associated with
higher earnings, there is also no signicant relationship between experience and hourly earnings on
freelance platforms in any of the regression models.
Appendix 4. ILO surveys, interviews and statistical analysis
25
X
Table A4.15 Regression results: Workers on freelance platforms, global survey, by development
(dependent variable: log of hourly earnings in US$)
(1)
Total
(2)
Developed countries
(3)
Developing countries
Coecient
Percentage
change
Coecient
Percentage
change
Coecient
Percentage
change
Female
(male)
0.03 3.4 – 0.11 –10.8 0.05 5.6
(0.126) (0.245) (0.155)
Age 0.14*** 14.8*** 0.11 11. 2 0.20*** 22.6***
(0.050) (0.156) (0.061)
Age-squared –0.00** – 0.1** –0.00 –0.1 –0.00*** –0.2***
(0.001) (0.002) (0.001)
Education (secondary or below)
Bachelor’s degree 0.30 34.5 0.46 58.5 0.24 27.4
(0.182) (0.332) (0.250)
Postgraduate degree and above 0.37* 44.3* 0.53* 69.8* 0.31 35.9
(0.191) (0.286) (0.274)
Experience (below 6 months)
6 months to just under 1 year 0.12 13.2 0.56 74.2 –0.07 –6.7
(0.204) (0.357) (0.243)
1 year to just under 3 years –0.02 –1.8 0.03 2.6 –0.04 –4.3
(0.192) (0.410) (0.220)
3 years to just under 5 years 0.14 14.5 0.31 36.4 0.01 1.1
(0.216) (0.376) (0.262)
5 or more years 0.07 6.9 0.62 86.6 –0.22 –20.1
(0.244) (0.435) (0.293)
Married
(not married)
0.14 14.5 0.31 36.4 0.04 3.6
(0.135) (0.246) (0.164)
Household size –0.04 – 4.1 –0.13 –12.1 0.01 1.1
(0.048) (0.100) (0.055)
Has children under 6 years 0.16 17.6 0.28 31.8 0.06 5.8
(0.155) (0.293) (0.184)
Urban
(rural)
0.00 0.4 0.03 3.1 –0.01 –1.1
(0.149) (0.261) (0.201)
Migrant –0.12 –11.7 –0.25 –22.4 –0.01 –0.9
(0.192) (0.259) (0.299)
Has another job –0.35*** –29.3*** –0.06 –5.9 –0.48*** –37.9***
(0.127) (0.262) (0.164)
Health status is poor or very poor
(good or very good health)
–0.47 –37.2 0.77 116.8 –0.82* –56.2*
(0.424) (0.572) (0.443)
The role of digital labour platforms in transforming the world of work
26
(1)
Total
(2)
Developed countries
(3)
Developing countries
Coecient
Percentage
change
Coecient
Percentage
change
Coecient
Percentage
change
Upwork
(Freelancer)
–0.44 –35.4 –0.35 –29.5 –0.50 –39.6
(0.295) (0.424) (0.372)
Regular client 0.33* 39.3* 0.66** 94.1** 0.25 29
(0.176) (0.324) (0.194)
Number of platforms used 0.11 11.5 0.20 22.1 0.06 5.7
(0.075) (0.126) (0.095)
Number of customers/week (1)
2–3 customers per week – 0.11 –10.4 – 0.12 –11.5 –0.07 –6.9
(0.141) (0.231) (0.181)
4–5 customers per week 0.62*** 86.8*** 0.61* 83.5* 0.68*** 97.6***
(0.183) (0.362) (0.256)
Business services tasks 0.09 9.5 0.21 23.9 0.04 3.8
(0.130) (0.223) (0.183)
Technology-related tasks 0.13 13.6 0.11 11. 5 0.13 13.8
(0.162) (0.346) (0.201)
Data analytic tasks 0.05 5.3 0.17 18.7 0.04 4.1
(0.143) (0.314) (0.162)
Creative tasks 0.04 3.9 0.22 24.9 –0.11 –10.5
(0.130) (0.253) (0.166)
Sales and marketing tasks –0.09 –8.4 –0.53* –40.9* 0.04 4.1
(0.152) (0.268) (0.194)
Professional services tasks –0.10 –9.7 –0.47* –37.8* 0.03 3.5
(0.123) (0.239) (0.159)
Developing
(developed)
–0.91*** –59.6***
(0.141)
Constant –0.94 –0.64 –2.98***
(0.974) (3.066) (1.143)
Observations 294 294 91 91 203 203
R-squared 0.313 0.313 0.358 0.358 0.234 0.234
Notes: Robust standard errors in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.
Percentage changes are calculated using the formula 100 × [exp(coecient) – 1].
Source: ILO calculations based on ILO global surveys of workers on freelance platforms (2019–20).
X
Table A4.15 (cont’d.)
4B.2.3 Workers on online web-based platforms in China (EPWK and 680)
To capture dierences in hourly earnings of workers undertaking online web-based work in China, the
OLS regression method was used. The analysis was done separately for respondents undertaking online
work on the platforms EPWK, 680 and ZBJ. The analysis for platform k68 was not undertaken as the
sample size was too small. The results for ZBJ are not presented as the R-squared value of the model was
very low. The dependent variable is workers’ log total hourly earnings (paid and unpaid) in a typical week.
Various covariates were introduced including demographic and several online work-related variables.
Appendix 4. ILO surveys, interviews and statistical analysis
27
The results of the regression show that some variables are signicantly correlated with hourly earnings in
at least one of the models (see table A4.16). A gender pay gap is present among workers on the platform
680, where female workers tend to earn 32 per cent less than male workers, while there is no such
signicant dierence on EPWK. Furthermore, online workers with a postgraduate degree tend to earn
more on both platforms than those without a university degree, while there is no such dierence in the
case of a bachelor’s degree. In addition, more experience in online work is associated with signicantly
higher earnings on EPWK, while this is not the case on 680. Undertaking particular tasks is not associ-
ated with any signicant dierence in hourly earnings for most types of tasks. The exceptions include
technology-related tasks, which are associated with higher earnings on EPWK, and microtasks, which
are associated with lower earnings on 680.
X
Table A4.16. Regression results: Workers on online web-based platforms, China
(dependent variable: log of hourly earnings in US$)
EPWK 680
Coecient
Percentage
change
Coecient
Percentage
change
Female
(male)
–0.16 –14.7 –0.38* –31.5*
(0.233) (0.200)
Age –0.10 –9.3 0.01 1.0
(0.124) (0.099)
Age-squared 0.00 0.2 –0.00 –0.0
(0.002) (0.002)
Education (below bachelor’s degree)
Bachelor’s degree –0.50** –39.2** 0.15 16.5
(0.226) (0.207)
Postgraduate degree and above 0.86* 135.5* 0.75** 111.7**
(0.484) (0.368)
Married
(not married)
0.10 10.3 0.10 10.1
(0.297) (0.277)
Household size –0.05 –4.9 –0.02 –1.8
(0.104) (0.078)
Has children under 6 years –0.12 –10.9 0.26 30.0
(0.297) (0.268)
Community type (rural/outer suburbs)
Country city 0.34 41.1 –0.49 –39.0
(0.510) (0.346)
Small and medium city 0.22 24.9 0.06 5.9
(0.478) (0.283)
Big city
(non-provincial city)
0.46 58.3 –0.14 –13.0
(0.477) (0.331)
Provincial capital city 0.67 94.7 –0.24 –21.6
(0.460) (0.278)
Migrated to current community –0.35 –29.6 0.38* 46.8*
(0.294) (0.209)
Hukou (local rural)
Non-local rural –0.57* –43.4* 0.30 35.2
(0.318) (0.252)
Local urban 0.50 64.9 0.01 1.3
(0.311) (0.242)
Non-local urban 0.52* 68.1* 0.28 32.4
(0.301) (0.406)
The role of digital labour platforms in transforming the world of work
28
EPWK 680
Coecient
Percentage
change
Coecient
Percentage
change
Experience (below 6 months)
6 months to just under 1 year 0.82** 127.6** 0.36 42.9
(0.346) (0.256)
1 year to just under 3 years 0.95*** 159.0*** 0.22 24.6
(0.308) (0.251)
3 years to just under 5 years 1.24*** 245.5*** 0.14 14.8
(0.383) (0.328)
5 or more years 1.30*** 266.8*** –0.12 –11.1
(0.424) (0.415)
Has another job 0.34 41.2 –0.00 –0.2
(0.217) (0.191)
Has physical or mental health conditions –0.37* –30.9* –0.04 –3.8
(0.221) (0.190)
Number of platforms used 0.10 10.7 –0.01 –0.9
(0.110) (0.083)
Technology-related tasks 0.79* 120.8* –0.47 –37.5
(0.466) (0.373)
Creative tasks 0.35 41.5 –0.22 –20.1
(0.460) (0.281)
Sales and marketing tasks 0.52 69.0 0.32 37.3
(0.602) (0.378)
Professional services tasks 0.37 45.3 0.06 5.7
(0.489) (0.286)
Microtasks 0.18 19.5 –0.57** –43.4**
(0.435) (0.256)
Other tasks 0.26 30.0 –0.44 –35.5
(0.748) (0.652)
GDP of province (lowest province GDP)
Lower province GDP –0.40 –32.7 0.25 28.5
(0.303) (0.273)
Higher province GDP –0.22 –19.8 0.17 18.6
(0.280) (0.292)
Highest province GDP 0.08 8.6 0.53** 69.2**
(0.287) (0.257)
Constant –0.03 –0.01
(1.856) (1.502)
Observations 210 210 260 260
R-squared 0.277 0.277 0.158 0.158
Notes: Robust standard errors in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.
Percentage changes are calculated using the formula 100 × [exp(coecient) – 1].
Source: ILO calculations based on ILO survey of platform workers in China (2019).
X
Table A4.16 (cont’d.)
Appendix 4. ILO surveys, interviews and statistical analysis
29
4B.2.4 Workers on online web-based platforms in Ukraine
To capture dierences in the hourly earnings of workers undertaking online web-based work in Ukraine,
the OLS regression method was used. The dependent variable is workers’ log total hourly earnings (paid
and unpaid) in a typical week. Various covariates were introduced including demographic and several
online work-related variables.
The results show that some covariates are signicantly correlated with hourly earnings (see table A4.17).
A gender pay gap is present, with female workers expected to earn 26 per cent less than their male
counterparts. Furthermore, having a postgraduate degree is associated with 36 per cent higher earnings
compared to those with secondary education or below, which is not the case for a bachelor’s degree.
Some other factors, including larger household size, having another paid job, uploading work portfolio,
and asking past clients to complete feedback or rating, display a signicant positive correlation with
hourly earnings. Other variables, such as having children under 18 years, undertaking microtasks, or
having physical or mental health conditions, are associated with lower hourly earnings. Furthermore,
age, marital status, urban location, migration status, experience, platform, number of platforms used,
and most task types as well as most strategies are not associated with any signicant dierence in the
hourly earnings of online workers in Ukraine.
X
Table A4.17 Regression results: Workers on online web-based platforms, Ukraine
(dependent variable: log of hourly earnings in US$)
Coecient
Percentage
change
Female
(male)
–0.31** –26.3**
(0.126)
Age –0.01 –1.5
(0.038)
Age-squared 0.00 0.0
(0.001)
Education (secondary or below)
Bachelor’s degree 0.25 28.3
(0.160)
Postgraduate degree and above 0.31** 35.7**
(0.136)
Married
(not married)
–0.12 –11.0
(0.113)
Household size 0.15*** 16.2***
(0.050)
Has children under 18 years –0.24* –21.5*
(0.146)
Urban
(rural)
0.12 12.3
(0.186)
Migrant 0.06 6.1
(0.256)
Years of experience with platform work 0.01 1.2
(0.018)
Has another job 0.20* 22.1*
(0.111)
The role of digital labour platforms in transforming the world of work
30
Coecient
Percentage
change
Platform (Freelancer)
Kabanchik.ua 0.01 1.3
(0.325)
Upwork –0.02 –2.1
(0.345)
Other Russian/Ukrainian platforms –0.17 –15.5
(0.351)
Other 0.41 50.8
(0.403)
Number of platforms used –0.06 –5.9
(0.060)
Task type (Business services)
Technology-related 0.40 49.9
(0.261)
Data analytic –0.67 –48.7
(0.572)
Creative 0.09 9.0
(0.262)
Sales and marketing 0.08 8.1
(0.295)
Professional services 0.05 5.2
(0.222)
Microtasks –0.65*** –47.9***
(0.240)
Manual work 0.83** 129.2**
(0.367)
I upload several examples of my work portfolio 0.36*** 43.9***
(0.115)
I underbid projects so that I can gain experience on the platform 0.12 12.4
(0.137)
I ask a client to give me a good rating in return for my giving them a good rating 0.13 13.8
(0.141)
I ask a client to give me a good rating in exchange for lower remuneration 0.32 37.5
(0.278)
I completed classes or trainings to obtain certications on the platform –0.37 –31.1
(0.262)
I actively ask past clients to complete feedback/ratings 0.22** 24.9**
(0.111)
I perform work only for the clients I know in real life 0.32 37.7
(0.214)
Has physical or mental health conditions –0.50** –39.6**
(0.246)
Constant –0.19
(0.776)
Observations 647 647
R-squared 0.180 0.180
Notes: Robust standard errors in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.
Percentage changes are calculated using the formula 100 × [exp(coecient) – 1].
Source: ILO calculations based on ILO survey of platform workers in Ukraine (2019)
X
Table A4.17 (cont’d.)
Appendix 5
ILO Interviews with unions and associations
Table A5.1 List of interviews with unions and associations
Name of association/union Person interviewed Country Date of interview
1. Sociedad de Fomento Fabril Director of Public Policies Chile 6 February 2020
2. Philadelphia Limousine
Association & Philadelphia
Drivers Union
President United States, Philadelphia 2 April 2020
3. United Private Hire Drivers
(UPHD)
Co-founder United Kingdom 7 April 2020
4. App Personal Lawyer Argentina 10 April 2020
5. Gig Workers Matter Chair United States, Chicago 14 April 2020
6. Unionen Policy analyst specializing
in digital labour markets
Sweden 15 April 2020
7. National Union of Professional
e-hailing Driver Partners
(NUPEDP)
Representative Nigeria 15 April 2020
8. Asociación de Conductores
Uruguayos de Aplicaciones
(ACUA)
Representative Uruguay 15 April 2020
9. Sindicato Independiente
Repartidores por Aplicaciones
(SIRA)
President Mexico 22 April 2020
10. APOPL ATEC Representative Costa Rica 24 April 2020
11. The Movement Representative South Africa 28 April 2020
12. Ride-Share SACCO Limited
and the Digital Taxi Forum
Representative Kenya 28 April 2020
13. NiUnRepartidorMenos Representatives Mexico 30 April 2020
14. Riders’ Union Representative Republic of Korea 18 May 2020
