N A T U R A L R E S O U R C E S I N S T I T U T E

NRI Report No: 2768

SSR Project No. R8077 2002-4

Gender, Rights & Participation

in the Kenya Cut Flower Industry

by

Catherine Dolan
(Northeastern University, USA)

Maggie Opondo
(University of Nairobi, Kenya)

Sally Smith
(Institute of Development Studies, UK)
 NRI Report No. 2768.
SSR Project No. R8077 2002-4

GENDER, RIGHTS & PARTICIPATION

IN THE KENYA CUT FLOWER INDUSTRY

By

Catherine Dolan
(Northeastern University, USA)

Maggie Opondo
(University of Nairobi, Kenya)

Sally Smith
(Institute of Development Studies, UK)
 Dr Catherine Dolan
Department of Sociology and Anthropology
Northeastern University, Boston, MA 02115, USA
Tel. +001 617 373 3858. Email: [email protected]

Dr Maggie Opondo
Department of Geography, University of Nairobi, Nairobi, Kenya
Tel. +254 (0)20 33 4244. Email: [email protected]

Ms Sally Smith
Institute of Development Studies, University of Sussex, Brighton BN1 9RE, UK.
Tel. +44 (0)1273 873663. Email: [email protected]

2
Contents

Acknowledgements 5
Abbreviations 6
EXECUTIVE SUMMARY 7
SECTION 1: BACKGROUND 11
1. Introduction 11
2. Social Accountability in the Kenya Cut Flower Industry 13
2.1 Cut Flowers in the Global Economy 13
2.2 Kenya Cut Flower Industry 13
2.3 Structure of the Kenya Cut Flower Industry 15
2.4 Codes in the Kenya Cut Flower Sector 16
3. The Regulatory Context 19
3.1 Kenya Labour Legislation 19
3.2 ILO Declaration on Fundamental Principles and Rights at Work 20
3.3 Collective Bargaining Agreement 21
SECTION 2: CASE STUDY OF WORKERS IN THE CUT FLOWER INDUSTRY 23
4. Methodology & Case Study Sample 23
4.1 Methodology 23
4.2 The Company Sample 24
4.3 Companys Use of Codes 25
4.4 Composition of the Worker Sample 28
5. Employment Conditions on Cut Flower Farms 33
5.1 Security of Employment: Regular Employment Provided 33
5.2 Employment is Freely Chosen 34
5.3 Contracts 34
5.4 Working Hours are not Excessive 355
5.5 Remuneration: Wages must be Sufficient to meet Basic Needs of Workers 36
5.6 Harsh and Inhumane Treatment and Intimidation 388
5.7 Equity in Wages and Non Discrimination 40
5.8 Freedom of Association, Collective Bargaining and Worker Participation in Decision Making 42
5.9 Grievance Procedures 45
5.10 Health and Safety Provision of a Safe and Healthy Work Environment 45
5.11 Pregnancy 47
5.12 Beyond Codes: Other Worker Welfare Issues 48
SECTION 3: THE AUDITING PROCESS 53
6. Participatory Social Auditing 53

3
6.1 Standard Auditing Practice 53
6.2 Participatory Social Auditing 54
6.3 Assessment of Participatory Social Auditing Methods 55
7. Multi-stakeholder Approaches 58
SECTION 4: RECOMMENDATIONS 60
8. Ways Forward 60
8.1 Employers 60
8.2 Buyers (importers, wholesalers, supermarkets) 61
8.3 Auditing Bodies 62
8.4 Trade Unions 63
8.5 HEBI 63
8.6 Non Governmental Organisations 63
8.7 Kenyan Government 63
9. Conclusion 65
Appendices 68
References 75

4
Acknowledgements

We gratefully acknowledge funding for this research from the UK Department for International
Development. We are also very grateful to the companies who participated in the research, for
permitting their workers to participate in the project during working hours, and for taking the time to
share their views with us. We extend sincere thanks to all the workers who shared their experiences
of working in the cut flower industry with us, and the Kenyan Research Assistants who so diligently
elicited and recorded those experiences. Finally, we thank the two anonymous, independent referees
who gave valuable feedback on an earlier draft of this report, and the other members of the project
team for their advice and comments.

Disclaimer:
This working paper has been produced as part of a two year research project into gender and ethical
trade in South Africa, Kenya and Zambia. The project is funded by the UK Department for
International Development, SSR Project No. R8077 2002-4. The Department for International
Development (DFID) supports policies, programmes and projects to promote international
development. DFID provided funds for this study as part of that objective, but the views and
opinions expressed are those of the authors alone.

Paper reproduced by the Natural Resources Institute, a partner in the Ethical Trade in African
Horticulture project, as part of its Project Report Series.

ISBN: 0 85954 551 2

5
Abbreviations

AEAAZ Agricultural Ethics Assurance Association of Zimbabwe

BGI Federation of German Importers & Wholesalers of Cut Flowers
BRC British Retail Consortium
BVQI Bureau Veritas Quality International
COLEACP Comit de Liason EuropeAfrique-Caraibe-Pacifique
DFID Department for International Development
ETI Ethical Trading Initiative
EUREP European Retailers Representatives Group
FAO Food and Agriculture Organisation
FIAN Food First Information and Action Network
FLO Fairtrade Labelling Organizations International
FLP Flower Label Program
GAP Good Agricultural Practice
FPEAK Fresh Produce Exporters Association of Kenya
HACCP Hazard Analysis Critical Control Point
HCDA Horticultural Crops Development Authority
HEBI Horticultural Ethical Business Initiative
ICFTU International Confederation of Free Trade Unions
ICC International Code of Conduct for the Production of Cut flowers
IFOAN International Federation of Organic Movements
ILO International Labour Organisation, Geneva
ISEAL International Social and Environmental Accreditation and Labelling Alliance
KARI Kenya Agricultural Research Institute
KEPC Kenya Export Promotion Council
KEWWO Kenya Women Workers Organisation
KFC Kenya Flower Council
KPAWU Kenya Plantation and Agricultural Workers Union
LNGG Lake Naivasha Growers Group
MPS Milieu Project Sierteelt
NSSF National Social Security Fund
SAI Social Accountability International
SAN Sustainable Agriculture Network
SASA Social Accountability in Sustainable Agriculture
SA8000 Social Accountability standard
USAID United Agency for International Development

6
 EXECUTIVE SUMMARY

Cut flower production is now a major part of the Kenyan economy, but the countrys success in
supplying European markets has brought with it increased attention to the industrys social and
environmental impacts. Driven by the various concerns of consumers, retailers, auctions, European
regulators and civil society organizations, flower growers have to comply with a number of codes of
conduct developed by their major market-brokers, international organizations, national industry
associations and multi-stakeholder partnerships. These codes are part of a global increase in non-
governmental regulation, but nonetheless make reference to both Kenyan labour and health and
safety law, and the core conventions of the International Labour Organisation.

Our research examines how such codes address employment conditions in the industry, especially the
experiences of women who comprise the majority of the workforce. We use criteria from a selection
of labour codes adopted by major producers to assess how women are being affected by the
industry, and to show the extent to which womens priority issues are included. We have also used
the study to make a case for advancing participatory social auditing and local multi-stakeholder
initiatives as significant elements of code implementation.

The ways workers are affected depends to a degree on the employment contract, with permanent
workers experiencing significantly more security, opportunity and fringe benefits compared to
seasonal and especially casual workers. In the more progressive companies at least there is a shift
towards more permanent employment, reflecting the demand for more skilled labour and the
pressure brought through codes of conduct. However, significant numbers of workers have only a
non-permanent status, and it is these who are least likely to benefit from national labour law or
voluntary codes.

The fact that women make up the majority of non-permanent workers is a key reason why many are
unaware of codes and feel limited benefit from them. However, the gender dimensions of codes are
not straightforward, and are also related to age, marital status and place of origin. Nonetheless, in
the areas of job security, working hours, wages, discrimination and harassment, women appear to
have separate concerns from men, and moreover feel that these are not being addressed (or in some
cases even recognized) by any of the formal channels currently available to them. The study found
that there were a number of problems common to workers on all the farms in our sample, regardless
of the size of farm, the market outlet, which codes were applied and whether or not the code had
been audited externally. These issues include:

7
Employment Insecurity: Security of employment was a major concern for many women workers,
but especially non-permanent workers. While many employers in the study have begun to issue more
permanent contracts, and shift seasonal and casual workers to permanent employment, thirty-three
percent of the sample workforce remains in insecure jobs. Further, since this study was conducted
among code adopting farms, there is a high probability that insecurity is more prevalent in the
industry as a whole.

Overtime: Overtime is often compulsory and frequently exceeds the maximum hours set out in
codes and national law. Overtime payments also vary with many workers feeling that they are not
properly compensated. The likelihood of performing overtime is also linked to the employment status
of workers. For example, permanent workers, who are paid monthly, are more likely to work
overtime than casual workers, who are paid daily. For women, regular overtime makes it difficult to
balance productive and reproductive roles. Arranging childcare when overtime is required at short
notice is particularly problematic. Personal safety can also be an issue if transport is not provided
after dark.

Sexual harassment: Sexual harassment was reported on all farms, particularly by women who were
subject to sexual harassment from male supervisors. The persistence of sexual harassment, despite
the intentions of those managers wishing to eliminate it, is related to the hierarchical employment
structure of companies, coupled with the lack of senior female staff, both of which intimidates women
from reporting incidences of harassment.

Representation: Although there has been an increase in union membership over the past few
years, trade union representation is still low on cut flower farms. It also excludes those workers in
non-permanent job categories. To date there is no systematic forum through which all workers can
freely express their opinions and voice their concerns, particularly women.

Complaints procedures: Generally, companies have an informal approach to complaints, ranging

from an open door policy to suggestion boxes. This fails to recognize that many workers are in
vulnerable positions and fear approaching management. This is especially true for women workers,
who are more fearful of discussing issues with male supervisors.

Communication: Poor communication between workers, supervisors and management is the

source of many worker welfare issues. Management delegates authority and power to supervisors
(who are predominantly men), who often abuse this authority in their relations with workers.
According to workers, those who express grievances to the management without the supervisors
consent risk being sacked.

8
Wages: Wages in the industry are typically better than the government minimum wage and those
that could be obtained through viable alternatives in the region. However, they are nevertheless low.
While low wages impact on both men and women, women with sole responsibility for children are
particularly affected as they cannot afford to pay someone to look after their children while they are
at work.

Opportunities for Promotion: There are generally few opportunities for upward career
progression, which tends to diminish worker motivation. Among women, this partly relates to the
gendered allocation of jobs, rooted in socio-cultural perceptions and norms, which provides few
prospects for women to move into management positions. Promotion is also linked to corruption,
sexual favours/demands, tribalism, and nepotism.

Health and Safety: There have been substantial improvements in the health and safety practices
of companies. However, chemical exposure remains an important concern among workers. The main
issues raised were a lack of provision of protective clothing for workers handling chemicals, the non
observance of re-entry periods, and the exposure of pregnant women to chemicals. In some cases,
the perception of risk among workers exceeds actual risk, suggesting that companies may not be
providing sufficient training and/or accurate information to workers on where the dangers of chemical
exposure lie. On the other hand, several companies mentioned the difficulties of ensuring that
workers abide with health and safety regulations, especially the wearing of protective clothing.

Pregnancy and maternity leave: Lack of access to adequate maternity leave creates anxiety
about income security and can lead women to seek abortions and/or hide their pregnancies, both of
which can carry long-term health implications. Adequate childcare was widely recommended as a
significant benefit. Currently there is only one farm that has childcare facilities available, although
other companies understand that women face difficulties with childcare, and some are planning to
establish day care facilities in the future.

Transport: Where transport is provided, workers noted problems of unreliability, overcrowding and
lateness. Several workers also mentioned safety concerns about travelling home at night, particularly
women.

Lack of awareness of rights and codes: Although the introduction of codes of conduct may be
contributing to improved labour conditions, most workers are unaware of their employment rights and
of national legislation that is relevant to them. They are also unaware of codes despite the fact that
this research was conducted on farms that had a comparatively long history with codes. Poor
communication with workers about codes and their importance reflects the problem of communication

9
that exists between management and workers in the industry. This may change as companies
recognise the importance of worker involvement to achieve code compliance.

These employment problems are not insurmountable. The fact that labour conditions have improved
over the last year provides cause for optimism. However, greater participation of workers and other
stakeholders is essential to improving conditions in the industry, particularly for women who form the
majority of the workforce. This study shows that participatory social auditing is not only effective in
discovering areas of non-compliance that are not detected by other methods, but may be a means of
fostering better dialogue and change. This method has been used as part of the multi-stakeholder
approach to implementing codes found elsewhere in Africa, which is being introduced into the Kenya
flower industry through the Horticultural Ethical Business Initiative, linking government, civil society
and industry representatives.

There are various actions that different stakeholders can take to improve working conditions and the
situation of women in the industry, and our report includes suggestions for employers, buyers,
auditing bodies, civil society organisations and the Kenyan government. Underlying these
suggestions is a recognition of the need to consider the gender-specific impacts of the industry, and
for gender-sensitive solutions, both of which require not only capacity building amongst employers
and workers, but equally an understanding of how improving labour conditions relates to productivity
and quality, and can constitute an aspect of comparative advantage.

10
 SECTION 1: BACKGROUND

1. Introduction

Each day thousands of roses, carnations and other fresh flowers are shipped from Kenya to Europe,
where within 24 hours they embellish the shelves of supermarkets and florists. These flowers are
big business for Kenya. The sector is now the fastest growing in the Kenyan economy, outpacing
Kenya's traditional hard currency earners - coffee and tourism - and bringing new employment
opportunities for the unemployed. Women in particular have benefited from these opportunities, and
now comprise between 65-75 percent of workers employed in the industry.

Over the last few years, however, the industry has been beset by allegations of poor labour practices
and environmentally damaging production processes. Women, in particular, face difficult working
conditions due to their predominance in the most labour intensive aspects of production, where long
working hours and insecure employment are common. One way of tackling gender issues and
improving the employment conditions of men and women working in the sector is through social
codes of conduct, which gained prominence in the supply chains of African horticulture during the
1990s. Kenyan cut flower producers have been at the forefront in embracing such codes, both
through the development of their own industry codes and the adoption of overseas buyer codes.
Despite the positive steps that growers have taken to comply with codes, however, a number of
employment problems persist.

This report discusses how the process of code implementation can be enhanced to improve the
employment conditions of workers in the Kenyan cut flower industry. It presents the findings of
Phase II of a research project on gender and social codes of conduct in the African export horticulture
sector. Phase I provided an initial mapping of codes of conduct and their gender content, and the
extent to which stakeholders participated in the code development process (Barrientos, Dolan and
Tallontire 2001). It found that a plethora of codes were introduced in the sector, coming from
European supermarkets, importers, exporters and trade associations, but that the extent to which
code content addressed gender concerns was highly variable. The study also found that while codes
can be effective instruments to raise the standards of permanent workers, they generally overlook the
conditions faced by men and women in insecure forms of employment (temporary, casual, migrant
and seasonal), which often form the majority of the horticulture workforce.

Phase II built on these findings by focusing on the process of code implementation. Through in-
depth research, Phase II examined the gendered needs and rights of cut flower workers, as

11
articulated by workers themselves, and how these could best be addressed by codes of conduct. The
research paid particular attention to vulnerable groups such as seasonal, casual and migrant workers,
who typically face a different set of constraints and opportunities in employment. The aims of the
project were to:

analyse how ethical trade can enhance the economic and social rights of women and men
workers in African export horticulture;

identify best practice in implementing gender-sensitive ethical trade based on worker and
stakeholder participation.

Phase II found that for codes to deliver substantive changes in working conditions, they need to
ensure that the rights of all workers are protected, including those of marginal workers and women.
This is not only related to the content of codes but also to the capacity of auditing processes to move
beyond a surface assessment of working conditions. While social codes can be an effective tool for
improving working conditions, this is unlikely to occur if the process of code implementation fails to
identify poor employment practices and issues of greatest concern to workers.

This report is organised as follows. Chapter 2 provides an overview of the cut flower industry in
Kenya, and the types of codes that are applied in the sector. Chapter 3 presents the employment
context, and the national and international labour legislation that governs workers in the sector.
Chapter 4 outlines the methodology used for the research. Chapter 5 presents the nature of
employment and working conditions found in the industry and the varying perspectives of workers
and employers toward these conditions. Chapter 6 explores the benefits of participatory social
auditing tools for eliciting workplace issues, especially gender issues, and identifies which tools are
most successful at obtaining certain types of information. Chapter 7 describes the value of a local
multi-stakeholder approach to code implementation and the extent to which stakeholders in Kenya
have embraced the process thus far. Chapter 8 identifies policy recommendations for best practice in
code implementation and Chapter 9 summarises and concludes.

12
 2. Social Accountability in the Kenya Cut Flower
Industry

2.1 Cut Flowers in the Global Economy

The world market for cut flowers has grown consistently since the early 1980s. The total acreage
allocated to cut flower production worldwide is now over 200,000 hectares, with roses, carnations,
and chrysanthemums the dominant varieties (ITC 2001). While the Netherlands remains the largest
producer of cut flowers worldwide, 1 developing countries such as Colombia, Kenya, Ecuador, and
Zimbabwe have become strong players in global markets. By 1998 exports from developing countries
comprised 29 percent of the world total (ITC 2001), and provided employment to approximately
190,000 people (PANUPS 2002). African countries in particular are important suppliers to European
markets, with ACP (African Caribbean and Pacific) countries accounting for 43 percent of EU imports
in value and 49 percent in volume (see Table 1).

Table 1: Imports into EU from non EU Countries (in million ECU)

%
Top
Increase/
Suppliers to 1994 1995 1996 1997 1998 1999 2000
Decrease
EU
94-00
Kenya 65.9 75.7 84.2 99.1 110.8 130.1 153.0 57
Colombia 88.1 94.6 94.3 104.0 102.5 93.4 104.3 16
Israel 109.7 110.5 139.1 132.2 145.8 103.1 100.6 -9
Ecuador 15.1 21.7 26.9 38.5 54.9 62.5 78.1 81
Zimbabwe 27.7 36.0 40.0 45.3 50.4 51.2 66.1 58
Thailand 20.7 20.5 18.4 18.9 16.6 16.1 18.3 -13
Zambia 3.4 4.4 6.8 8.5 12.2 16.0 17.5 80
Uganda 1.0 2.1 3.2 4.4 4.8 5.6 10.6 90
South Africa 7.6 8.3 8.1 8.6 8.2 8.3 9.1 16
Tanzania 2.3 3.2 3.9 5.1 5.4 7.6 8.4 73
Total Non 402.5 441.8 488.5 518.9 565.7 540.1 612.3
EU
Total ACP 108.1 128.6 145.1 169.8 191.9 217.3 263.8
% ACP 27 29 30 33 34 40 43
Source: Based on Eurostat-Elaboration, COLEACP (2002)

2.2 Kenya Cut Flower Industry

Kenya boasts the oldest and most successful cut flower industry in Africa. The industry has expanded
from a small scale trade in the 1950s/1960s to become one of the most important off-season
suppliers of cut flowers in the world (see Appendix 1). This expansion has been supported on the

1
In 1996 the Netherlands accounted for 59 percent of global cut flower exports (Thoen et al. forthcoming).
13
demand side by more affluent consumers in Northern countries, and on the supply side by Kenyas
preferential trade status into European markets coupled with the governments hands-off approach
(Thoen et al. forthcoming).

The industry is widely considered an economic success story, with the value of cut flower exports
increasing from KSh 940 million in 1990 to Ksh 14, 792 million in 2002 (see Table 2). Cut flowers are
now the nations second largest source of foreign exchange in agriculture, bringing in US $110 million
in 2001 (FAO 2002) and providing employment to an estimated 40,000 -50,000 workers2 (HCDA 1999
statistics cited by KFC 2002).

Table 2: Export Volume and Value: 1980-2001 Selected Years

Year Volume Value
(Tons) (Kshs. Millions)
1980 7422 227
1985 10,000 463
1990 14,425 940
1995 29,374 3,642
1996 35,212 4,366
1997 35,853 4,900
1998 30,221 4,857
1999 36,992 7,235
2000 38,757 8,650
2001 41,396 10,627
2002 52,106 14,792
Source: HCDA Statistics (Unpublished)

By far the largest proportion of Kenyan flower exports is supplied to Europe. Kenya has surpassed
Colombia and Israel as the largest supplier to the EU, accounting for 58 percent of all ACP cut flower
exports to the region (Hennock 2002, COLEACP 2002). Within Europe, the Netherlands is the leading
destination for Kenyan flowers (see Table 3), followed by the UK and Germany. A significant quantity
(approximately 80 percent) of Kenyan flowers is also re-routed to other countries through the re-
exports of the Netherlands.

The Dutch flower auctions have historically been the most important channels through which Kenyan
flowers are distributed to European wholesalers and retailers. This remains the case overall;
however, in several European countries supermarkets now dominate the cut flower trade. In
Switzerland, the two major supermarket chains (Migros and Coop) together account for 65 percent of
all Swiss flower sales (Videa 2001, The International Floriculture Quarterly Report 2002). Similarly,
by 2000 the share of supermarket cut flower sales in the UK reached 43 percent, 3 up from 30 percent

2
These levels of employment are quickly approaching those of coffee (59,864) and tea (77,024), the nations
most prosperous export crops, and are highly significant in a country where high unemployment and land
constraints limit opportunities for income generation (Data provided by Central Bureau of Statistics, 2003).
3
See international trade statistics from the International Floriculture Quarterly Report cited on
www.pathfastpublishing.com archives:
(http://www.pathfastpublishing.com/Archive02/Market%20share%20uk%202000.htm).
14
in the late 1990s (Galinsky 1997). African producers have benefited from this trend with 60 to 70
percent of Kenya flowers now destined for retail chains (van Liemt 1998). Whether cut flower
producers supply the retail market directly or via the Dutch Auctions greatly influences the codes that
are applied, the type of monitoring and verification that occurs, and the subsequent potential to
achieve gender sensitivity. This will be explored in the Chapter 4.

Table 3: Market Destination of Kenya Cut Flower Exports by Volume, 1999

Market Destination Volume (Tons) Percent

Holland 25,233 67.8
UK 7,170 19.3
Germany 2,207 5.9
Switzerland 491 1.3
South Africa 508 1.4
Sweden 356 1.0
Saudi Arabia 243 0.7
Italy 119 0.3
Belgium 7 0.0
Dubai 74 0.2
Others 496 1.3
Total 37,206 100.0
Source: HCDA statistics (unpublished)

2.3 Structure of the Kenya Cut Flower Industry

In contrast to the situation in the 1970s and 1980s when only one or two companies accounted for
the bulk of cut flower exports, there are presently over 500 producer/exporters growing cut flowers in
Kenya. Production for export is largely concentrated on some 60 or so medium to large scale flower
operations of which the twenty-five largest producers account for over 60 percent of total exports.
The larger flower operations range in size from 20 to over 100 hectares under production with a
labour force ranging from 250 to 6000. Supplementing these larger growers are approximately 50
medium scale commercial growers and an estimated 500 small growers (see Table 4).

15
 Table 4: Kenyan Cut Flower Producers

Category Approximate production area and crop Typical Features

Large >20 ha. of protected production plus Manage own export operations
producer- open field production May buy in from outgrowers
exporter Sophisticated infrastructure, expatriate Diversified markets including direct sales
management to supermarkets
Rose + carnation + open field flowers Large employers (250 6000 staff)
Small - Medium 2-20 ha, including up to 10ha of Own export and/or act as outgrowers
sized producers protected production May in turn buy in from outgrowers
Sophisticated infrastructure Sell through auction system, limited direct
Mostly rose + some open field flowers sales
Employment ~100 staff
Account for 20-30 percent of exports
Smallholder 0.25-2 ha, open field crops, particularly Act as outgrowers
producers alstromeria Product sold through auctions
Low input system with very little
investment
Mostly family labour
Account for 5-10 percent of exports
Source: Updated from Blowfield et al., 1998

2.4 Codes in the Kenya Cut Flower Sector

Over the last decade consumers have placed increasing pressure on companies to account for the
ethical impact of their operations. As a result, voluntary codes, labels and regulatory standards
covering the production processes of Southern producers exporting to Northern markets have
multiplied. These initiatives have become increasingly widespread in labour and resource-intensive
industries such as export horticulture. 4

During the 1990s a number of regulatory measures were adopted to address the quality, social and
environmental impact of global trade. In the horticulture industry these standards were initially
imposed to protect the health and safety of consumers in importing nations and include various trade
agreements and standards enacted at international, regional and sub-regional levels. By the mid
1990s, the emphasis on governmental regulation was augmented by a number of voluntary
approaches that extended beyond the impact on consumers in the North to the implications for the
health, labour and human rights of people involved in the supply chains of developing countries. In
the cut flower industry this was evident in two trends. Firstly, consumers took an increased interest
in the environmental dimensions of cut flower production as concerns about pesticide exposure and
environmental contamination became widespread. Exposure to such pressures, and the fear of the
commercial loss associated with failure to remedy the situation, prompted flower traders and
producers to develop their own codes of conduct with clear environmental standards (Videa 2001),
notably in the Netherlands but also in several developing countries such as Zambia, Zimbabwe,

4
See van Liemt (1998), Murray (1997), Caldwell (1998), and Roberts and Robins (2000).
16
Uganda, and Colombia. Secondly, by the late 1990s, the cut flower industry was exposed to a
number of well organized and vocal campaigns from NGOs and consumers regarding labour rights
abuses such as excessive working hours and low wages. In response, growers in both Latin America
and Africa adopted codes of conduct to address the working conditions on flower farms.

In Kenya, the cut flower industry quickly became one of the most codified agriculture sectors in the
world. By the mid 1990s most of the leading Kenyan producers had applied codes in order to gain or
maintain access to markets and/or to satisfy the requirements of customers for environmental and
social responsibility. These codes were introduced from three different origins:

dominant buyers such as supermarkets and importers/wholesalers (company codes);

trade associations linked to the horticultural sector (sectoral codes); and

independent bodies comprising business and a range of civil society organizations (multi-
stake holder codes).

Company Codes: In the mid 1990s major importers and retailers established voluntary standards
for reasons of brand definition, quality and safety assurance. 5 Major supermarkets, especially in the
UK, implemented company-specific codes along their supply chains (e.g., Waitroses Responsible
Sourcing, the Co-ops Sound Sourcing and Tescos Natures Choice). Implementation guidelines
were developed by the companies themselves and monitoring was internal.

Sectoral Codes: Company codes have become less common since the adoption of sectoral and
multi-stakeholder codes. Several influential sectoral codes have been developed in the North such as
EUREPGAP6, BRC 7 and MPS 8, which have been adopted by Kenyan growers either voluntarily or as a
requirement to supply certain buyers. MPS, in particular, is extremely important for Kenya flower
growers who supply the Dutch Auctions (MPS has certified between 70 and 80 percent of all flowers
in the Dutch auctions). Like EUREPGAP, MPS originated as a technical standard to reduce the
environmental impact of cut flower production, but added an optional social chapter drawing on the
Universal Declaration of Human Rights and ILO Conventions in 2001. Separate labour standards have
also been applied to the flower industry through the International Code of Conduct for Cut Flowers
(ICC), a sectoral code developed by NGOs and trade unions in Europe, and supervised by the

5
The implementation of standards along the chain is closely associated with managerial and technological
innovation in the supply chain such as sophisticated monitoring/tracking systems (EDI, EpoS and ECR) and
category management.
6
European Retailers Produce Working Groups (EUREP) code for Good Agricultural Practice (GAP) in Fresh Fruits
and Vegetables, and in Flowers and Ornamentals.
7
British Retail Consortiums (BRC) Technical Standard for Companies Supplying Retailer Branded Food Products.
8
Milieu Programma Sierteelt (MPS) environmental certification standard.
17
directorate of the German-based Flower Label Program (FLP). Most recently MPS' Social Chapter has
been benchmarked against the ICC although MPS will retain its own environmental and social
qualification system (MPS-A, B, C and MPS SQ). In addition, although not a code of conduct per se,
the Max Havelaar criteria for Fairtrade Cut Flowers have also been adopted on several Kenyan farms. 9

At the same time a number of horticultural producer associations in developing countries have
developed their own codes as a way to protect the image and the legitimacy of their industries in
European markets. In Kenya these include the Kenya Flower Council (KFC), the Fresh Produce
Exporters Association of Kenya (FPEAK), and the Kenya National codes. Each of these targets a
slightly different type of grower and requires different approaches to verification. However, it is
uncertain how long the separate initiatives will continue, especially given the recent cooperative
agreement between KFC and MPS, which together account for over 90 percent of Kenya's export
flower business.

Multi-stakeholder Codes: Most recently multi-stakeholder codes such as the ETI Base Code and
HEBIs Kenya Base Code on Social Accountability for the Flower Industry have emerged in the
industry10 . Growers supplying UK supermarkets are especially affected by the ETI Base Code as
seven of UKs largest supermarkets are ETI members 11 and have agreed to apply the code to all their
fresh produce suppliers. While the ETI Base Code is not an auditable standard, the ETI has initiated a
process of learning to identify good practice in multi-s takeholder approaches to monitoring and
verification.

9
A new consumer label, Fair Flowers & Plants (FFP), has also been developed through an alliance of trade
unions, non governmental organisations (NGOs) and international flower trade organisations. It has not yet
been formalized but is based upon the ICC Rules for Implementation and its Guidelines as well as the MPS
qualification standards.
10
The Horticultural Ethical Business Initiative (HEBI) is explored in more detail in Chapter 7.
11
Supermarket members of the ETI are ASDA, the Co-Op Group, J Sainsbury, Marks & Spencer, Safeway,
Somerfield, and Tesco. They are applying codes to all their own brand products, including fresh produce.
18
3. The Regulatory Context

Codes of conduct are part of a trend towards self-regulation that is itself an element of the dominant
pattern of globalisation (Blowfield 2003). However, codes are nonetheless linked to national and
international frameworks, and must be used in conjunction with them to protect workers rights. In
fact, national and international legislation, as well as the provisions embodied in Collective Bargaining
Agreements, can complement and, in some cases, strengthen the provisions of codes. Many (though
not all) codes are explicitly based on core ILO conventions and the UN Declaration of Human Rights,12
and state that relevant national legislation must be applied. In this case, the standard set by the code
will largely depend on the broader legal and social framework operating in the country and the extent
to which it addresses issues of gender inequality. For example, in countries where labour laws are
progressive and rigorously enforced, this can give codes more extensive coverage than the specific
criteria within the code itself. On the other hand, in countries such as Kenya with comparatively weak
labour legislation, code compliance can provide a minimum floor for employment conditions
(Barrientos, Dolan and Tallontire 2001).

Assessing the potential of codes to ensure gender sensitivity in the workplace therefore requires
reviewing relevant international conventions and national legislation. The following briefly reviews the
ILO Conventions, Kenya national legislation and the Collective Bargaining Agreement (CBA) of
relevance to workers in the cut flower sector.

3.1 Kenya Labour Legislation

Kenya has a range of legislation and government policy addressing employment conditions including
both labour and health and safety legislation (see Appendix 2); however, the coverage of these
provisions varies. This is particularly the case with regard to laws designed to protect women
workers from discrimination and guarantee them equal opportunities in the workplace. The Kenyan
Constitution extends equal protection of rights and freedoms to men and women, but only in 1997
was the Constitution amended to include a specific prohibition of discrimination on the basis of
gender (IBLF 2002). The government has yet to enact the statutes that recognise equal pay for equal
work, nor are there any specific legal protections against discrimination on the basis of sex.

With regards to labour, the Employment Act (Chapter 226 of the Kenyan Constitution) and
Regulations of Wages and Conditions of Employment Act (Revised Edition, 1980) are the governing
pieces of legislation for regulating employment conditions in Kenya. The Employment Act codifies the

19
rules of the workplace, including basic conditions and benefits of employment such as protection of
wages, housing, and health and welfare. While the Act contains certain provisions regarding women
and young persons, it fails to address gender inequality and workplace violations such as sexual
harassment and discrimination (IBLF 2002). Even those laws that aim to safeguard womens
employment rights are biased against, rather than in favour, of them. For example, Section 28 of the
Employment Act prohibits women and children to work in certain areas and the employment of
women in certain types of employment at night (Dwasi 1999).

The Regulations of Wages and Conditions of Employment Act (Chapter 229) covers matters related to
remuneration and conditions of work such as minimum wages, working hours, and payment for
overtime and holiday leave. Specifically, it aims to maintain wages and conditions of work at a level
determined by the Minister. Again, this law does not address discrimination (IBLF 2002). Thus, in a
national context such as Kenya, where legislation can be weak, the standards embodied in codes,
coupled with relevant international conventions, are important vehicles for ensuring and improving
upon the quality and gender-sensitivity of workers employment.

3.2 ILO Declaration on Fundamental Principles and

Rights at Work
The labour and human rights criteria used in many codes of conduct draw in part on the international
conventions of the ILO, particularly the core conventions set out in the Declaration on Fundamental
Principles and Rights at Work. These cover freedom of association, collective bargaining, forced
labour, child labour, discrimination and equal remuneration (ILO 1998). ILO conventions are open for
ratification by member States, and once ratified must be accommodated under national legislation
(ibid.). Kenya ratified core conventions on equal remuneration (100) and discrimination (111) 13 in
2001 along with conventions on family responsibility (156) and maternity rights (183), although it has
yet to enact enabling legislation (IBLF 2002). Moreover, Kenya has yet to ratify the core convention
on freedom of association, although there are signs of change in the governments position on labour
law as reflected in the recent request to the ILO for assistance in the review and reform of relevant
legislation (e.g. elimination of casual employment).

Codes that stipulate compliance with ILO conventions can strengthen workers employment rights,
however this depends on the type of employment that workers perform. Conventions are based on

12
None of the codes covering the Kenya cut flower industry currently make reference to the African [Banjul]
Charter on Human and Peoples' Rights, which covers economic, social and cultural rights as well as civil and
political rights, and has been ratified by more than forty African states.
13
Convention 111 does not specifically prohibit sexual harassment. Nor does Recommendation 111 clarify that its
prohibition of sex discrimination incorporates sexual harassment, although the ILO Committee of Experts has
alluded to a prohibition of harassment by virtue of Convention 111 (IBLF 2002).

20
the notion that employment is full time and permanent, and their coverage of temporary employment
is limited (Ladbury and Gibbons 2000; Seyfang 1999) (see Figure 1). Despite the improvements that
some growers have made, a large number of cut flower workers remain in non-permanent positions,
the majority of whom are women. As Figure 1 illustrates, codes, together with national legislation
and international agreements, can potentially affect the employment conditions of permanent
workers, however, they are weak in their capacity to reach informal categories of employment.

Figure 1: Mapping of the Reach of Labour Protections

National Legislation Codes ILO Conventions

Formal
Employment
Permanent

Temporary
Informal
Employment
Seasonal

Casual

Contract Smallholders

KEY:
Employment below this line is typically informal but can be formalized through provision of
employment protections

3.3 Collective Bargaining Agreement

A growing number of voluntary codes include criteria on the right to collective bargaining. Many
workers in the cut flower sector are covered by the collective bargaining agreement (CBA) between
the Flower Growers Group of the Agricultural Employers Association (AEA) and the Kenya Plantation
and Agricultural Workers Union (KPAWU). 14 The CBA, signed in February of 2002, establishes

14
KPAWU is an ICFTU-affiliated union representing workers in horticulture, tea, coffee and other agriculture
sectors.
21
minimum wages and terms and conditions of employment for unionised workers on commercial cut
flower farms. This is being implemented by 20 companies in Kenya (three of which were among
those researched in this study). In contrast to national legislation, the CBA extends several benefits to
seasonal workers, providing some protections to workers on farms that have signed the CBA.

In the next section, we take a closer look at the situation of workers through a presentation of
information derived from a case study of employment on farms and in packhouses. This information
points more clearly to the need for government, trade unions, and private sector initiatives to be
strengthened to ensure the provision of full employment protections.

22
 SECTION 2: CASE STUDY OF WORKERS IN
THE CUT FLOWER INDUSTRY

4. Methodology & Case Study Sample

4.1 Methodology
This report is based on research conducted during July-August, 2002, which examined how the
process of code implementation could be enhanced to improve employment conditions, particularly
for women workers, who comprise the majority of the workforce. The study was also concerned with
the potential for advancing participatory social auditing and local multi-stakeholder initiatives in the
process of code implementation.

The methods included the use of semi-structured interviews (SSIs) and participatory focus group
discussions (FGDs) with workers. One hundred SSIs and 13 in-depth FGDs were conducted with
workers employed in five and four companies, respectively, located in Nakuru, Kiambu and Thika
districts. The companies were purposively selected to ensure coverage of different types of codes
(see Appendix 3). Another important criterion for the selection of companies was the willingness of
the management to participate in the research project. 15 Because companies in this study were
applying codes, and were willing to participate in the research project, they may be among the most
progressive in the industry, and therefore are not representative of the industry as a whole.

The sample of 100 workers was drawn from the five companies in proportion to their total population
of workers (see Appendix 4). The sample was selected and stratified according to gender,
employment status (permanent, seasonal, or casual) and work site (packhouse, greenhouse, and
Crop Development Unit (CDU)). The 13 FGDs were conducted with 85 workers from the four firms
applying codes. 16 The focus groups ranged from five to eight participants and the composition varied
according to gender, nature of work, and employment status. Sixty-six percent of FGD participants
were women, and 80 and 20 percent were permanent and seasonal workers, respectively.

15
Collection of primary data through semi-structured interviews and focus group discussions with workers was
time consuming. On average one semi-structured interview and focus group discussion took 40 minutes and
three hours respectively.
16
One of the five farms on which we conducted worker interviews did not apply a code. They were selected to
compare the experience of worke rs on code farms with those on farms without them.
23
Worker interviews and focus group discussions were supplemented by in-depth interviews with
management personnel on seven cut flower farms (of which 5 were included in the above sample).
Some or all of the following management staff participated in the company interviews: company
directors/owners, general managers, human resource managers/directors, and administration
personnel. In addition, over 25 key informant interviews were conducted with government officials,
trade associations, NGOs and union representatives.

4.2 The Company Sample

The selected companies have an average of 17 hectares of flowers in production (see Table 5), with
three of the seven also producing vegetables. Rose production is dominant: four of the farms
produce only roses and two produce roses and other flowers such as lisianthus and carnations. One
farm produces non-rose flowers including hypericum and zantedeschia, and the last produces
chrysanthemum cuttings. Four of the companies are Kenyan owned, including white, Indian-origin
and African owners (and in one case all three), and two are European owned. The last is a joint
venture between a Kenyan and a Dutch man.

Table 5: Flower Production (hectares) on Sample Farms

Farm Hectares of
Flowers
Farm 1 15.0
Farm 2 36.0
Farm 3 5.0
Farm 4 11.3
Farm 5 13.5
Farm 6 30.0
Farm 7 8.0
Average Size 17.0
Source: Research Data, 2002, Company Interviews

The principal market region for all seven companies is Europe, with five of the seven exporting
exclusively to European countries (see Table 6). Within Europe, the UK and the Netherlands are the
main market destinations, with two companies selling almost 100 percent of their product into UK
supermarkets.

24
 Table 6: Market Destinations of Sampled Cut Flower Farms

Region Farm 1 Farm 2 Farm 3 Farm 4 Farm 5 Farm 6 Farm 7

Netherlands majority 31% 30%
UK approx 100% 60%
100%
Germany 20%
Switzerland small 6% majority
amount
France 26%
Other mainland 8%
Europe
Australia
Middle East
Reunion Islands
USA
Other 10%
Source: Research Data, 2002, Company Interviews

4.3 Companys Use of Codes

Most of these companies began applying codes of conduct in the mid-1990s to satisfy the
requirements of their overseas customers, and have steadily increased the number they apply (see
Table 7). All but one of the seven companies uses at least one code of conduct relating to
employment practice, with an average of three codes per farm. Several companies were also
applying codes largely unrelated to working conditions (e.g., EUREPGAP, BRC and the HAACP
system). The most widely applied code among the sample is the Kenya Flower Council (KFC) code
(five farms), followed by the Lake Naivasha Growers Group (LNGG) code (three farms). Two farms
are MPS certified, 17 one of which is Socially Qualified, and two farms are FLP certified, one of which
is also certified by Max Havelaar to sell on the Fair Trade market in Switzerland. In general, the type
of code that an exporter is likely to apply is directly related to the market they are supplying. Kenyan
producers supplying the Dutch Auctions (the majority) tend to apply MPS, while those supplying UK,
Swiss, and German markets are likely to adopt the KFC, Max Havelaar, and FLP codes respectively.

Two of the local codes (KFC and LNGG) are audited by the industry itself (first party auditing), while
the FPEAK code is audited by a third party commercial auditor. The other codes are generally subject
to second party auditing or monitoring by buyers or code organisations. Although EUREPGAP requires
third party auditing, the social content of the code is minimal and according to producers only
superficially audited. Some third party audits using the ETI Base Code have taken place, but only in
isolated cases. The details of auditing will be covered in Chapter 6.

17
One farm was previously MPS certified, but withdrew from MPS when MPS announced its collaboration with
KFC.
25
 Table 7: Codes of Conduct Applied on Farms, Including Date of Uptake

Code Farm 1 Farm 2 Farm 3 Farm 4 Farm 5 Farm 6 Farm 7

KFC 1997 1996 1999 1998 yes Only
through
MPS
FPEAK Member Member 1990s
but dont but dont
use code use code
ETI 2000 2002
MPS 1997-SQ ex-MPS 1996
FLP 1998 1999
Max Havelaar 2001
EUREPGAP 2002/3 2001
BRC 2000 2000
LNGG 1996 1996 Member yes
but dont
use code
M&S Before
Safeways 1995
Source: Research Data, 2002, Company Interviews

Despite initial reservations, most companies found codes to be constructive, both as a management
tool and in enhancing their awareness of their legal and social obligations to their workers. There
were a number of benefits reported, including a better understanding of national and international
labour laws, improved welfare and training for workers, and increased understanding of sensitive
employment issues and advice on how to handle them. Most companies also realised that achieving
the quality standards required to access the more profitable overseas markets required more
professional farm management, including better human resource management, and some equated
worker welfare directly with the productivity of the company (Smith et al. 2003).

However, codes are not without their critics and several companies raised a number of concerns.
These included the large number of codes that growers were faced with, the perception that codes
were forced upon them by overseas buyers, and the costs of implementation. With the exception of
the costs associated with maternity leave, 18 most of the growers did not perceive the costs of
implementation as insurmountable. However, the costs and time entailed in ongoing monitoring and
auditing were seen as an unfair imposition. Similarly, some interviewees felt that the way codes were
introduced, as a top-down Northern-driven initiative, reflected the unequal power relations between
buyers and producers, with increasing demands being made by buyers with little reward for
compliance. For example, there were some feelings of resentment about buyers demanding that
suppliers comply with codes without any guarantee of long-term purchasing agreements. Other
growers felt that buyers had double standards, with demands for code compliance made in some
countries (e.g., Kenya) and not in others (e.g., Egypt). Several claimed that buyers will continue to

26
source from countries where working conditions were far worse than Kenya if they can get a better
price. Hence, in a context of increasing demands for ethical production, some producers intimated
that more consideration should be awarded to the nature of producer-buyer relationships (Smith et
al. 2003).

Nevertheless, the introduction of codes (coupled with pressure from media and civil society
organisations) is contributing to a heightened awareness of ethical production in the Kenya
floriculture industry and several producers have joined a local multi-stakeholder initiative to eliminate
unethical business practices. Indeed one company claimed that it would be better for rogue
companies to go out of business, as they damage the reputation of the industry and jeopardise the
benefits it brings to the country. This partly reflects the fact that the farms in our study were
among the more progressive, better practice farms, but it also signals the growing importance of
ethical business practices for both Northern buyers and Southern suppliers. This business climate,
including public concern for labour rights and the pressures for code compliance, have led some
companies to institute a number of positive changes, including:

Health and Safety: Increased use of personal protection equipment (PPE), regular health and
safety (H&S) training, and improved H&S facilities. Improved sanitation, including the provision
of drinking water, and separate toilets and showers for men and women.

Maternity Leave: Increases in the duration of maternity leave from 8 to 12 weeks and the
provision of maternity leave to seasonal workers.

Formalisation of employment relations: Introduction of written contracts detailing terms and

conditions, formalisation of grievance procedures, an increase in permanent workers and a
reduction in the use of long-term casual workers, and the introduction of pro-rata benefits to
seasonal and temporary workers.

One company has taken a number of additional steps to specifically address the needs and concerns
of women workers (see Box 1). All of these positive changes are an indication of the progression of
codes towards greater gender sensitivity. For example, a safe and hygienic working environment is
particularly important for pregnant women and breastfeeding mothers. Similarly, the formalisation of
employment relations increases the proportion of female employees receiving important benefits such
as sick leave and maternity leave. Such changes will be discussed further in Chapter 8. We now turn
to an analysis of employment in the study sample, identifying the composition of the labour force and

18
In contrast to Kenyan law, MPS and FLP require three months maternity leave, although MPS maternity leave
may include the period of annual leave.
27
the nature of employment concerns, as well as the perception of workers and management toward
the key provisions contained in codes.

Box 1: Example of Good Practice Relating To Women Workers

One company in Kenya has established a Gender Policy and Gender Committee specifically to address
issues of concern to women workers. The gender committee comprises nine female representatives
elected from all areas of the workforce, who meet with senior management to raise issues and seek
solutions. The following are some of the issues they are currently trying to address:

Empowerment The committee seeks to achieve empowerment through greater gender equity
in representation on various committees. In 2003 there are plans to sensitise women workers on
the importance of having female representatives on the various committees throughout the farm.

Pregnant Women Currently this farm gives three months paid maternity leave, one month in
lieu of annual leave. Women workers can take maternity leave as often as they need, but the aim
of the gender policy is to encourage women to space their pregnancies for health and economic
reasons. Pregnant women are given special treatment in their seventh month of pregnancy such
as the provision of light duties and no work in the cold room.

Mothers - Mothers with small babies (less than four months) work for a maximum of 8 hours a
day.

Health Education The Gender Policy commits to providing education on HIV/AIDS, family
planning, nutrition and hygiene. The farm has also trained peer counsellors for HIV/AIDS.

Domestic violence and sexual harassment The company has contracted a local NGO the
Coalition on Violence against Women (COVAW) to provide counselling and training on gender
rights.

Proposed crche The gender committee is currently working to develop a crche through the
funds from the Max Havelaar social premium.

4.4 Composition of the Worker Sample

Demographic Characteristics
Over sixty percent of the study sample is female, a proportion that is typical of the industry as a
whole but significantly greater than the proportion of women engaged in waged employment in
Kenya (12 percent in 2000) (World Bank 2002). In large part the female dominance in the industry
stems from a rigid division of labour in which female tasks constitute the majority of employment.
Women are concentrated in the segments of the production process that are most labour intensive
and that also hold the most significance for the cosmetic quality of the final product. These include
picking, packing, and value-added processing activities, all of which require intense concentration and
long periods of standing and bending (Figure 2). Men on the other hand are primarily engaged in
spraying, irrigation, construction, packing in the cold store and maintenance, which can carry
significant physical risk.

28
The majority of sampled workers are migrants: 70 percent and 75 percent of all men and women
respectively have migrated from other areas of the country (Nakuru, Kiambu or Thika Districts),
where they were engaged in smallholder agriculture and/or informal activities prior to entering
employment in cut flowers. 19 The industry is also highly dependent on young workers, who are
perceived by employers as better suited for quick, labour intensive work. Over eighty percent of
sampled workers are between 20-34 years of age, 38 percent of whom are between 25-29 years of
age. When disaggregated by gender, more than half the female employees are aged between 20-24
years of age (Figure 3). While there is more variation in the age distribution of permanent workers,
casual and seasonal workers are predominantly between 20-29 years (Figure 4), suggesting that
young workers may be more likely to rearrange their lives to fulfil the flexibility that companies seek.
However, in contrast to some other global export industries, the profile of the workforce is not the
young single woman, as workers are more likely to be married than not (52 percent of women, 72
percent of men).

Figure 2: Gender by Nature of Work

35

30

25

20

(%) Employees
Female
15 Male

10

0
CDU GH M&C PH
CDU=Crop Development Unit GH= Greenhouse M&C= Maintenance & Construction PH=Packhouse

19
Most workers sought employment in the industry due to high levels of unemployment (in 2002 the national
unemployment rate was 14.6%) and/or land shortages in rural areas.
29
 Figure 3: Employment Status by Age

25

20

15

(%) Employees Permanent

Seasonal
10
Casual

0
<20 20-24 25-29 30-34 35-39 40-44
Age group of respondents in years

Figure 4: Gender by Age

25

20

15

(%) Employees
Female
Male
10

0
<20 20-24 25-29 30-34 35-39 40-44

Age group of respondent in years

30
 Table 8: Highest level of educational achievement by Gender
Gender
Level of Education Male Female
Primary Education 53.1 38.9
Secondary Education 40.6 50.0
Other 6.3 11.1
Total (%) 100.0 100.0
Source: Research Data, 2002

Employment Status
As noted earlier, the ILO conventions reflected in labour codes and Kenyan labour law recognise and
thus favour certain types of workers, in particular benefiting workers on permanent and seasonal
contracts over casual workers on short term contracts. All but one of the companies referred to a
change in labour strategy within the last year or so, from the use of low-skilled workers on repeated
short-term contracts with minimal benefits to a more stable workforce on permanent and seasonal
contracts. This change has been driven by a combination of factors including: pressures exerted by
civil society organisations and code provisions that discourage the use of temporary labour; the
quality imperatives of the European market that necessitate a somewhat more skilled, trained
workforce to perform value added processing (e.g., bunching, sleeving, labelling and mixed
bouquets); and year round production with less seasonality in production cycles.

This is reflected in Table 9 below, which highlights the fact that less than two percent of labour on
the seven farms is classified as 'casual' (according to the companies' own definitions). However, the
ratio of permanent labour to seasonal labour varies between farms, from 9 percent permanent versus
91 percent seasonal, to 81 percent permanent versus 15 percent seasonal. While this may reflect the
fact that some farms experience greater fluctuations in labour requirements (due to less consistent
orders, for example), it also reflects differing attitudes towards the role that labour stability plays in
overall company competitiveness.

Table 10 shows the proportion of men and women engaged in permanent, seasonal and casual work.
Women comprise 61 percent of our sample workforce, of which nearly 66 percent are on permanent
contracts. While the latter figure is skewed slightly upward by the number of permanent workers
employed on Farm 2, it is nevertheless a significant departure from the traditional dependence of the
industry on employing women on a casual and seasonal basis. 20 As Table 10 highlights, there is very
little difference in the proportion of men and women found in informal categories of work. Again,
this may not necessarily be reflective of the industry in general, given that the farms in our sample

20
Farm 2 moved toward a labour strategy based on permanent employment in 2002. The motivation for doing
so was both to reduce overall labour costs through a reduction in overall employment and enhanced productivity
gains gleaned from permanent workers, as well as to address the increased quality imperatives of overseas
buyers.
31
may be considered better practice companies which are making deliberate efforts to reduce their use
of informal labour.

Table 9: Percentage of Permanent, Seasonal and Casual Labour, Disaggregated By Gender

Total No Permanent Seasonal Casual Total %
M% F% M% F% M% F%
Farm 1 201 7.5 1.5 52.2 38.8 0.0 0.0 100.0
Farm 2 844 39.0 57.2 3.0 0.8 0.0 0.0 100.0
Farm 3 250 3.2 25.2 12.8 58.8 0.0 0.0 100.0
Farm 4 332 21.1 24.7 22.3 31.9 0.0 0.0 100.0
Farm 5 370 33.8 47.6 5.7 9.7 1.4 1.9 100.0
Farm 6 177 23.2 27.1 16.9 32.8 0.0 0.0 100.0
Farm 7 200 10.0 40.0 10.0 30.0 5.0 5.0 100.0
All Farms 2374 25.6 39.4 12.9 20.7 0.6 0.7 100.0
Farm 1: seasonal = 3 mth contracts, often rolling; casual = hired and paid daily
Farm 2: seasonal = 1 wk to 3 mths contracts, 1-3 contracts per year; casual = hired and paid daily
Farm 3: seasonal = 6 mths contract with 10-15 day break btwn contracts; casual = hired and paid daily
Farm 4: seasonal = 1 or 6 mths contracts, often rolling; casual = hired and paid daily
Farm 5: seasonal = 3-8 mths contracts; casual = 2 wks to 3 mths contracts, paid daily
Farm 6: seasonal = 3 mths then 6 mths contract, then may be made perm; casual = hired and paid daily
Farm 7: seasonal = 3-8 mths contracts; casual = 3 mths contracts, paid fortnightly
Source: Research Data, 2002, Company Interviews

Table 10: Percentage of Male and Female Labour, Disaggregated by Employment Status
Men Women
Perm Season Casual Total Perm Season Casual Total
Farm 1 12.5 87.5 0.0 100.0 3.7 96.3 0.0 100.0
Farm 2 92.9 7.1 0.0 100.0 98.6 1.4 0.0 100.0
Farm 3 20.0 80.0 0.0 100.0 30.0 70.0 0.0 100.0
Farm 4 48.6 51.4 0.0 100.0 43.6 56.4 0.0 100.0
Farm 5 82.8 13.9 3.3 100.0 80.4 16.4 3.2 100.0
Farm 6 57.7 42.3 0.0 100.0 45.3 54.7 0.0 100.0
Farm 7 40.0 40.0 20.0 100.0 53.3 40.0 6.7 100.0
Total 65.4 33.0 1.6 100.0 64.8 34.1 1.2 100.0
Source: Research Data, 2002, Company Interviews

32
5. Employment Conditions on Cut Flower Farms

The research revealed a pronounced gap between managements understanding of the nature and
severity of problems that workers faced and what workers reported, and similarly between
managements and workers understanding of company policy. Despite the relatively widespread
adoption of codes and the explicit commitment of several companies to improve working conditions,
none of the companies interviewed has a formal procedure for informing workers about codes. As a
result, workers had very little awareness of codes (4 percent of workers understood what a code of
conduct was), or how they might be used to improve their conditions.

This discrepancy is highlighted in the remainder of this section, where we summarise the views of
workers and management regarding employment in relation to social codes. For comprehensiveness,
we have drawn on the key social provisions from the KFC, MPS and ICC codes as a framework for this
analysis. This is not meant as an endorsement of any one code but rather reflects a sample of codes
applied in the sector.

5.1 Security of Employment: Regular Employment

Provided
Many codes stipulate that, as far as possible, producers should provide workers with regular
employment, and caution against the use of fixed term contracts in order to avoid legal obligations to
workers. Workers did confirm a trend towards permanent employment (as reported by management)
in the industry stating that previously workers could be employed on the same farm for 2-5 years
before being awarded permanent status. However, seasonal and casual workers (who represent over
thirty percent of sampled workers) expressed considerable feelings of job insecurity. They are most
exposed to potential dismissal (e.g., in the case of market decline or presumed poor performance),
often without mandatory warnings. In several instances, seasonal and casual workers said they
worried about becoming pregnant, sick or injured because they risked losing their jobs. This
insecurity had specific gender implications related in some cases to women's exclusion from benefits
such as maternity leave and sick pay, as well as the fact that many workers leave their children
behind in rural areas due to job insecurity. As one migrant woman said, If one is sacked and the
children are here and some are in school, where will I take them if I am dismissed? So it is better for
my children to stay at home so that if I am sacked, I will get problems alone.

Feelings of job insecurity were not unique to seasonal or casual workers; on some farms even
permanent workers perceived their jobs as insecure due to their mistrust of management, and lack of
adequate trade union protection from unfair dismissal. For example, on one farm half of the women
in two all-female FGDs claimed that although their contracts stipulated that two warnings be given

33
prior to dismissal, they did not trust the company as the company had not honoured their
contractual commitment to working hours. Moreover, in many cases workers feelings of insecurity
stem from a lack of clarity regarding their employment status. This is sometimes warranted as
companies interpret casual, seasonal and temporary labour variously. Some companies apply the
term seasonal for contracts that extend from two weeks up to three months, others use it for
contracts of one or six months, and still others for contracts of between three and eight months (see
Table 9 above). To confuse matters further, some use the term casual to refer to contracts of up to
three months, whereas others only refer to casual workers who are paid daily for specific tasks. Yet it
is not so much the length of the seasonal or casual contract that is important, as whether workers
become locked into seasonal/casual employment permanently. For example, if a seasonal worker is
given repeated contracts such that they are effectively working 12 months a year (as has been the
case), it is typically a strategy to avoid paying certain benefits or to acquire more flexibility in labour.
In this respect, three farms referred specifically to the use of rolling or repeated contracts with all
three having high proportions of seasonal workers. While this situation does not contravene the
industry CBA or national law, it is cautioned against in most codes covering the sector.

Box 2: Links between Job Insecurity and Migration

Job insecurity is exacerbated by the prevalence of a migrant labour force. Without family and kin locally
workers who are laid off have no money to pay the rent and often lose their houses. They lamented I
have no brother or cousin here as I am a migrant, whom can I stay with when chased away by the
landlord?.

5.2 Employment is Freely Chosen

All codes address the right to freely enter into employment. Neither forced nor slave labour was
found in any of the companies studied, nor was there any indication that workers were required to
deposit their identification papers with their employer.

5.3 Contracts
Most codes stipulate that workers be given a written, legally binding employment contract signed by
both parties detailing their obligations, rights and statutory deductions. Such contracts are indicative
of freely chosen employment. According to management interviews on all farms, permanent and
seasonal employees have written contracts, with all but one company providing a copy for the
worker. Casual workers are given verbal contracts only. However, workers reported that they were
not always provided with a copy of the contract, nor did they always understand its content since it
was generally written in English. The coverage of contracts also varies widely between farms

34
ranging from the provision of the most basic information to a more detailed account of rights and
obligations. 21

5.4 Working Hours are not Excessive

According to Kenyan law, regular working hours are 46 hours of work per week, with a paid day of
rest after every six days worked. However, on all farms overtime is generally compulsory and is
particularly common at certain times of the year, such as in the run-up to Christmas, Valentines Day
and Mother's Day, when packhouse workers are under great pressure to pack flowers before the
nightly flight to Europe departs. 22 Packers and pickers on three out of four farms complained of
working long hours, up to 16-18 hours a day (e.g., 7:00 a.m. to 11:00 p.m.), and sprayers on half of
the farms claimed to be working 11-12 hours, spraying from 7.00 a.m. to 6.00 p.m. One employer
acknowledged that graders were working more than 12 hours overtime a week, a situation which
they are trying to ameliorate. More female workers interviewed were involved in overtime work (since
they formed the majority of packhouse staff), which can have serious implications as they typically
also bear the responsibility for work in the home (childcare and domestic tasks).

In and of itself, overtime is not necessarily problematic if: it is not excessive; workers are
compensated appropriately; and it is chosen freely. However, participants in several FGDs
complained that overtime is not voluntary and that they are not given the opportunity to refuse in
accordance with codes. The responses also indicate that there is no awareness of how many hours
per week are allowable under a code or the law, pointing to communication issues between
management and workers.

Such long hours are partly a function of the seasonality of the industry. However, they are also a
function of the production imperatives that filter through the supply chain from overseas buyers, who
demand that exporters respond swiftly to fluctuations in consumer demand. Hence, from an
employers' perspective, providing workers with advance notice of overtime is not always possible, as
the use of 'just in time' supply chain management means that orders can arrive suddenly, often
during the same day that they are expected to be airfreighted. Compulsory overtime at short notice is
therefore not the producers responsibility alone. The ability to uphold these standards is directly
related to actors further up the supply chain.

21
On some farms the content is restricted to salary, duration of contract, and number of leave days while on
others contracts are more comprehensive, including duties, working hours, probation period, termination
conditions, health and safety, and details of benefits to which workers are entitled such as annual leave, sick
leave, maternity leave, medical coverage, and funeral expenses etc.
22
This is reflected in the CBA that stipulates every employee is required to work overtime when called upon to
do so.
35
All companies reported paying employees overtime at one and half times basic pay for normal days
and double basic pay for Sundays and public holidays, the exception being the three which pay
harvesters/graders/packers piece rate for overtime. These payments are issued in accordance with
the CBA and range between KSh.18.00 to 38.00 per hour depending on the basic wage.23 But
according to workers overtime payments are not always forthcoming. Over half of the respondents in
SSIs indicated that they are not paid overtime and in four of the FGDs overtime, particularly without
compensation, was ranked as one of the key worker concerns. In some cases workers claimed to
have been given days off in lieu of wage compensation, which is less satisfactory as the majority are
migrants who have come to the area solely to work. In general considerable confusion surrounds
overtime payments; most workers are not very conversant with the system of payments and do not
understand how they are calculated. This confusion is indicative of the extent to which company
policy is not properly communicated to the workers, particularly the use of piece rates for overtime.

Box 3: Consequences of Womens Overtime Work on Children

For those women whose children have accompanied them to flower producing areas, overtime
work has particularly negative consequences as they often face difficulties arranging for childcare
when they have to work late. For the most part female workers who live with their children are
single parents, and therefore the sole breadwinner and carer. They are very concerned about
having no-one to look after their children during the time they are at work, which very often
includes night hours. When women had to work until midnight, they could often go for days
without seeing their children. A female participant in a FGD aptly captured the impact of
extended overtime work, lamenting, Last year I had a small child and could even go back at
12.00 midnight and the childs health was really affected. There is a time I stayed for two days
without seeing my child because I was leaving very early in the morning and arriving home very
late at night.

5.5 Remuneration: Wages Must Be Sufficient to Meet

Basic Needs of Workers
With the exception of casual workers (who are paid daily), most workers are paid monthly with a mid-
month advance. The daily rate for a starting level general worker varies between KSh 80 and KSh
127 per day, with little variation between permanent, seasonal and casual employees (see Table 11).
It is also true, however, that the two farms with the most seasonals are also the farms that have the
lowest wages and award the least benefits. Two farms have a grading system that workers can be
promoted through, either with or without changing jobs, while others have set rates for specific jobs.
Non-promotional pay rises are given annually by two companies and on a discretionary basis by
another.

23
In August, 2002, one US dollar was equivalent to 78.5756 KShs.
36
 Table 11: Starting Wage by Employment Status
Farm Permanent Seasonal Casual
KSh/day KSh/day KSh/day
Farm 1 110 100 No data
Farm 2 127 127 127
Farm 3 103 103 No data
Farm 4 85 85 85
Farm 5 80 80 80
Farm 6 87 77-82 77
Farm 7 117 117 117
Average 101.29 98.79 97.20
Source: Research Data, 2002, Company Interviews

The normal weekly wage reported by workers varied widely from a minimum of KSh. 499.00 to a
maximum of KSh.2750.00. Table 12 sets out the weekly wages of male and female workers, and
shows that the wages for normal and higher earning males are slightly higher than those for women.

Table 12: Descriptive Statistics of Weekly Normal Wages and High Wages by Gender

Descriptive Statistics

95% Confidence
Std. Interval for Mean
N Mean Std. Error Minimum Maximum
Deviation
Lower Upper
Bound Bound

Female 54 1056.10 391.93 53.33 949.14 1163.09 569.00 2750.00

Normal
wage Male 32 1137.91 399.42 70.61 993.90 1281.91 499.00 2200.00

Total 86 1086.55 394.39 42.53 1001.99 1171.10 499.00 2750.00

Female 17 1197.24 433.76 105.20 974.21 1420.26 650.00 2095.00

High
wage Male 10 1392.00 541.90 171.36 1004.35 1779.65 696.00 2425.00

Total 27 1269.37 476.05 91.6154 1081.05 1457.69 650.00 2425.00

Source: Research Data, 2002, Workers Interviews

Codes also often include criteria requiring producers to pay living or minimum wages. Such criteria
do not refer to ILO conventions, but reflect concerns in some quarters that neither market forces nor
CBAs always ensure that a fair price is paid for labour. While KFC, FLP and ETI Base Codes stipulate
that wages be sufficient to meet basic needs of workers and their families and provide some
discretionary income, nearly all workers in the SSIs (85 percent of respondents) claimed that their
wages did not meet their basic needs (defined by workers as enough for a worker and his/her
dependents to feed, clothe and house themselves). In all the FGDS low wages were ranked as one of
the top four impediments to improving their situation and in three FGDs it was the number one
obstacle. The workers described how difficult it was for them to provide food, shelter, clothing and
school fees for their children, and how low wages forced them to rely on various income earning
strategies (e.g., receiving remittances from relatives, securing monthly credit facilities for household
37
items from the local shopkeeper, and borrowing from friends to supplement their wages). In fact, one
indication of the permanent female workers desperate need of cash in their hands was their
articulated perception, on two farms, that seasonals are better off than permanent workers because
they do not have deductions taken from their pay and can therefore afford more food. Their
immediate subsistence needs therefore overrode any potential future benefits derived from national
social security fund (NSSF) contributions, maternity leave or other benefits provided to permanent
workers. However, most workers felt that it was difficult to question wages as they were fully aware
(and were reminded by supervisors) that there was a surplus of unemployed workers in Kenya who
would be happy to replace them.

Box 4: Relationship Between Wages And Basic Needs

As a typical example of a workers response to whether their wage covered their basic needs, a
woman replied, "It is not enough at all. For housing I pay KSh 400, school fees are about KSh 500
per month, food about KSh 1500, water about KSh 200, clothing about KSh 600, and sickness which
varies and since the salary is about KSh 3000 per month, then I strain [to make ends meet]." This
was illustrated in a role play where workers enacted a room with a simple wooden bench (no other
furniture) and a man sleeping on the floor. Friends come by to borrow money to buy food, he has
none and they go to another worker who says he himself was coming to them to borrow money. The
first two then agree to go to others, while the sleeping man remains on the floor to sleep to avoid
feeling hungry.

However, while wages are low in the industry, they are also higher than statutory minimums and
generally exceed the wages that are available through alternative economic undertakings in domestic
industries or informal work. Companies also claimed that if they were obliged to raise wages, either
they would go out of business or Kenya could no longer compete with other countries where labour is
remunerated far less favourably. The issue of wage rates, therefore, highlights the importance of
government policy and regulation in establishing a fair minimum wage for the sector.

Finally, while most workers had a reasonable understanding of how their wages were calculated and
how deductions were made from their payslip, there were nevertheless a number of workers who
expressed confusion over the way their wages were calculated. Even those who received detailed
payslips did not always understand them, or did not think the calculations were correct. Once again,
these concerns show that workers are either not given sufficient information regarding their
entitlements or do not understand that which is provided.

5.6 Harsh and Inhumane Treatment and Intimidation

Many social codes prohibit harsh or inhumane treatment or intimidation in the workplace (including
verbal, psychological, physical and psychological abuse, and corruption). The clearest case of harsh
treatment relates to the relationship between supervisors and employees. Supervisor abuse was
ranked as one of the most significant problems facing both male and female workers in all 13 FGDs.

38
The poor relationships with supervisors is manifest in several ways including job harassment and
verbal abuse (experienced by both male and female workers), corruption and favouritism, dismissal
without just cause, and sexual harassment of female workers by male supervisors.

Non-permanent workers were particularly vulnerable to this abuse, as supervisors were often
responsible for hiring them, and workers feared not being rehired, or being fired, if they complained
or resisted. As one worker reflected, temperature inaweza panda, akasirike na akufute kazi (the
supervisor may simply lose his temper and get very angry and dismiss a worker). Incidences of
24
supervisor abuse were reported on all farms. However, workers also acknowledged that supervisors
were merely transferring the pressures from management onto workers such that the supervisor is
pushed by the management who then pushes the worker to do more work.

Sexual Harassment
One aspect of abusive treatment is the sexual harassment by male supervisors reported on all farms,
in both packhouses and greenhouses. Respondents claimed that supervisors request sexual favours
in exchange for employment, time off, promotion, and bonuses. As one respondent said, supervisor
akiwa na nafasi ya kazi ya wamama wawili, anawambia angewataka na basi anaweza kuwaanjili (a
supervisor can easily seduce women if he has a job for them). Resistance to such advances can
place women at considerable risk. One picker recounted how she has had difficulties with her male
supervisor ever since she refused his sexual advances and now lives in fear of dismissal. Workers
feel that they cannot complain because there is no suitable channel through which to communicate
such incidents to management.

Box 5: Sexual Harassment

In some FGDs male workers claimed that women would offer sexual favours to reap benefits such as
promotion or lighter work, particularly in the grading shed. However, the same group noted that
supervisors can sack female workers who refuse their advances in order to employ women that will
provide them with sexual favours. Alternatively, supervisors can deploy their power as supervisor to
compel a woman to comply. One respondent reflected, When a male supervisor seduces a female
worker and this doesnt bear fruit, he can use thorax [i.e. job power] to win that female worker.

The majority of management staff interviewed didn't think that sexual harassment occurred on their
farms. Even where it was acknowledged, sexual harassment was viewed as a cultural rather than
company issue. As one employer claimed, there is a problem between male supervisors and female
workers women are treated badly. This is largely a socio-cultural phenomenon women have

24
For example, on one farm, there were allegations of supervisors sacking workers and informing the
management that workers had deserted duty. Bribery was also reported by some men who claimed that
supervisors will help a worker to obtain employment on the condition that he receives a portion of the workers
first pay check. Some supervisors also borrow money from the workers (particularly those that they have had a
hand in employing or promoting) and delay paying it back.
39
traditionally been abused by men.25 On a positive note, on three farms several workers felt that
management was making efforts towards confronting sexual harassment.

5.7 Equity in Wages and Non Discrimination

Hiring Practices
Only one of the companies in the sample had a written Equal Opportunities Policy, claiming that since
they have more female than male staff they practice equal opportunities. However, as noted earlier,
there are clear gender divisions of labour, derived from both management preferences (e.g., women
are more dexterous, which is good for flowers) and socio-cultural traditions: women tend to work as
harvesters, graders and packers, whereas men tend to be employed as spraymen, irrigators, scouts,
maintenance workers, security men, and greenhouse workers. In addition, management (including
supervisors) is predominately male on most farms, with the exception of the lower level team
leaders, among whom the gender proportions within the workforce are better reflected. Even among
workers, such gender segmentation was viewed as a natural outcome of social norms and biological
difference. According to one Kenyan greenhouse worker, In African culture, women tend to respect
men leaders more than women leaders and this is a weakness women have, it is a weakness that
women are born with.

Training
Insofar as unequal treatment is attributable to different skill levels, training is important to overcome
discrimination. Training and/or performance-related pay were identified by most companies as key to
improving productivity, quality and efficiency. One farm has introduced a system of 'integration' with
multi-skilled workers responsible for a whole section of production (e.g., crop husbandry, bed
maintenance, harvesting etc.). This not only allows workers to earn more and do less repetitive
work, but it has also improved labour productivity and efficiency on the farm.

However, interviews with workers indicate that most receive very little training and/or opportunities
to improve their skills beyond that which is provided by supervisors or peers upon employment. In
one FGD female packhouse workers said that they would like to have more training in the packaging
of flowers in order to master new techniques and avoid the possibility that they would face
deductions from their wages or be dismissed because of damage to flowers. In addition, while there
is little gender bias in training opportunities, the gender segregation of tasks functions as a barrier to
the attainment of training as the jobs that women perform have fewer opportunities for skills
upgrading. For example, formal training is provided to spraymen, storemen, irrigators, scouts and

25
However, one interviewee admitted that sexual harassment had been disclosed on another of the companys
farms and of the three perpetrators, only one was fired, since the others were too valuable to the company.
40
health and safety officers (e.g., fire wardens), positions that are typically occupied by men and
associated with higher wage levels. 26

Promotion
Promotion for most of the workers is synonymous with being upgraded to supervisory positions or
being made a permanent worker. According to most employers promotion is based exclusively on
merit, although one company bases it on both merit and length of service. Promotion in our sample
was relatively uncommon (only 16 percent of the respondents stated that they had been promoted).
This reflects the nature of our sample, which was primarily comprised of general workers, as well as
the nature of work in the industry (primarily low skilled). Despite its infrequency, few workers
identified any gender-specific barriers to promotion. More than half of the workers in the SSIs said
that companies provide equal opportunities for promotion based on workers productivity and job
performance. However, as noted, formal training, which can lead to promotion, is usually restricted
to jobs that are predominantly performed by men.

However, not all workers thought that promotions were based solely on merit. Several participants in
the FGDs claimed that supervisors and junior managers based promotional decisions on nepotism and
bribery (sexual or financial). The workers also resented the fact that management sometimes
recruits supervisors from people who have worked in other companies before them, even if they have
been on the farm for a shorter period of time. On half of the farms male workers in FGDs complained
that men are brought in from outside for senior positions, thus discriminating against local workers.
Finally, although workers did not identify gender discrimination in promotion, pregnancy testing prior
to employment of permanent workers was reported to occur on one farm as part of the general
medical examination (administered to both men and women). Workers claimed that there are some
seasonal workers who were slated to be upgraded to permanent status, but were not after failing the
medical examination, allegedly due to pregnancy.

Equal Pay for Work of Equal Value

All farms formally practised equal pay for equal work, yet the gender segregation in job categories
and skill levels creates male-female wage differentials in both production and processing. Pay scales
are thus different for predominantly male or female jobs, with men earning more on average than
women (see Table 12). This is often due to an allowance being paid to sprayers or coldstore workers
(male positions) for working in what are seen as difficult conditions 27. Standing for long hours

26
Formal training is administered by organisations such as the Agricultural Society of Kenya (ASK), and the Safe
Use Project, and is provided mainly for workers in the CDU (predominantly men), who are taught how to handle
chemicals, use PPE, and maintain inventories of chemicals. One farm trains members of the Workers' Committee
(predominantly men) on disciplinary procedures, grievance handling, and statutory and company regulations.

27
On one farm sprayers receive an allowance of KSh 20 per day whereas on another sprayers are given a
monthly allowance of KSh 550.00
41
without breaks, performing repetitive tasks typical of the work carried out by women is not
considered by employers as difficult.

5.8 Freedom of Association, Collective Bargaining and

Worker Participation in Decision Making

Trade Unions
Freedom of Association and the Right to Collective Bargaining are two of the fundamental principles
of the ILO core conventions, and are part of many codes. Furthermore, according to the Kenya Law
of Association, workers are protected and are allowed to join and form unions of their choice,
although there are limits on freedom of association and the government has yet to ratify ILO
Convention 87.

Unionisation of employees is a fairly recent phenomenon (most respondents who are members joined
in 2002) and just over thirty percent of all respondents are union members (see Table 13). Four of
the seven farms have employees who are members of the Kenya Plantation and Agriculture Workers
Union (KPAWU). Only permanent workers are members, and the number unionised varies from only
a few workers up to 90 percent of the permanent workforce (see Table 13). Only two farms actually
have sufficient membership to elect shop stewards (i.e. at least 51 percent), there being three female
and two male shop stewards on one farm, and six female and six male shop stewards on the other.
According to company interviews, many workers who had joined the union prior to 2002 left when
the membership fee increased in 2002. 28 Several companies also claimed that the union is not
proactive in its work, citing instances where the unions had been invited to speak to workers, but
failed to show up.

Since workers have only recently joined the union they have yet to reap all of the potential
advantages of membership. However, there are conflicting opinions about the effectiveness of trade
unions in representing workers. Approximately 20 percent of workers were satisfied with union
efforts. They cited a number of changes that were spearheaded by the union including the
negotiation of salary and leave, transport allowance increments, observance of re-entry periods after
spraying, reductions in supervisor abuse of power, and the warning system. They indicated that the
frequency with which verbal and/or written warnings were issued had also decreased, and that the
union had also begun representing dismissal cases.

28
Workers now pay two percent of their basic pay to KPAWU, plus KSh 50 per month to the Central Organisation
of Trade Unions (COTU).
42
 Table 13: Union Membership
Percentage of Workers Unionised Number of Unionised Presence of a Workers
(acc. to Management) Respondents Committees
Farm 1 Less than 50% of permanents 0 yes
Farm 2 45% of permanents 26 yes
Farm 3 None 1 no
Farm 4 None 0 yes
Farm 5 90% of permanents 8 yes
Farm 6 None n/a yes
Farm 7 90% of permanents n/a yes
Total 35
n/a = No worker interviews conducted on these farms
Source: Research Data, 2002, Company and Worker Interviews

On the other hand, a larger number of workers were less sanguine, claiming that unions were
inactive in recruiting members and failed to adequately represent the workers who did join. Several
said that they dont know where the union they registered with went to and it is as if the union
wanted just money from them. Other workers attributed low figures of union membership to
management intimidation, with 19 workers (over half of whom were from one farm) saying it was not
allowed by management, or believing they would be sacked for joining. Workers in one FGD
performed a role play that depicted a local farm (not a flower farm), where the entire unionized
workforce was sacked and replaced with new workers, without the union taking action to reinstate or
seek compensation for the employees who were terminated.

While the trade union technically extends membership to non-permanent workers, union recruitment
efforts are limited to permanent workers. In this study none of the non-permanent workers were
union members. Likewise, the CBA restricts the provision of most benefits to permanent workers.
Given the numbers of workers remaining in non-permanent work, this exclusion is cause for concern.
Indeed, workers in three FGDs expressed concern that casual workers were not entitled to benefits
such as maternity leave and felt that the trade union should extend coverage to casual workers.

Workers Committees
Codes such as KFC, ICC, and Max Havelaar stipulate that workers should be free to form their own in-
house welfare associations and elect their own leaders with full mandate to negotiate with the
employer on their behalf. All but one farm has such an association in the form of a Workers
Committee (or a Welfare Committee). Representatives are normally elected by the workforce from
different sections of the company, with men slightly better represented than women (only two farms
specify the election of equal numbers of men and women) (see Table 14). On one farm a Welfare
Committee manages a co-operative fund to assist workers in financial need (mainly for funeral
expenses), and the Max Havelaar Committee manages the social development fund that is derived
from exporting to the Fair Trade market. Moreover, this is the same farm that has also set up a
Gender Committee, as described earlier (Box 1).

43
Box 6: Reasons for under-representation of women in Workers' Committees
Both men and women workers in Kenya expressed the opinion that there were more men than
women elected to committees because men made better leaders. For example, one woman said,
"women are scared of asking questions and therefore not good representatives, and another said,
"Men champion better the workers rights as they do not fear questioning." This lack of confidence in
women's abilities, rooted in cultural beliefs, demonstrates the importance of giving women support to
voice their opinions. This can be done through having predetermined seats for men and women in
committees, proportionate to their number in the workforce, or through having separate Women's
Committees (Smith et al. 2003).

Table 14: Gender Representation on Workers Committees

Men Women
Farm 1 10 10
Farm 2 5 3
Farm 3 n/a n/a
Farm 4 70% 30%
Farm 5 mostly men on other Have a separate
committees Gender Committee
Farm 6 6 6
Farm 7 No information No information
Source: Research Data, 2002, Company Interviews

Generally, worker committees are seen as more effective in addressing worker concerns than the
trade unions. This is partly due to the vacuum in trade union representation, but also due to the
quality of the representation where it does exist. About thirty-three percent of all respondents
claimed that workers committees had succeeded in making several positive changes in working
conditions such as a shift from seasonal to more permanent labour; salary and house allowance
increments; and provision of PPE. Workers also believed that the mediating role played by
committees had helped to mitigate supervisor harassment and increase dialogue between the
management and the workers. However, the majority of workers (66%) still found the committees
ineffective in representing their interests, claiming committees to either be inactive and/or lax at
presenting workers grievances to management. They cited cases where workers elected dynamic
committee members to represent them but that these committee members were quickly silenced by
management through promotion to supervisory positions.

Box 7: When Workers Committees Fail

One role-play enacted the ineffectiveness of workers committee in championing the workers
grievances. When the women went to talk with the committee representatives they were very keen to
listen to them and had very cordial discussions. However when the committee representatives
brought workers concerns to management, they were inarticulate and sometimes reported the
contrary just to please the managers.

44
5.9 Grievance Procedures
Some codes stipulate that workers should have the right to be heard on matters relating to
contractual terms, dismissals and general welfare, but this can prove difficult due to the hierarchical
nature of company structure. Complaints from workers are typically received through line
management and committees on most farms, though it varies in terms of whether there is a formal
documente d procedure or not. Two employers said that senior management has an open door policy
if workers prefer to go directly to them, while another said that since they are a small company it is
easy for staff to communicate freely. On two farms there are Complaints/Suggestions Boxes, with
one company providing a KSh. 1,000 bonus for the best idea for improved management. All of these
are predicated on the assumption that workers, the majority of whom are women, will feel sufficiently
secure to avail of these options. Even employers acknowledged that communication was often poor,
a situation that was reflected in the fact that senior management was clearly unaware of many of the
problems that employees faced. One employer said: Our main difficulty is communication. Workers
have preconceived notions about management, and we need to improve communications to help with
this. Another employer remarked, There are no complaints channels or forums and I admit to
existing job insecurity felt by the workers. The KFC auditor has actually requested that these farms
establish a formal grievance procedure and there are some indications that companies are trying to
improve communication between management and workers.

Issues of poor communication and lack of effective representation were echoed by general workers
and few, if any, had ever interacted with senior management. In four farms the majority of
respondents (85 percent) said that company policies are not properly explained to them but rather
are issued as directives from supervisors. Workers also felt that supervisors do not forward their
grievances to management as they do not want to be perceived as pro-worker by the management.
There are also salient gender implications to this structure as women must communicate their
problems to predominantly male managers before reaching management. This decreases the
probability that women will share any sensitive information such as sexual harassment, health
concerns or childcare problems, and at the same time leaves management unaware that there may
be problems among their workforce.

5.10 Health and Safety Provision of a Safe and

Healthy Work Environment
All codes require that companies provide a safe and hygienic working environment. This study found
that codes have engendered notable improvements in occupational health and safety, specifically
concerning the safe use of chemicals and provision of PPE, as well as separate toilets and washing

45
facilities for male and female employees. 29 According to interviews with employers, three companies
provide all workers with general health and safety training every one or two years, while the
remaining companies only train workers regarding special risks, such as those handling chemicals.
Most of the farms have also carried out, or been included in, educational campaigns to raise
awareness of HIV/AIDS among workers (two companies also mentioned handing out free condoms to
workers), and the KFC code contains particular provisions on HIV/AIDS awareness.

However, serious issues persist. In all FGDs women raised concerns related to pregnancy, exposure
to chemicals, and extended periods of bending and standing, while men raised the issue of chemical
exposure and inadequate protective clothing. Fifty percent of the respondents in the SSIs indicated
that the provision of PPE is inadequate not all workers receive it and even if they do, it is not
replaced regularly. Participants in six FGDs listed this as one of the key problems they faced at work.
Casual workers, in particular, face greater health risks since they are commonly not provided with
PPE. Graders on the farms raised the issue of working with either worn or inadequate gloves to
protect their hands from thorns and hand injuries were widely reported. In response, several
employers argued that it is difficult to convince workers to abide by health and safety regulations,
especially the wearing of protective clothing. For example, spraymen complain that clothing is too
hot or uncomfortable, and flower graders elect not to wear gloves as they cannot work as quickly
with them on, which affects their remuneration through piece rate bonuses. As a result of such
problems, one company introduced a temporary bonus scheme to reward those who wear their
protective clothing.

Workers in FGDs on three farms cited chemical exposure as one of the main problems affecting both
men and women, giving rise to a number of skin, throat and respiratory health problems. The
workers said that while notices regarding re-entry periods were posted, re-entry periods were not
strictly adhered to, and spraying and harvesting were reported to occur simultaneously. The sprayers
in these FGDs complained that their spray suits are usually worn out exposing them to chemicals.
They also claimed that although they undergo cholinesterase testing regularly, none of them are
transferred to other sections regardless of the results of the tests. According to workers the
mandatory job rotation for the sprayers is not observed, resulting in long periods of exposure to
chemicals. These problems are often more severe for workers with less education, who face
difficulties with detailed spraying instructions. Even women (who are not permitted to handle
chemicals on farms in accordance with company policy), can be affected by chemicals in cases where
re-entry periods are not observed. In particular, there are risks associated with the effects of
pesticides on unborn babies and breast feeding children. According to workers in one FGD, one

29
Provisions mentioned by some or all companies include: health and safety committees; protective clothing for
all workers, or alternatively just for those handling chemicals; cholinesterase testing (from random testing every
month and base-line testing of new employees, to annual testing of those handling chemicals); alternation of
sprayers and irrigators; light duties for pregnant women; first aid training and equipment; fire safety precautions
and training.

46
woman miscarried due to the chemicals in the greenhouse and another mother said that her baby
refused to breastfeed when she returned to work because the mothers milk is smelling of
chemicals.

Box 8: Exposure to Chemicals

In FGDs sprayers were very vocal about their difficulties regarding chemicals such as inadequate
protective clothing, long hours spent spraying and the potential danger to their health. In one all
male role play, workers depicted a supervisor who forced workers to continue spraying after their
original job was completed. They also only had one change of protective clothing and one bar of
soap per month to wash them. However, this is also an illustration of where workers are ill informed
of company policy as the CBA only requires that employees be issued with one standard bar soap per
month.

5.11 Pregnancy
Most codes covering the sector stipulate that women shall not be discriminated against during ante-
or post-natal periods in regard to recruitment, selection or termination. Expectant mothers should
also be given due consideration and assigned to duties appropriate to pregnancy. These light duties
typically consist of sweeping, picking up litter, and cleaning. However, one farm manager said that
she makes sure that women go on leave one month before their due date since it is difficult to find
light enough duties for heavily pregnant women. In addition, respondents claimed that supervisors
often deride light duties, claiming that, mimba iko kwa tumbo si kwa mikono (pregnancy is not in
the hands but in the stomach), and that pregnant women should carry on as normal.

The analysis of the FGDs indicates that special treatment for pregnant women and mothers with
babies depends on a number of factors such as the specific farm and/or section in which one works
and his/her employment status. In FGDs participants from three farms said that pregnant women are
given lighter work but only those who work in the packhouse. In the harvesting section, women who
are pregnant work up to the last day without a reduction in workload. According to one worker,
When I was pregnant, I was not treated differently and used to go home very tired and the company
said it will be changing pregnant women to do lighter work but this has not happened yet. In
addition, even on the better practice farms the provision of light duties is not extended to all pregnant
women as it is at the discretion of the supervisor, who may or may not sympathise with the
condition of the pregnant workers. However, given that light duties are required under codes, their
provision should not be contingent upon supervisor benevolence. This also highlights the need to
train supervisors on company policy and gender awareness.

47
 5.12 Beyond Codes: Other Worker Welfare Issues

Benefits
In hazardous, relatively low-paying sectors such as floriculture where demand for labour often
fluctuates, employment benefits other than fair wages are of great importance to workers. Benefits
such as maternity leave, childcare provision and transport may be especially important to women,
enabling them to balance unpaid caring work with paid work in the industry (Barrientos, Dolan and
Tallontire 2003).

One of the findings of this research was that many of the benefits workers value are not covered by
codes. For example, while all companies provide a range of fringe benefits to workers (e.g., food,
transport to and from work, and medical care), the extent to which codes address these provisions is
uneven. Furthermore, permanent workers are more likely than others to receive benefits such sick
pay, medical care and maternity leave. However, one of the positive changes in the industry is the
increasing access of non-permanent workers to benefits (Table 15). The following describes some of
the key benefits provided to workers in the industry.

Table 15: Wage Benefits Provided by Employment Status

Benefit Farm 1 Farm 2 Farm 3 Farm 4 Farm 5 Farm 6 Farm 7

Overtime 02 or 02 or 02 02 or 02 02 n/a
piece rate piece rate piece rate
Paid Annual 02 02 02 02 02 02 02
Leave after 6 mths after 2 mths
Paid Sick 02 02 02 02 02 02 01
Leave work related
Paid Maternity 01 02 02 01 02 01 01
leave 1+1 mth 1+1 mth 2 mths 3 mths 2+1 mths 1+1 mth 1 + 1 mth
NSSF 01 + some 02 02 02 02 02 No info
seasonals
Private 01 01 or 02
pension Gratuity
Production 02 02 02 02 No info
bonus
Xmas bonus 02 No info
Other bonus Drivers, Spraymen/ Spraymen No info
supervisors Coldstore
Funeral 01 co-op 01
expenses
01 = permanents only; 02 = permanents and seasonals; 03 = all workers
Source: Research Data, 2002, Company Interviews

48
Wage Benefits
Annual Leave & Sick Leave
According to companies, annual and sick leave is provided to all permanent and seasonal workers.30
Yet there is some discrepancy between management and worker accounts. For example, only 33.3
percent of seasonal workers claimed to have received annual leave despite the fact that most had
contracts indicating their leave entitlements. Similarly, companies indicated that permanent and
seasonal workers are given paid sick leave provided that they present a doctor's certificate, however,
only 61.1 percent of the seasonals (and 33.3 percent of casuals) said they have access to this benefit.
On one farm sick leave is only given if the illness is work-related, as determined by the company's
contracted doctor, and there were some reports that the doctor did not believe workers when they
reported falling ill. Once again this may suggest that workers lack an understanding of the terms and
conditions of their employment.

Maternity Leave
Maternity leave is typically only extended to permanent workers, however, on three farms seasonal
workers are also provided with maternity benefits. According to both Kenyan law and the CBA
women are eligible for two months maternity leave, one of which is their annual leave. Most
Northern codes have stronger provisions, requiring a longer term of leave. However, most women
prefer to work up to full term and take leave after or on delivery. This was confirmed in FGDs, There
are some who work until they deliver the baby while at work. Most pregnant women hide their
pregnancy, so that they can have a longer maternity leave to spend with the baby.

Anxiety about becoming pregnant or informing their employers that they are pregnant is widespread,
particularly among those in non-permanent positions. According to one worker, when a woman
displays symptoms of pregnancy (e.g., vomiting), she is monitored closely and all eyes are focused on
her like a telescope (macho yote inawekwa kwako kama darubini). The majority of seasonal
workers and casual workers claimed that not only are they excluded from maternity leave
entitlements, but they are also more likely to lose their jobs in case of pregnancy. Workers claimed
that pregnant women in casual/seasonal positions are typically not given another contract when their
current one ends and are the first to be dismissed in case of a mchucho (when there is no work in
the company). As a result abortions are common. According to a group of casual female packhouse
workers, wanawake wengi sana hapa wanatoa mimba hata mtoto wa miezi nane ili asifutwe kazi
(many female workers have had abortions, even at eight months to avoid the risk of being sacked).

30
Companies provide between 1.75 and 2.5 of paid annual leave days per month and three farms provide a
travel allowance for workers to return home during leave. Sick leave is more variable. Five companies give 30
days on full pay and 30 days on half-pay per year (pro-rata for seasonal workers), one gives 40 days full pay and
45 days half-pay, and one gives up to 90 days full pay.
49
Non-Wage Benefits
Permanent workers were also more likely to have access to non-wage benefits than seasonal or
casual workers. Yet the decisive factor influencing access to non-wage benefits was company policy
(see Table 16). The following describes the non-wage benefits provided to workers.

Housing
In compliance with national law and the CBA, companies are required to provide employees with
housing or a housing allowance in lieu thereof. Housing is provided for some or all workers on three
farms,31 while all other companies pay seasonal and permanent employees a monthly housing
allowance equivalent to 15 to 30 percent of their basic wage. Most workers interviewed complained
that the housing allowance was insufficient to acquire decent housing, particularly for workers living
in close proximity to Nairobi (e.g., Kiambu and Thika Districts) where rents are higher. With an
average monthly house allowance of KSh. 800 workers can only afford single rooms without
electricity, water, and toilets despite the fact that many are shared with children and/or spouses.
Where housing is provided, workers usually live in one-room units. The lack of adequate housing is
one of the main reasons that children remain with their grandparents or other relatives in the rural
home districts.

Healthcare
According to the Employment Act, it is the responsibility of employers to ensure the provision of
appropriate medicines during illness. All companies provide employees with free primary healthcare,
in most cases through a resident company nurse, and three companies also have their own doctors.
Thus far, HIV/AIDS has not been a significant issue for companies, although they all recognised it as
a potential threat. On one farm the management said that they have an anti-discrimination policy
regarding HIV/AIDS sufferers, commenting that they have one employee who has AIDS and is still
productive. In the SSIs and FGDs workers had little awareness of HIV/AIDS as a problem on the
farms; those that were aware of the disease indicated that they had only heard about and not
experienced it first hand.

31
Two of the three provide housing for permanent workers only.
50
 Table 16: Non-Wage Benefits Provided to Permanent, Seasonal and Casual Workers

Benefit Farm 1 Farm 2 Farm 3 Farm 4 Farm 5 Farm 6 Farm 7

Housing/ 03 03 02 03 01 03 01
housing 15-20% HA KSh900 or 15% H 15% HA or KSh 870 HA 15% HA H
allowance after 6 mths HA house
Primary 03 03 ? 03 03 03 03
healthcare nurse nurse clinic nurse clinic nurse see below
Secondary 01 03 03 03 03 03 03
Healthcare Up to KSh Comp. doctor Comp. doctor Comp doctor KSh subsided up to 5%
10,000/yr once/wk 1,000/mth doctor salary
Hospital Care 03 03 03 03 01 03 03
NHIF NHIF NHIF NHIF insurance NHIF see above
Creche 01 or 03 planned planned
Transport 03 03
Leave travel 01 01 03
allowance
Childrens 03 03
education community primary
support school
Food or food 03 03 03 KSh10/day uji for some 03
allowance uji canteen & uji uji
Sport facilities KSh 250,000 03 03 03 03
budget
Community Hall 03
Staff outings
Other Annual party community social events
work
01 = permanents only; 02 = permanents and seasonals; 03 = all workers; HA = house allowance; H
= housing provided; NHIF = National Hospital Insurance Fund
Source: Research Data, 2002, Company Interviews

Childcare and Education for Employees Children

Childcare facilities make a significant difference to womens ability to participate in the labour market.
Indeed, many women prioritised childcare ahead of wages, signalling the importance of womens
labour constraints, and the difficulties they face in taking proper care of their children while working
in the industry. A number of female workers (whose children had accompanied them) said they were
obliged to leave younger children with their older siblings or neighbours while they worked as they
could not afford to employ household help. Alternatively, children were left unsupervised and out of
school. Other workers were forced to leave their children behind in rural areas where they could be
assured that their kin could provide them with decent care.

On-site childcare facilities were thus highly valued by working mothers, but they were only provided
by one company (although two others reported plans to build a crche). Permanent and seasonal
employees of the same company can also send their children to the primary school on a neighbouring
farm. One other company provides support for the education of their employees' children, in the
form of paying the salary of two teachers at the local school, and paying for construction of new
classrooms.

51
Box 9: Migration and Childcare
In Kenya many women had migrated from rural areas to seek work, but had left their children to be
cared for by their grandparents, as they could not afford to keep them with them. For those
respondents who work and their children live away from them in rural areas it is an emotionally
difficult situation. They missed their children greatly and many expressed concern about their
upbringing, "I only pray to God to keep them safe. One worker said that, people here really miss
having joy with their family. But the work demands that they work here and children live at home due
to low wages.

Transport
Free transport to and from work is provided by two companies, with the remainder either having on-
farm housing, or employees that live close by. On one farm there is a bicycle purchase scheme that
is jointly financed by the social premium derived from the Max Havelaar scheme and the workers.
Where transport is provided, workers complained of problems with unreliability, overcrowding and
lateness in the evening. Several workers were also concerned about security claiming that if they
work late there is sometimes no transport provided or they are dropped too far from their homes.
Safety in travelling to and from the farms is especially problematic for female workers, especially
since women have been mugged on the way home in the dark.

52
 SECTION 3: THE AUDITING PROCESS

6. Participatory Social Auditing

6.1 Standard Auditing Practice

As discussed in Chapter 4, most codes applied in the cut flower industry have arisen in response to
civil society and consumer pressures in developed countries. As a result, they have been introduced
by Northern companies at the top of the supply chain, and imposed on Southern suppliers and
workers without a great deal of consultation. Even codes developed in the South (e.g., KFC, FPEAK,
and the Kenya National Code) tend to draw criteria, as well as legitimacy from modelling Northern
codes. For example, while four of the seven sample cut flower farms were involved in developing the
KFC code, none of the principles within the codes arose through workplace consultation.
Consequently, the actual priorities of flower workers in Kenya have been excluded from the code
development process.

The exclusion of workers perspectives in code design can be ameliorated by ensuring that workers
voices and concerns are included in the auditing process. However, in many cases codes have been
monitored by existing quality assurance staff and/or through a technical, check list approach, which
draws significantly on employer interviews, yet incorporates little worker consultation. Experience has
shown, however, that constraints of time, knowledge, and culture mean that Northern technologists
are not always suitable for monitoring employment conditions (Auret 2001). Even in cases where
independent auditing has occurred, the methods used have often been insufficient to capture
sensitive issues such as gender discrimination, sexual harassment and abuse, which are most
common amongst workers in non-permanent job categories (see Box 10). Many of these issues are
often deeply embedded in social norms, practices, and institutions, and are therefore not easily
picked up by parachuting monitors who make snapshot auditing assessments. While many Kenyan
flower farms are taking monitoring seriously (for example, three companies sent management staff
on a SA8000 training course given by BVQI in 2002), the depth of auditing is variable in terms of both
the methods used and the range of external stakeholders integrated into the process.

Box 10: How Reliable are Audits?

On one farm discussants in a focus group claimed that if a worker is allowed to talk to KFC auditors,
he/she is accompanied by the supervisor. One of the respondents wondered, inawezakanaje mtu
akuje akitaka kuniuliza siri ya mke wangu na amekuja na huyo bibi yangu?, ( i.e. how can one seek
to know secrets of my wife from me yet he comes accompanied by the same wife?).

53
6.2 Participatory Social Auditing32
The benefits that codes can bring to business are unlikely to be realised if the process of code
auditing is flawed. One way to achieve a deeper understanding of labour conditions is through the
adoption of a participatory social auditing33 approach, which can be particularly helpful in illuminating
gender issues (e.g., sexual harassment and sexual discrimination) that are difficult to discern through
conventional snapshot audits. While participatory methods have been developed and applied by
development practitioners for decades, it is only recently that these methods have been adapted to
private sector initiatives. Adopting a participatory approach in social auditing is more likely to build
trust and to promote dialogue, and expose workplace issues that are often impossible to identify
through a one-size-fits-all auditing approach. In particular, by using techniques and tools such as
group discussions and visual representations (see Box 11), the participatory approach has several
advantages for ensuring that the voices of marginalized workers are included, who are typically
excluded in a number of other institutions, such as trade unions, national legislation and international
conventions. Aurets work among Zimbabwean producers has found that a participatory social
auditing approach brings several benefits including:

facilitates consulting with, and gathering data from, all categories of management and workers of
varying age, gender, and race;

contributes to a process of awareness creation that can produce long-term change;

facilitates the involvement of the poorest and the powerless, particularly seasonal, casual,
contract and migrant workers, and those who are semi-literate or illiterate;

is flexible and can be adapted to different social and cultural situations, providing insights into
local differences; and

is cost effective. Although it may entail greater up-front costs in terms of employee time, and
requires special skill-sets among auditors, the depth and relevance of information acquired render
it cost effective in the long run (Auret 2001).

However, participatory social auditing is not a one-off activity; to be effective it must be undertaken
as a process, beginning with raising awareness among employers and employees about the principles
and criteria embodied in codes, followed by a series of pre-audits and audits, and ultimately leading
to remedial action in order to cultivate improved management/worker relations (Smith et al. 2003).

32
This section is based on Auret 2001.
33
In this report, the term social audit refers specifically to the auditing of labour standards, which includes
standards applicable to both working and living areas. Participatory social auditing refers to the techniques and
tools developed during the ETI Pilot Project in Zimbabwe and later used by the Agricultural Ethics Assurance
Association of Zimbabwe (AEAAZ). The theory and practice of participatory social auditing is discussed in detail
by Di Auret (2001), on which this section is based.

54
Box 11. Participatory Social Auditing Tools
Semi-structured one-to-one interviews (SSIs): In a social audit, SSIs should be based on
the social code, but have the flexibility to explore any relevant information raised by the
informant. SSIs should be conducted with key informants drawn from management and different
categories of workers, taking care to ensure proportional gender representation.
Focus group discussions (FGDs): FGDs are informal, conversational group interviews based
around the provisions contained in the social code. Ideally FGDs should include 6-8 participants,
selected from different managerial and worker categories. Generally, the groups should be
relatively homogenous (e.g., all seasonal pickers, packhouse workers, supervisors and single sex
groups) to capitalise on the identification that stems from shared, common experiences.
Group Exercises: Where time allows FGDs can include group exercises, during which the group
is asked to visually depict key issues raised in their discussions. Exercises include mapping
working and/or living areas, role-play and pictorial ranking of key issues.
Participant observation: A 'transect walk' conducted by auditors through the working and/or
living areas in the farm/packhouse can cross-check the verbal data collected. The inclusion of two
or three workers in this walk enables it to serve as a listening survey during which a deeper
understanding of attitudes and behaviour is acquired (ibid.).

6.3 Assessment of Participatory Social Auditing

Methods
One of the aims of the research was to explore the advantages and disadvantages of social auditing
methods, especially in terms of how to enhance gender sensitivity. A number of participatory tools
(including SSIs, FGDs, ranking exercises, role plays, and transect walks) were employed to assess
how best to identify gender-sensitive information and encourage workers to participate in workplace
improvement. The following describes the findings drawn from the data collected and feedback from
research assistants (or 'facilitators' in the FGDs).

SSIs
The SSIs, which consisted of both closed and open-ended questions, were a valuable tool for
obtaining information on demographic and employment-related data (e.g., contracts, benefits, and
working hours), as well as an overview of key issues of concern to workers. Overall, women workers
were more apprehensive of the interview process than their male counterparts and were less
forthright in their responses. Young women, particularly, required much more prompting to disclose
information that required more explanation than a yes/no. Nevertheless, while women were more
apprehensive than men during SSIs, they were also more likely to share their feelings and
perceptions of sensitive issues in this format.

SSIs are a useful tool for data collection; however, they have their limitations, particularly in their
capacity to provide a more comprehensive and in-depth assessment of working conditions. Firstly,
they are time-consuming and can therefore interfere with workers ability to meet their obligations at
home or at the workplace. Workers may attempt to hasten the interview process by providing curt

55
and/or compliant responses, which may not always reflect the reality of their situation. Secondly,
since there is little opportunity to probe in depth, there is a high risk that the responses will only
provide a partial understanding of workplace issues. Finally, the individual interview can be an
intimidating experience for workers (especially insecure non-permanent workers) who, despite the
assurance of confidentiality, are likely to worry that disclosing certain information could cost them
their jobs. For these reasons, it is important to combine SSIs with the use of other participatory
social auditing tools.

Focus Group Discussions

SSIs can be a useful springboard to FGDs. There were 13 focus group discussions conducted on four
of the sample farms (see Appendix 4), which facilitated a deeper exploration of workplace issues and
provided a clearer picture of the problems that workers face. The interaction and group dynamics of
the participants appeared to enhance the security of workers, who became more open to sharing
emotionally-sensitive information. A number of pertinent issues emerged in FGDs that would have
been unlikely to surface using more structured methods including: workers perception towards their
working environment; their relationship with the supervisors and management; sensitive matters such
as sexual harassment and sexual favours; and the degree of empowerment experienced among the
different groups of workers. For example, one of the participants who had rejected the sexual
advances of a supervisor was able to express her fears of being dismissed with the group, which
elicited empathy from the group participants. Three main trends emerged in FGDs:

1. Although all participants in the FGDs actively participated, those who are articulate tended to
contribute more information.

2. In the mixed-sex FGDs, there was a tendency for male participants to dominate. On the other
hand, mixed FGDs based on either employment status or nature of work did not reveal any
differences in the level of participation between the permanent and seasonal workers nor among
those working in the packhouses and greenhouses.

3. There were a number of gender issues that emerged more clearly in FGDs than in SSIs, including:
exposure to chemicals; verbal abuse and sexual harassment; impact of low income on family
responsibilities; and the implications of employment on the ability to care for children. All of these
issues are extremely important to women workers, and would not necessarily have been raised
had the research been limited to SSIs.

Group Exercises
Mapping:
Participants were requested to create a simple map of their working environment, identifying key
sites such as packhouses, greenhouses, toilets, drinking water points, schools, and childcare and

56
housing facilities. This exercise, which was undertaken before the start of the FGDs, was a good ice-
breaker and helped to generate lively discussions. However, mapping is time consuming and can
significantly increase the length of the FGD if not managed carefully.

Role Play:

A total of six role-play exercises were carried out in the FGDs, each addressing an area of work-
related concern. Role plays are particularly useful for acquiring a deeper understanding of certain
issues (e.g., harassment by supervisors came out very clearly as the same abusive language used by
the supervisors was imitated during the role-play), for promoting group cohesion, and for
encouraging greater participation of reticent group members. Role plays were especially effective at
eliciting the perspectives of women, who expressed themselves most fully in this format.

Ranking Exercises:

Following a discussion of their problems at work, participants were asked to rank the problems in
order of importance and draw a diagram (e.g., a wheel or spider) to represent the rankings. Some
participants were able to rank issues only after concrete examples were provided; they also found it
relatively straightforward to rank the first three issues, and thereafter found it difficult to determine
the relative importance of the remaining issues. Facilitators might consider reducing the number of
issues to be ranked to six pertinent ones.

Transect walk:

Finally, a farm walk of the working environment was conducted on two farms. It was particularly
useful in capturing experiences that could not be conveyed by either SSIs or FGDs. For example, it
was only through the transect walk that the research team was able to experience firsthand the
intensity of heat and humidity in a greenhouse.

57
7. Multi-stakeholder Approaches

It is clear that the application of various participatory social auditing techniques and tools could help
to mitigate criticisms that auditing has failed to adequately capture the employment conditions of
workers. However, as stated above, participatory social auditing is about more than the application of
tools and methods. For participatory social auditing to enhance the code implementation process, and
improve conditions for workers, it can only be one component in a wider process of social change in
which all stakeholders (consumers, buyers, producers, workers, government, and civil society
organisations) work together. This points to the need for a local multi-stakeholder approach to code
implementation, which can help to ensure that the information derived through participatory methods
is understood in the local context and acted upon.

By the end of the 1990s, a number of initiatives emerged in various industries to foster the
participation of non industry parties in the process of monitoring and verification, most recently in the
form of multi-stakeholder initiatives that encourage stakeholder dialogue and social learning. The
idea was that encouraging a constructive dialogue among diverse stakeholders (that represented the
interests and concerns of workers and producers) would enhance the depth and accuracy of audits,
and help to build capacity among local stakeholders to address the issues that emerged. Over time,
the multi-stakeholder approach to code implementation has gathered momentum and is now
associated with a range of advantages including enhanced credibility, compatibility with existing
regulations and standards, and a greater likelihood of addressing the needs and priorities of both
growers and workers (NRI 2002).

In African horticulture, there has also been a growing recognition that a multi-stakeholder approach
to code implementation may have greater potential to effect improvement in labour conditions,
particularly in the capacity to ensure the representation of marginal workers such as women. The
experiences of the Ethical Trading Initiative (ETI) Wine Pilot in South Africa which led to the creation
of the Wine Industry Ethical Trade Association (WIETA), and the Agricultural Ethics Assurance
Association of Zimbabwe (AEAAZ) have provided models of locally-owned multi-stakeholder initiatives
oriented toward a continual process of workplace improvement. These efforts are being adapted in
other countries, where multi-stakeholder bodies are bringing together a wider group of perspectives
(private sector, government, trade unions, and civil society stakeholders) to bear on the auditing
process and to develop solutions for change through on-going dialogue. In Kenya a multi-
stakeholder Steering Committee has also been established to raise social standards in the horticulture
industry. This initiative the Horticultural Ethical Business Initiative (HEBI) aims to ensure that the
auditing process is inclusive of all stakeholders (industry, NGO, trade union and government) who
come together to identify problems and to develop solutions for improvement (See Box 12).

58
 Box 12. Horticultural Ethical Business Initiative (HEBI), Kenya
During 2002, local civil society organisations spearheaded a campaign against poor working
conditions on Kenyan flower farms, spawning a series of articles in the Kenyan press. These
activities generated concern about the reputation of the industry in overseas markets, and were
responsible for bringing together a range of stakeholders to engage in dialogue on the labour
practices of flower farms. A Steering Committee the Horticultural Ethical Business Initiative
(HEBI) was formed to guide social accountability in the horticulture sector. It is a tri-partite body
composed of members from government, civil society organisations and trade
associations/employers. Unions were invited to participate, although to date they have declined to
take part. There are also a number of observers, including the ETI, DFID, the United States Agency
for International Development (USAID), the Royal Netherlands Embassy and Africa Now (an
international NGO). HEBI has also received a commitment of support from the Kenyan
government.

HEBI plans to systematise a multi-stakeholder approach to code implementation, and has identified
a number of organisational goals to facilitate this. However, it has experienced teething problems
stemming from the difficulties entailed in managing the expectations of stakeholders with varying
interests. Nevertheless, there is consensus among the committee that a multi-stakeholder approach
to code implementation (including worker participation) may be the only way to resolve some of
the persistent employment problems facing workers in the industry, while simultaneously
maintaining Kenyas reputation and access to European markets.

Initiatives such as HEBI, AEAAZ and WIETA have several advantages. Firstly, the establishment of an
independent body comprised of diverse stakeholders provides a degree of protection from any one
stakeholder dominating the process. Secondly, because gender issues are socially embedded at a
local level, ensuring the participation of organizations that reflect those interests is likely to improve
the prospects for genuine change. Thirdly, a multi-stakeholder process can engender a sense of local
ownership over the code implementation process and counter the criticisms that codes are merely a
Northern imposition.

59
SECTION 4: RECOMMENDATIONS

8. Ways Forward

The employment problems discussed in Chapter 5 are not insurmountable. The fact that labour
conditions have improved over the last year provides cause for optimism. However for codes to
provide a long-term shift in employment conditions and contribute to gender rights, all stakeholders,
including workers, must have a role in the process of code implementation. The following identifies
several actions that can be undertaken to create sustainable and continuous improvement for workers
in the sector, broken down into recommendations for different actors. Most of these
recommendations were raised during the FGDs by workers themselves and in a stakeholder workshop
held in Kenya as part of research process. Taken together they highlight the fact that the
responsibility for better working conditions needs to be borne by all stakeholders.

8.1 Employers

Education and Training

Provide information and training on company policies and procedures, and terms and conditions
of employment to all workers. All workers should have a copy of their contract (in the local
language) and should be given pay slips.

Provide workers with training on labour legislation and their rights as workers. Because workers
are often reluctant to express their views in front of management personnel, independent bodies
should carry out training for the workers.

Provide workers with a full and comprehensible explanation of the code (both verbally and in
written form), and how it is implemented and audited. Involve workers in internal monitoring
and auditing of codes.

Codes should be posted in a prominent place in the packhouse/farm in the local language(s) of
the workers. Copies of the code should also be made available to workers if and when requested.

Ensure that all management personnel are knowledgeable about the code and how it is being
implemented/audited.

Provide supervisors with training on appropriate methods for managing workers, including
supervisor/staff training on gender sensitivity and workers rights.

60
Gender-Sensitive Work Environment
Ensure gender sensitivity in task allocation and provide women with the training required to
achieve promotion to supervisory and middle-management levels.

Provide light duties for pregnant women and breaks for breastfeeding mothers.

Provide reliable, high quality childcare facilities.

Establish an equal opportunities policy.

Benefits
Continue positive trend of transferring non-permanent workers to permanent contracts. Provide
seasonal workers with preferential treatment for re-hiring. Extend benefits to non-permanent
workers on a pro rata basis.

Aspire to the provision of living wages and benefits that meet the basic needs of workers and
their families and allow some discretionary income.

Provide safe, timely and reliable transport services.

Provide more comprehensive health care and personal protective equipment to all categories of
workers.

Communication Channels
Improve communication with workers through unions, workers' committees, management-worker
meetings and notice boards. Language(s) used for communication should be understandable to
all workers. Develop channels of communication that women feel comfortable using.

Establish a confidential forum for workers grievances appropriate to all categories and gender of
workers. Workers should also be provided a confidential and accessible channel through which to
report code violations or to make anonymous complaints when they believe their rights are being
violated.

Workers committees need to be strengthened, and include the representation of both men and
women in all categories of work. Consider establishing Women's Committees.

Workers should be supported if they wish to join trade unions.

8.2 Buyers (Importers, Wholesalers, Supermarkets)

Develop Worker-Sensitive Supply Chain Relationships

All buyers should visit farms and familiarize themselves with the issues facing producers and
workers.

61
 Buyers need to be fully aware of the link between the pressure to meet order deadlines and
hours of work and overtime practices. Buyers should provide more lead time on orders and stable
production orders.

Ensure that all suppliers are knowledgeable about codes, and are committed to fully complying
with its provisions.

Supermarkets and other buyers should assist suppliers to meet the costs of compliance, for
instance, by contributing funds for a multi-stakeholder body to oversee a local auditing process or
by including the costs of compliance in product pricing.

Harmonise codes. Cut flower producers are confronted with several codes of conduct,
encompassing somewhat different standards and monitoring requirements. Harmonisation of
codes would reduce confusion and compliance costs, and also simplify worker education on
codes.

Make clear to suppliers that the provisions of codes should be extended to non-permanent
workers.

Reward suppliers that are positively implementing codes, e.g., with Preferred Supplier status.

Actively support the use of participatory social auditing methods.

Allow suppliers to combine audits of different codes to save both time and money. Dovetailing
audits will also minimise disruption to production.

Work with suppliers to correct problems when they are identified. Suppliers should be given
sufficient time, opportunities and support to achieve compliance. A corrective action plan with
reasonable deadlines should be developed and followed.

Ensure that smaller producer/exporters and smallholders are not discriminated against in the
process of code implementation.

8.3 Auditing Bodies

Consider using focus group discussions and other participatory social auditing techniques and
tools in order to enhance the quality and gender-sensitivity of audits.

Provide auditors with training in participatory social auditing methods and gender training. Invest
in training local women (and men) as social auditors.

All auditors should be knowledgeable of relevant local laws and the social, economic and political
context, and fluent in the local language(s).

Ensure that workers interviewed during the auditing process are selected on a random basis,
without management interference, that workers anonymity is protected, and that there are no
management attempts to influence what workers say or to punish workers for reporting
62
 violations. If workers appear reticent or intimidated, off-farm interviews should also be carried
out to ensure that workers feel safe to disclose their views.

Audits should be evaluated to ensure that all categories of workers (at different levels of authority
and in different work categories) are consulted during the auditing process.

Ensure that local and civil society organisations are substantively incorporated into the auditing
process.

Ensure that auditors are aware of potential sensitive gender issues and are cognizant of the need
to probe for embedded problems.

8.4 Trade Unions

Ensure that non-permanent workers are eligible for membership.

Extend Collective Bargaining Agreements to cover non-permanent workers.

Educate workers about national and international legislation, workers' rights and codes.

Promote gender equality through collective bargaining. Encourage and support women to assume
leadership roles within the union at national, regional and company levels.

8.5 HEBI
Institute gender-sensitive participatory social auditing, encouraging the involvement of
marginalized groups.

Ensure that the process of code implementation is locally owned and that the process of
auditing/verification emphasises flexible application that addresses the local social/cultural
context.

8.6 Non Governmental Organisations

Train workers on national and international legislation, workers rights and codes.

Demand secure channels for worker complaints.

Maintain pressure on producers and retailers to account for employment practices in the cut
flower supply chain.

Provide expertise to growers, especially in areas of training and gender issues.

Participate in monitoring.

8.7 Kenyan Government

Grant non-permanent workers protection in all labour legislation.

63
 Ratify ILO Convention 183, which extends the length of maternity leave from 12 to 14 weeks, or
adopt the maternity protection of 90 days in the Equality Bill.

Ratify ILO Convention 87 covering workers right to Freedom of Association.

Implement all ILO Conventions that have been ratified, especially the Core Conventions.

Enact the proposed Equality Bill, 34 2000 that recognises equal pay for equal work and anti-
discrimination provisions.

Incorporate the Convention on the Elimination of All Forms of Discrimination against Women
(CEDAW) into domestic law and ensure that the provisions of the Convention are fully reflected in
the constitution and in all legislation.

Increase efforts to create awareness about the need to change stereotypical attitudes and
discriminatory behaviour towards women and girls through educational campaigns.

Improve enforcement of labour and occupational health and safety legislation.

Provide state assistance for childcare and maternity leave.

Keep abreast of the regulatory and voluntary standards of importing countries, and disseminate
this information to exporters/producers.

Provide training and support to small producers/exporters to meet the criteria contained in codes.

Support the capacity of multi-stakeholder bodies and help to broker cross-sector knowledge
sharing in order to strengthen the learning process from code implementation.

Taken together, these initiatives could contribute to a process approach that embodies a commitment
to sound labour practices that can improve working conditions, and enhance the overall productivity
of cut flower producers. It is key, however, that these actions be adopted in the spirit of a
developmental approach, where education and awareness raising occur throughout the supply chain,
including among consumers and Northern stakeholders, so that a local, multi-stakeholder process
approach becomes more widely accepted as essential to achieve substantive and sustainable change
for all workers.

34
See Republic of Kenya (2002a) and (2002b) for a discussion of the Equality Bill and other gender-related
legislation.
64
9. Conclusion

Kenyan producers have been at the centre of initiatives to raise social and environmental standards in
the cut flower industry both through the development of their own industry codes and the adoption of
overseas buyer codes. As a result, over the last year there have been a number of positive changes in
the employment practices found on cut flower farms. However, despite the constructive steps that
growers have taken to comply with codes and improve working conditions, a number of employment
problems persist ranging from occupational health issues to gender-related concerns such as sexual
harassment and maternity leave.

To an extent, the problems identified by workers can be ameliorated by successful implementation of

existing codes (e.g. improved wages, improved worker representation and better health and safety).
But some of the challenges are more systemic. These include:

Gender and social norms in Kenya: Many employment conditions (e.g., sexual harassment,
discrimination and occupational segregation) are deeply entrenched in gendered social relations and a
patriarchal employment context that exists in Kenya. This is reinforced by a legislative context which
upholds gender discrimination in many laws and statutes. Such embedded social norms are not easily
surmounted by the adoption of codes or employment policies alone, and require broader processes of
education and empowerment.

Hierarchical organisation of companies: As noted, the organisational structure of companies is

often quite rigid with communication occurring through vertical channels. Supervisors are delegated
substantial power for hiring, firing, discipline, and promotion, and serve as a barrier between workers
and management. Without training supervisors tend to replicate the gender and social norms
mentioned above. This is exacerbated by the fact that there are relatively few female supervisors,
and an absence of grievance mechanisms or viable workers committees through which workers can
articulate their concerns.

Buyer behaviour: Buyers in Europe (particularly supermarkets) tend to pass risk and insecurity
down the supply chain. For example, the institution of just-in-time production methods passes the
costs of inventory holding to exporters, creating long and unpredictable hours for workers in Kenya.
At the same time, competitive pressures are intense. Cut flower producers are squeezed in returns,
and deflect this pressure onto the workforce, obliging employees to work longer and harder so that
producers can maintain market access. Finally, the costs of code compliance are also passed to on to
producers. This fosters resentment on behalf of suppliers, and works against creating conditions
conducive to long-term social improvement.

65
Top down approach: Codes in the cut flower industry have been largely driven by developed
country consumers and business, rather than producer opinions or worker priorities. Until recently,
the absence of multi-stakeholder engagement in code development in Kenya led to a process of ad
hoc worker involvement.

Snapshot and technical auditing: The inclusion of worker and stakeholder participation into the
auditing process is at an early stage. In most cases auditing has been largely a technical exercise that
missed many of the deeper issues that concern workers. This was because the motivation for
adopting codes was to satisfy the requirements of Northern markets rather than to effect substantive
changes in employment practices and working conditions. Yet a checklist, compliance approach is ill
equipped to identify the types of problems workers face, leaving poor working conditions intact.

The diversity of these factors points to the importance of having a process of code implementation
that acknowledges the uniqueness of local conditions, and the role that all stakeholders play in
improving working conditions. However, adding to the ever-increasing number of codes by designing
more that are specific to every employment situation is not feasible. There is a need for widely
agreed codes that can be flexibly implemented to accommodate diverse cultural contexts and
workers own perceptions of their employment needs. Hence, for workers to achieve full employment
and social rights, including the recognition of gender-specific concerns, there needs to be a more
inclusive and systematic approach that incorporates several interdependent actions.

Towards a Strategy for Inclusive, Systematic and Gender-Sensitive Ethical Trade

Firstly, code development, monitoring and verification are more effective if there is a sense of local
ownership over the process and if a range of stakeholders buy into the process. These stakeholders
include all workers, including those in less secure forms of employment, as well as local institutions in
which workers have trust. From a gender perspective, it is essential to include local organisations
that are able to reflect women's interests. Such an approach aims to ensure local ownership in the
compliance process by eliciting worker participation and by bringing a wider group of perspectives
(private sector, government and civil society stakeholders) together to identify common solutions for
improvement. A multi-stakeholder approach to code implementation (including worker participation)
that begins to resolve some of the employment problems in the industry will not only improve labour
practices but will also help companies to realize commercial benefits through a more productive
workforce and higher quality products.

Secondly, it is essential that gender-sensitive auditing be undertaken. This entails two pivotal changes
in the auditing process. At one level auditors need to have familiarity with local conditions, including
local language skills, knowledge of relevant national legislation, as well as an understanding of social

66
and gender norms. Auditing teams also need to be gender balanced. Although female auditors do
not necessarily guarantee a gender-sensitive process, female workers may feel more comfortable
discussing sensitive issues such as harassment and abuse with female auditors. Irrespective of sex,
all auditors require gender awareness training to be equipped to identify gender employment
concerns and discrimination in the workplace.

At another level, participatory social auditing procedures should be developed and implemented as
part of a local multi-stakeholder initiative. This would include the systematic use of participatory
methods of social auditing as described in this report, which would help to ensure the active
involvement of different categories and genders of workers in the auditing process. Such auditing
encourages the participation of marginalised workers, especially non-permanent workers and women,
who may lack other forms of representation. There also needs to be a mechanism through which
workers concerns are fed back to management (anonymously) to ensure corrective action that
addresses the needs of all workers.

Finally, for codes of conduct to be truly effective, there must be a continual process of awareness
raising and improvement with an ultimate aim of fostering a work environment where the social and
economic rights of workers are respected. This requires a process of education of management and
workers, a recognition by management that improving labour conditions can enhance the productivity
and quality of work, and a change in employment culture that aims to ratchet up conditions as a
means of competitive advantage. It is in the interest of all stakeholders, including the government,
trade unions, workers, buyers, consumers, and civil society, to ensure that this occurs.

67
Appendices

Appendix 1
World Exporters of Floricultural Products, 2001
Rank Country Change 2001 Change 2000 Change
00/01 Total 99/20 Total 98/99
98 99 00 01 % $000 $000
Total -4% 7,321,619.25 7,662,924.00 -6%
1 1 1 1 Netherlands -2% 3,730,959.71 -7% 3,810,619.89 -8%
2 2 2 2 Colombia -23% 441,311.55 4% 570,334.66 -9%
7 6 3 Canada -7% 285,341.99 14% 305,593.00 8%
6 5 4 4 Belgium 5% 280,032.23 -2% 267,603.00 0%
4 3 3 5 Italy 2% 275,085.94 -9% 268,926.56 -1%
3 4 5 6 Denmark 3% 270,138.76 -9% 261,628.45 -9%
5 7 6 7 USA 5% 231,658.01 1% 220,871.04 -22%
10 9 8 8 Germany 13% 224,044.76 -1% 197,978.18 -1%
9 8 7 9 Ecuador -8% 198,018.80 2% 215,977.34 4%
13 12 9 10 Kenya 14% 186,668.76 5% 164,204.09 12%
All values are in US$000. In 2001 the value of the Euro dropped by three percent
Source: www. pathfastpublishing.com from Customs and Excise

68
Appendix 2

National Legislation and International Conventions Relevant to Cut Flower

Workers

I. Kenya National Legislation

The Employment Act Cap 226

The Workmens Compensation Act Cap 236
The Regulation Of Wages And Conditions Of Employment Act Cap 229
The Factories Act Cap 514
The National Hospital Insurance Fund Cap 255
The National Social Security Fund Cap 258
The Trade Unions Act Cap 233
The Trade Disputes Act Cap 234

II. ILO Conventions

Core Conventions:
ILO Conventions 29 and 105 & Recommendation 35 (Forced and Bonded Labour)
ILO Convention 87 (Freedom of Association)
ILO Convention 98 (Right to Organise and Collective Bargaining)
ILO Conventions 100 and 111 & Recommendations 90 and 111 (Equal remuneration for male
and female workers for work of equal value; Discrimination in employment and occupation)
ILO Convention 138 & Recommendation 146 (Minimum Age).
ILO Convention 182 & Recommendation 190 (Worst forms of Child Labour)

Although not core ILO conventions, the following are relevant to cut flower workers:
ILO Convention 81 (Labour Inspection)
ILO Convention 110 (Plantations Convention)
ILO Convention 122 (Employment Policy)
ILO Convention 131 (Minimum Wage Fixing)
ILO Convention 135 & Recommendation 143 (Workers' Representatives Convention)
ILO Convention 155 & Recommendation 164 (Occupational Safety & Health)
ILO Convention 154 (Collective Bargaining)
ILO Convention 156 (Workers with Family Responsibilities)
ILO Convention 159 & Recommendation 168 (Vocation Rehabilitation & Employment/Disabled
Persons)
ILO Convention 175 (Part time work)
ILO Convention 177 & Recommendation 184 (Home Work).
ILO Convention 183 (Maternity Protection).
ILO Convention 190 & Recommendations (Safety and Health in Agriculture)

III. UN Conventions and Declarations

Slavery Convention (1927); Supplementary Convention on the Abolition of Slavery, the Slave
Trade and Institutions and Practices Similar to Slavery (1956)
Universal Declaration of Human Rights (1948)

69
 Convention on the Suppression of the Traffic in Persons and the Exploitation of the
Prostitution of Others (1949)
Convention on the Political Rights of Women (1952); Fourth World Conference on Women:
Beijing Declaration and Platform for Action (1995)
International Convention on the Elimination of All Forms of Racial Discrimination (1965)
International Covenant on Economic, Social and Cultural Rights (1966)
International Covenant on Civil and Political Rights (1966)
Convention on the Elimination of All Forms of Discrimination against Women (1979)
Convention on the Rights of the Child (1989)
World Conference on Human Rights: Vienna Declaration and Programme of Action (1993)
UN International Conference on Population and Development: Cairo Programme of Action
(1994)
International Convention on the Protection of the Rights of All Migrant Workers and Members
of Their Families (2003)
Protocol Against the Smuggling of Migrants by Land, Sea or Air (not yet in force)

Sources: Gammage et al. 2002, Office of the High Commissioner for Human Rights 2003

70
 Appendix 3

Labour Force Population and Sample Size of Flower Firms

Type of Employees
Farm Codes No. Employees Permanent Seasonal Casual Sampl
Applied e Size
F M Total F M F M F M Total
1 KFC, BRC, 78 123 201 3 15 78 105 13
EUREPGAP (10%) (6%)
2 KFC, ETI 490 354 41 483 329 7 25 41
Base Code, (42%) (5%)
BRC,
EUREPGAP,
M&S,
Safeways,
LNGG code
3 FLP, KFC & 188 144 332 82 70 106 74 19
FPEAK (17%) (6%)
4 None 210 40 250 63 8 147 32 13
(13%) (5%)
5 KFC, MPS, 219 151 370 176 125 36 21 7 5 14
FLP, Max (19%) (4%)
Havelaar
Total 1185 812 1997 807 547 374 257 7 5 100
(59%) (41%) (100%) (67%) (27%) (19%) (13%) (.4%) (.3%) (5%)

In July-August 2002 the five farms/packhouses had a total workforce of 1997. The sample of 100
workers (5 percent) was selected using a stratified random sampling technique according to gender,
employment status (permanent, seasonal, or casual) and nature of work (packhouse, greenhouse and
Crop Development Unit (CDU)). This sample of workers drawn from the five firms is in proportion to
the total population of workers in these firms. This ratio was calculated using the following formula:

f = n
N
Where: f = the ratio of the sample of the population
n = size of the sample
N = size of population

F is equivalent to 100/1997 = 0.0501, this ratio was then used to determine the number of workers
to be selected from each firm, in some of the farms/packhouses the number of workers selected were
increased or decreased to ensure that all the strata were included in the sample.

71
Appendix 4

Case Study Methodology

Aims and Objectives

The aims of this project were to analyse how ethical trade can enhance the economic and social
rights of women and men workers in the Kenyan cut flower industry, and to identify best practice in
implementing gender-sensitive ethical trade based on worker and stakeholder participation.

The objectives of the project were threefold:

1. Examine the gendered needs and rights of workers in the cut-flower industry in Kenya.
2. Assess the extent to which these gendered needs and rights are being, or could be,
addressed through worker and wider stakeholder participation.
3. Assess the implementation/impact of codes in the cut-flower industry in Kenya.

Methods for Data Collection and Analysis

This study utilized qualitative methods to collect primary data. These methods consisted of: semi
structured interviews and in-depth focus group discussions conducted with workers. Primary data
were collected using two semi-structured interview schedules administered to firstly, the management
of cut flower producers and secondly, the workers in selected cut-flower firms. The latter were those
directly engaged in the production of cut-flowers and excluded the management and the
support/administrative staff. Thereafter, substantive issues were further explored in small focus group
discussions (FGDs). During the FGDs, a variety of tools were used in order to both elicit information,
and to assess and compare their effectiveness for doing so.

Sampling Procedure
There are over 500 flower farms and/or packhouses in Kenya (Thoen et al., forthcoming). However,
the sample of the study was determined by selecting only those applying code(s) of conduct. This
formed the population of the study. The core fieldwork sample consisted of five flower production
companies from the Lake Naivasha, Nairobi and Ruiru areas selected according to the type(s) of
codes in application; this was an attempt to ensure a representative coverage of the codes. The
inclusion in the survey of one company that does not apply any codes was included as a control
factor. Two additional companies were included in the 'Company Interviews', but 'Worker Interviews'
were not conducted on their farms. However, the sample was not large enough to allow
generalisations across the population, especially since another important criterion for the selection of
the farm was the willingness of the management to participate in the research project.

72
These 5 farms/packhouses have a total workforce of 1997. A sample of 100 workers was selected
using a stratified random sampling technique according to gender, employment status (permanent,
seasonal, or casual) and nature of work (packhouse, greenhouse and CDU). In each case the sample
of workers was drawn from a roster of employees provided by the management. The sample of
workers from each farm/packhouse was in proportion to the total population. Following completion of
the SSIs, a preliminary analysis was conducted in order to ascertain the key workplace issues that
were raised, and which were to be probed in greater depth through focus group discussions and
other participatory methods.

Distribution of Sample by Company

Company No. of Worker SSIs No. of Workers per FGD

1 13 (6), (6)
2 41 (7), (8), (6), (7)
3 13 No FGDs conducted
4 19 (6), (6), (8), (6)
5 14 (6), (6), (7)
Total Workers 100 85
Source: Research Data, 2002

Purposive sampling was undertaken for the FGDs, in order to ensure participation of the different
types of workers. This classification was made on the basis of gender, nature of work (i.e. packhouse,
greenhouse and CDU concerned with spraying, irrigation and fertilizer application) and employment
status (permanent and seasonal). A total of 13 FGDs were conducted among the four firms applying
codes. The number of participants in the FGDs ranged from five to eight, for the most part drawn
from the sample of SSIs conducted previously. The composition of the groups varied. Nine were
comprised of mixed categories of workers across gender, nature of work and employment status
categories, while four were homogenous. During several FGDs, group exercises comprised of
mapping, role-play and ranking tools were facilitated, and transect walks through the farms were also
carried out. All FGDs were conducted. Four focus groups were facilitated by women, three by men,
and the remaining five by both men and women.

73
 Composition of Focus Group Discussions

Category of Worker Farm 1 Farm 2 Farm 4 Farm 5

Female field workers - - 1 1
P only P&S
Female packhouse - - 1 1
workers P&S P&S
Female packhouse 2 1 - -
and field workers 1 = P; 2 = P&S P only
Male CDU workers - 1 1 1
P&S P&S P&S
Female In-charges - - 1 -
P only
Mixed CDU, field & 2 - - -
packhouse workers P&S
Notes: P = Permanent; S = Seasonal
Source: Research Data, 2002

74
References
Auret, D., 2001, Participatory Social Auditing: A Handbook for Code of Practice Implementors,
Zimbabwe: AEAAZ.

Barrientos, S., Dolan, C., and Tallontire, A., 2001, Gender and Ethical Trade: A Mapping of the Issues
in African Horticulture, Working Paper No. 26, Chatham Maritime: NRI.

Barrientos, S., Dolan, C., and Tallontire, A., 2003, A Gendered Value Chain Approach to Codes of
Conduct in African Horticulture, World Development 31(9):1511-1526.

Blowfield, M.E., 2003, Ethical Trade: The Negotiation of a Global Ethic. DPhil Thesis. Falmer, UK:
University of Sussex

Caldwell D.J., 1998, Ecolabeling and Regulatory Framework A Survey of Domestic and International
Fora, http://www.consumerscouncil.org.

COLEACP, 2002, European Imports of Flowers and Ornamental Plants 1994 2000,
http://www.coleacp.org/

Dwasi, J., 1999, Kenya: A Study in International Labor Standards and Their Effect on Working Women
in Developing Countries: The Case for Integration of Enforcement of World Bank Policies, Wisconsin
International Law Journal 17: 347-462.

FAO, 2002, A Thorn on Every Rose for Kenyas Flower Industry, Newsroom: Rome: FAO.
http://www.fao.org/english/newsroom/news/2002/3789-en.html

Flowers Growers Group of the Agricultural Association, 2002, Memorandum of Agreement between
Agricultural Employers Association and the Kenya Plantation and Agricultural Workers Union. Nairobi.

Galinsky, B., 1997, UK Rose Market in Full Bloom, Market Asia, V. 4(2),
http://www.marketag.com/ma/news/archive/v42/Rose.stm

Gammage, S., Jorgensen, H., McGill, E., and White, M., 2002, Framework for Gender Assessments of
Trade and Investment Agreements, Washington, DC: Womens EDGE, Global Trade Program. Draft,
April 2002.

Hennock, M., 2002, Kenya's Flower Farms Flourish, BBC News 14 February
http://news.bbc.co.uk/1/hi/business/1820515.stm

IBLF, 2002, Violence against Women in the Workplace in Kenya: Assessment of Workplace Sexual
Harassment in the Commercial Agriculture and Textile Manufacturing Sectors in Kenya,
http://www.laborrights.org/projects/women/kenyareport.pdf

The International Floriculture Quarterly Report, 2002, UK Market Share 2000,

http://www.pathfastpublishing.com/qr27/qreport.htm

International Trade Centre (ITC), 2001, Product Profile: Cut Flowers and Foliage, Business Sector
Roundtable, Discussion Document, Brussels, 16 May. http://www.intracen.org/bsrt/ppcutflowers.pdf

KFC, 2002, A Growing Responsibility, http://www.kenyaflowers.co.ke/page4.html

Ladbury, S. and S. Gibbons, 2000, Core Labour Standards. Key Issues and a Proposal for a Strategy,
London: Department for International Development, January.

75
Van Liemt, G., 1998, The Social Policy Implications of Codes of Code, with Particular Reference to the
Relations between Companies Adopting such Codes and their Suppliers and Subcontractors. Paper
Presented at the International Workshop on Global Production and Local Jobs: New Perspectives on
Enterprise Networks, Employment and Local Development Policy. Geneva: International Institute of
Labour Studies, 9-10 March.

Murray, J., 1997, Corporate Codes of Conduct and Fair Labour Standards, Paper presented at
International Workshop on Global Production Systems and Labour Markets, International Institute for
Labour Studies, Geneva, 22-23 May.

NRI, 2002, Building Multi-stakeholder institutions for Developing and Managing National Codes of
Practice, NRET Theme Papers on Codes of Practice in the Fresh Produce Sector (No 2), Chatham
Maritime: Natural Resources Institute, http://www.nri.org/NRET/TP2.pdf

PANUPS (Pesticide Action Network Updates Service), 2002, Floriculture: Pesticides, Worker Health &
Codes of Conduct, June 12. http://panna.igc.org/resources/panups/panup_20020612.dv.html

Republic of Kenya, 2002a, Kenya Gazette Supplement The Equality Bill, 2002, Nairobi: Government
Printers.

Republic of Kenya, 2002b, Sessional Paper on Gender and Development, Nairobi: Ministry of Home
Affairs.

Roberts, S. and N. Robins, 2000, The Reality of Sustainable Trade, London: IIED.

Seyfang, G., 1999, Private Sector Self-Regulation for Social Responsibility: Mapping Codes of Conduct.
Ethical Trading and Globalisation: Self-Regulation And Workers Experience, Working Paper 1,
Overseas Development Group, Norwich, UK: University of East Anglia.

Smith, S., Auret, D., Barrientos, S.., Dolan, C., Kleinbooi, K., Njobvu, C., Opondo, M. and A. Tallontire,
2003, Ethical Trade in African Horticulture: Gender, Rights and Participation, Preliminary Report for
Multi-Stakeholder Workshop, June 26, Brighton, UK: Institute of Development Studies.

Thoen, R., Jaffee, S. and C. Dolan, (forthcoming), Equatorial Rose: The Kenyan-European Cut Flower
Supply Chain, in Supply Chain Development in Emerging Markets: Case Students of Supportive Public
Policy, Kopiki, R (ed.) Washington D.C.: World Bank.

VIDEA, 2001, Fair Flowers, Fair Trade and Informed Consumers: Building on European Strategies for
Public Engagement, Victoria: VIDEA, www.videa.ca/support_items/videa_fairflowers.pdf

World Bank, 2002, Gender Stats. Retrieved November 28, 2002.

(29http://genderstats.worldbank.org/genderRpt.asp?rpt=labor&cty=KEN,Kenya&m=home2)

76
For Further information please contact:
NRET
Natural Resources Institute
Central Avenue
Chatham Maritime
Kent ME4 4TB
United Kingdom

Tel: +44 (0)1634 883199

Fax: +44 (0)1634 883706
Email: [email protected]
Website: http://www.nri.org/NRET