A New Approach to Industrial Censuses and Surveys for Southern African Countries—A Case of Zimbabwe ()
1. Introduction
Industrial censuses and surveys are very important to an economy because they inform policy and also an important source of data for further research by both private researchers and academia. Economic data is generated through these censuses and surveys and hence there is need for them to be conducted with a high degree of precision by any producer of statistics within the statistical system. Due diligence has to be taken from the sample size selection and the selection of the experimental units in case of surveys. This must be done by real seasoned experts. The execution of the industrial surveys and censuses should be done by well knowledgeable personnel so that the data collected is of high quality which then inform policy accurately and which then also forms as a basis for further research by academia.
1.1. Statement of the Problem
With the demand of industrial data for policy formulation and research, government and researchers are finding it difficult to trust the statistics generated by the national statistical office and private statistical agencies in Zimbabwe due to very low response rates in industrial censuses and surveys. The low response rates in industrial censuses and surveys are increasingly become a major problem both for the producers of statistics and the users of statistics in Zimbabwe.
Almost all the establishments within an economy are registered under the government and always follow the government laws when operating. The major producer of official statistics is the national statistical office in an economy and is owned by the government and uses a statistical act enacted by the Parliament when conducting its operations. The private producers of statistics are also within the national statistical system and have the mandate to generate official statistics in partnership with the national statistics office.
Given such situation, low response rates continue to be experienced by producers of official statistics in Zimbabwe. The reports generated by producers of statistics do not reflect what is actually happening on the ground as most of industrial respondents refuse to complete the questionnaires by giving various reasons. The low response rates in these industrial censuses and surveys are a cause of concern and the causes are yet to be known. There is need to find solutions to address the problems causing low response rates in industrial censuses and surveys.
1.2. Purpose of the Study
The purpose of the research was to identify and develop procedures which need to be taken by any producer of statistics when conducting industrial censuses and surveys in Southern Africa particularly in Zimbabwe. Again, the research is aimed at identifying the root causes of poor response rates when these activities are carried out by producers of statistics in Zimbabwe. Higher response rates improve acceptance and credibility of research findings among key stakeholders [1]. On the other hand, low response rates will make the survey unrepresentative of the population under study [2].
1.3. Research Objectives
The objectives of the research were to:
1) Analyze the current methods of conducting industrial censuses and surveys by producers of statistics in Zimbabwe and other countries in Eastern and Southern Africa and deduce the problems associated with the methods.
2) Identify the weaknesses of the current methods being used by producers of statistics when conducting industrial censuses and surveys and develop modified systems of data collection for industrial censuses and surveys.
3) Identify the causes which leads to data suppliers not to respond to industrial censuses and surveys questionnaires.
4) Develop new strategies and methodologies which must be followed by producers of statistics in Eastern and Southern Africa including Zimbabwe, when conducting industrial censuses and surveys in order to get higher response rates.
1.4. Research Questions
The main research question is what causes the poor response rates in industrial censuses and surveys conducted by producers of statistics in Zimbabwe and other Southern African countries?
The sub research questions were:
1) What really caused the industrialists not to respond to the questionnaires sent by producers of statistics?
2) Why is it that the companies do not obey the laws of the government which they operate under and do not give data to producers of statistics as required by law?
3) What are the weaknesses of producers of statistics when conducting these activities that might lead to poor response rates?
2. Review of Literature
[3] stated that response rates, in industrial surveys can be increased by a number of factors with incentives as one of the factors. However, the authors did not say much about the root causes of very low response rates in industrial surveys in India. [4] noted that response rates in industrial surveys increased significantly due to the use of modern methods of data collection and these includes the use of emails, phone and web scrapping, however, the author did not say anything regarding the root causes of the poor response rates in industrial surveys and censuses. The stated reasons for none-response were failure to deliver the questionnaire to the target population and reluctance of respondents to complete the questionnaires.
According to other researchers giving incentives to respondents will increase the response rates. The authors went on to state that multiple response options (online, emails, etc.) increases also the response rates. However, the authors did not say much on the actual causes of the non-responses and why agriculture respondents did not want to complete questionnaires. [5] indicated that family businesses yielded an average of 21% response rate. However, the authors indicated that family members are not open about their businesses than non-family members and this led to poor response rates in family businesses. However, the authors did not say much about the reasons why family members are not open about their businesses. [6] indicated that the mean response rates published on industrial surveys were around 34%. The authors went on to describe several methods of enhancing response rates and little was said on the root cause of the non-response rates or very low response rates.
According to [7], there was no difference in response rates in surveys conducted using mails and online. However, the authors did not mention anything on the causes of poor response rates, although they have indicated that response rates increase with time. [8] highlighted that there was a low response rates in industrial surveys conducted in Sri Lanka in 2005 of 42%. However, nothing was said on the root cause of the low response rates. [9] mentioned that response rates increases with the size of company, with large corporates having a higher response rate than the medium and smallest companies. However, the research did not mention anything about the causes of low response rates. Table 1 shows response rates of surveys conducted by a few selected producers of statistics in Eastern and Southern Africa including Zimbabwe. [10] mentioned three ways of response enhancing strategies, that is, involving key stakeholders, leveraging technology and appealing to respondents’ interest. However, there are some challenges with the approach, as some respondents may not have access to internet and also if respondents are selected out of the sampled respondents, the sample becomes smaller and smaller. [11] went on to point out the features which must be improved on the paper questionnaire so that the response rate increases and nothing was said on the actual causes of the low response rate. [12] indicated that there was need to show response rate when producers of statistics carry out surveys and it is these response rates which can inform the degree of accuracy of the statistics generated. Nothing was said concerning the root causes of the low response rates. Whereas, [13] went on to compare the quality of the questionnaires submitted by late respondents and those submitted by early respondents and nothing was said about why the response rates were very low.
Table 1. Survey response rates by producers of statistics and country.
Producer of Statistics |
Survey |
Country |
Response Rate |
CZI |
Manufacturing Sector Survey-2018 |
Zimbabwe |
47% |
Zimbabwe National Statistics Agency |
Business Tendency Survey-2014 |
Zimbabwe |
49% |
Kenya National Bureau of Statistics |
Census of Industrial Production-2010 |
Kenya |
80.6% |
Tanzania National Bureau of Statistics |
Census of Industrial Production
Report-2013 |
Tanzania |
93.9% |
Statistics South Africa |
Manufacturing: Production and Sales Survey-January 2020 |
South Africa |
81.5% |
Statistics South Africa |
Manufacturing: Production and Sales Survey-August 2021 |
South Africa |
62.7% |
Statistics South Africa |
Manufacturing Industry:
Production-2017 |
South Africa |
74.0% |
Statistics South Africa |
Manufacturing Industry:
Financial-2017 |
South Africa |
74.0% |
Statistics South Africa |
Manufacturing: Utilization of Production Capacity by Large
Enterprises-2022 |
South Africa |
72.4% |
Knowledge and Research Consultancy |
Enterprises
Survey-2014 |
Namibia |
96.7% |
Bank of Zambia |
Quarterly Survey of Business Opinions and Expectations-2021 |
Zambia |
29.9% |
PWC |
2020 Zambia Bank and Non-Bank Industry Survey-2020 |
Zambia |
94% |
IPSOS Zambia |
Enterprise Survey Follow-Up on
COVID-19,
Round 3-2021 |
Zambia |
95% |
DERG |
Survey of Mozambican Manufacturing
Firms-2017 |
Mozambique |
100% |
Statistics Norway |
Industrial Statistics in Mozambique-2000 |
Mozambique |
54% |
Nathan Associates and TNS |
National Small Business Survey of Uganda-2015 |
Uganda |
91.95% |
2.1. Current Methods of Data Collection
The methods which are currently being used by producers of statistics in Eastern and Southern Africa including Zimbabwe when conducting industrial censuses and surveys are paper based questionnaires usually administered by enumerators, emails, web-based questionnaires, online, etc.
2.1.1. Paper Based Questionnaires
Paper based questionnaires are questionnaires printed on paper which are then delivered to respondent by enumerators who will in turn return some day to collect the completed questionnaires. The success of this method of data collection depends on the expertise of the enumerator to explain the survey or census objectives and how to complete the questionnaire to respondents and the ability of the enumerator to represent the producer of statistics adequately in terms of appearance and the survey or census knowledge. In some countries this method of data collection is combined with the use of tablets, where the enumerator will immediately transfer the collected data into a tablet upon collecting the hard copy. However, this method is now being used by few countries in the region as most of the countries have moved to the use of emails and other modern forms of data collection, when collecting industrial censuses and surveys data, which are less expensive. [14] indicated that the use of incentives does not improve the response rates, hence producers of statistics in Zimbabwe do not consider much in giving respondents incentives.
2.1.2. Data Collection by Use of E-Mails
Data collection by use of emails is where by the data collector sends emails to industries explaining the importance of the survey or census and also explaining how the questionnaire is to be completed. This is the most frequently used method by most of producers of statistics both private and state owned. This is the cheapest of all the methods but requires constant follow ups through emails and phone calls and usually works well when it is used after a physical meeting between the producer of statistics and the potential respondents. [15] have shown that there are higher response rates on surveys and censuses done after the respondents are initially met and briefed about the surveys or censuses than when respondents are not initially met. This method initially works well when there is mutual trust between the producer of statistics and respondents. According to other researchers, the use of emails and other forms of electronic data collection methods slightly improves the response rate, but a follow up of the questionnaire does not really increase the response rate. This then reinforce the fact that follow ups usually done by producers of statistics should be done after a physical meeting with data suppliers.
2.1.3. Web Scraping and Internet Scanning
Web scraping is mainly done on the websites of industries which do post information about their activities, prices, production outputs etc. Majority of industries in East and Southern Africa including Zimbabwe do not post important information on their websites making it very difficult to collect meaningful data using web scraping methods. The same also applies to internet scanning. Hence web scrapping and internet scanning cannot be used in Zimbabwe as a means of data collection. Table 2 shows a few selected industrial surveys conducted by producers of statistics in Eastern and Southern Africa showing the methods used during data collection. From the sampled producers of statistics, it shows that the more developed a country is, the more ways of data collection the producer of statistics uses. Producers of statistics who are still using the method of engaging the use of enumerators completely in industrial surveys and census, have small samples. Producers of statistics in South Africa whose sample size is quite large, mainly uses emails when conducting industrial surveys in addition to other methods of data collection. [16] indicated that web surveys are only good where the respondent has access to internet otherwise they might increase nonresponses.
Table 2. Producers of statistics methods of data collection on selected surveys.
Producer of Statistics |
Survey |
Country |
Method of Data Collection |
CZI |
Manufacturing Sector Survey-2018 |
Zimbabwe |
Paper questionnaire, electronic questionnaire, personnel visits, telephone and self-enumeration |
Zimbabwe National Statistics Agency |
Business Tendency Survey-2014 |
Zimbabwe |
Paper questionnaire delivered and collected by enumerators |
Kenya National Bureau of Statistics |
Census of Industrial Production-2010 |
Kenya |
Paper questionnaire delivered and collected by enumerators and some send direct to the producer of statistics. |
Tanzania National Bureau of Statistics |
Census of Industrial Production
Report-2013 |
Tanzania |
Paper questionnaire delivered and collected by enumerators |
Statistics South Africa |
Manufacturing: Production and Sales Survey-January 2020 |
South Africa |
Emails, telephone and personal visits |
Statistics South Africa |
Manufacturing: Production and Sales Survey-August 2021 |
South Africa |
Emails, telephone and personal visits |
Statistics South Africa |
Manufacturing Industry: Production-2017 |
South Africa |
Emails, telephone and personal visits |
Statistics South Africa |
Manufacturing Industry:
Financial-2017 |
South Africa |
Emails, telephone and personal visits |
Statistics South Africa |
Manufacturing: Utilization of Production Capacity by Large Enterprises-2022 |
South Africa |
Emails, telephone and personal visits |
Knowledge and Research Consultancy |
Enterprises
Survey-2014 |
Namibia |
Questionnaire delivered and collected by enumerators |
Bank of Zambia |
Quarterly Survey of Business Opinions and
Expectations-2021 |
Zambia |
Questionnaire delivered and collected by enumerators |
PWC |
2020 Zambia Bank and Non-Bank Industry
Survey-2020 |
Zambia |
emails |
IPSOS Zambia |
Enterprise Survey Follow-Up on COVID-19, Round 3-2021 |
Zambia |
CATI (Computer Assisted Telephone Interview) |
DERG |
Survey of Mozambican Manufacturing Firms-2017 |
Mozambique |
Paper questionnaire delivered and collected by enumerators and some send direct to the Producer of statistics. |
Statistics Norway |
Industrial Statistics in
Mozambique-2000 |
Mozambique |
Paper questionnaire delivered and collected by enumerators and some send direct to the Producer of statistics. |
Nathan Associates and TNS |
National Small Business Survey of Uganda-2015 |
Uganda |
Face to face interviews |
2.2. Weaknesses of the Current Methods of Data Collection
In Zimbabwe, the producers of statistics are facing a problem of very low industrial surveys and censuses response rates. The producers of statistics who mainly use the method of sending enumerators to industries need to ensure that the data collection personnel can represent the producer of statistics adequately and the objectives of the survey or census being conducted must be explained to respondents adequately. Suitable personnel with the required skills for data collection must be recruited by producers of statistics for each particular industrial survey or census. Producers of statistics need to equip data collectors with the necessary tools required for data collection in industry. In addition to the above, regular meetings with industrial captains and introductory meetings before any survey or census is necessary and also report back meetings with the industry are very important. Conducting a survey by phone call should only be done after these introductory meetings are done. According to other researchers, low response rates can lead to misleading conclusions and hence the weaknesses of the current methods of data collection should be urgently addressed.
2.2.1. Weaknesses in the Training of Data Collectors
Most of the poor response rates experienced by producers of statistics in Zimbabwe in industrial surveys and censuses are as a result of the use of inexperienced enumerators who then act as messengers between respondents and the producers of statistics. Data collectors need not to be messengers who carry the paper questionnaire from the office of the producer of statistics to the industry and leave the questionnaire without further explanation. This normally originate from the trainings which are conducted by producers of statistics, where they only concentrate on the questionnaires leaving the aspect of industrial respondent handling which is also crucial when conducting industrial surveys and censuses. The training should cover the questionnaire completion and should also be repeated again as refresher training over a certain period of time which is determined by staff turnover and change in the sample etc.
Industrial surveys and censuses are different from household surveys and censuses in the way they are conducted. When conducting industrial surveys and censuses, the enumerators have to be dressed formally because they usually meet respondents who are dressed formally. In addition, respondents are willing to share financial data with someone dressed formally and who is moving with a vehicle. This was tested and seen to be true in Zimbabwe. The greatest weakness with this system of data collection, especially in Zimbabwe, is that enumerators walks on foot from one industry to another and are poorly dressed. Enumerators trainings usually lack respondent handling, a key component in industrial surveys and censuses.
2.2.2. Weaknesses in the Recruitment of Data Collectors
In addition to the above, the type of recruitment which is normally done by the producers of statistics when conducting industrial censuses and surveys also contribute to poor response rates. In trying to cut done survey or census costs, producers of statistics, mainly in Zimbabwe, recruit personnel with minimum secondary qualifications and in some cases on contract basis, that is, the enumerators are only recruited for that particular survey or census and for the period of that particular survey or census. These enumerators do not have basic public relations skills and they normally do not care about the quality of the data collected and neither the response rates. This increases non-responses if there are no trained supervisors working with them. Industrial respondents, especially the owners of the industry or company, needs to trust the enumerator first before information is shared and as such, such type of mutual understanding and relationship usually takes time to be established, hence a temporary enumerator does not help to improve the response rate in industrial surveys and censuses. The enumerators recruited should possess relevant qualifications which is in line with the type of survey or census to be undertaken. For example, economic surveys and censuses should be undertaken by someone with an economics background who understands the basic language in that field.
2.2.3. Lack of Survey or Census Tools
Again, industrial enumerators should be well equipped with instruments of work, just similar to sales representatives of companies who are given vehicles and are usually dressed nicely and well versed in the products they sell. Likewise, industrial enumerators need such equipment, they need vehicles and adequate salary or allowances which enables them to purchase relevant clothes. Producers of statistics must invest more on the tools of work if meaningful data is to collected from the industry. Most of producers of statistics in Zimbabwe do not provide cars to their enumerators and only provides transport fares. The data collectors find their own means of transport to industry and most of them use public transport to reach to their assigned place of work. This normally compromise the quality of work as these enumerators cover very few industries per day and most of them will not be looking presentable after walking so many kilometers on foot. It has been observed that if enumerators use a vehicle in collecting data from companies, the response rate will rise than when they use public transport and sometimes on foot. The option of providing motorbikes to enumerators was explored by some producers of statistics in Zimbabwe and seen as feasible. The motorbikes would be purchased and loaned to enumerators so that they will gradually pay back.
2.2.4. Weakness on the Online Surveys and Censuses in Southern Africa
Apart from the method of sending enumerators, producers of statistics in Zimbabwe also make use of emails and online questionnaires. This form of data collection needs some follow ups in form of emails and phone calls. For this method of data collection to be successful, initial meetings with industrial captains or representatives of industries and associations of industries has to be done so that the objectives of the industrial surveys and censuses are well explained to data providers before they are asked to provide the data on emails. However, [17] have shown that mail-based surveys still have a higher response rates when compared to web-based surveys in developed countries, particularly the UK.
In addition, also issues to do with trust and confidentiality of the data should be cleared in such meetings before emails are dispatched to individual industries. In addition to the above, regular statistical meetings have to be conducted between producers of statistics and data providers where statistical issues affecting industries are discussed and new data requests from industries are made. In practice, these meetings are rarely done in Zimbabwe. Zimbabwe National Statistics Agency used to hold such a meeting once in five years and majority of the participants were government ministries and agencies with very few from the industry. In addition, producers of statistics should not only knock at industrial doors when they need data. There should be report-back meetings with industries after every industrial survey or census. Producers of statistics should also produce statistics which helps the industries and not only helps policy makers and academia. Industries should see value in having statistics produced in a country.
2.2.5. Weakness of the Method of Phone Calls
The other method of data collection which is becoming common among producers of statistics in Southern Africa particularly in Zimbabwe is, conducting an interview by phone call. This is also one of the cheapest methods of conducting an interview depending on the length of the questionnaire and type of respondents. No industrial survey or census can be done successfully from the start of the interview to the end of the interview by phone calls if the questionnaire is long. Household based surveys can be done by phone calls provided the sample size is small and the length of the questionnaire is short again. Phone calls are only necessary on industrial surveys and censuses as a means of follow ups, that is, checking whether the questionnaires were completed or not etc.
3. The Research
However, the above were mere reasons which the researcher gathered and to ascertain the above reasons, a research in form of a survey was carried out to establish the actual thinking of the industrial captains.
3.1. Research Methodology
The research was carried out following the below steps. A questionnaire was designed and distributed to a few selected industries who knows well the researcher and were in constant communication with the researcher in other areas of statistics. The questionnaire was to gather their opinion on why they do not respond to questionnaires send by producers of statistics. The sample was designed in such a way that all types of industries were represented and all the provinces of Zimbabwe we represented. Turnover as a selection variable was not used and each industry had an equal probability of being selected. The data collection method was via emails and follow ups were done by phone calls since the contact persons at these industries were in constant contact with the researcher and the researcher was well known by these companies. The respondents were asked to complete the questionnaires on behalf of the companies, that is, the responses to the questions were the general feeling of the top management of the company.
3.2. Questionnaire Design
The questions on the questionnaire were constructed in such a way so that the reason why an industry does not respond to the questionnaires send by producers of statistics was captured. The questions investigated whether the government had anything to contribute to the low response rates experienced by producers of statistics when conducting industrial censuses and surveys. The questions also investigated whether the industry had regarded the producers of statistics as having some links with revenue authority as most companies are known to have various ways of evading tax, since from the past experiences, it had been known that when producers of statistics visited industrial outlets revenue authorities will follow. The questions also investigated whether there exist, the fear for unknown, among the operating industries within Zimbabwe when dealing with government and research institutions.
The questions also investigated whether the industry regarded statistics generated by producers of statistics as inaccurate figures and hence see no reason for sharing data. The research questions also investigated whether the industry regarded producers of statistics as being controlled by government and hence concluded that the figures would be altered in favor of government anticipations. The questions also investigated whether the industry regarded the statistics generated by producers of statistics as of no value to industry.
The research questions went on to investigate the weakness of the producers of statistics when conducting industrial surveys and censuses. The questions investigated whether the low response rates were as a result of producers of statistics’ staff not explaining well on how to complete the questionnaires. If not explained well, this then leads to none responses due to respondents not knowing what to share.
In addition, the questions went on to investigate the effects of the dress code of producers of statistics’ staff. The investigation was whether the industry was comfortable in sharing data with staff dressed like ordinary people in the streets or would not share data with such people. This was as a result of numerous complaints by industrial contact persons that the dress code of staff from producers of statistics was not good and were not comfortable in sharing turnover data which such people. Furthermore, the research investigated whether staff from producers of statistics follows up the completion of the questionnaires with the selected industry.
Judging from some survey experiences, if there is no follow up on the completion of questionnaires, the industry will forget to complete and that will be a non-response. Lastly the questions investigated whether the industry regarded the statistics generated as false and hence see no reason of participating in such surveys or censuses. The industry was also given an opportunity to write other reasons which led to the industry not participating in the surveys and censuses but not mentioned in the response codes given. All the questions asked during the survey are shown in Table 3.
4. Conclusion
Summary of Key Findings
From the above distribution as shown in Figure 1, it shows that 15.1 percent of
Table 3. Results of the survey.
Question Code |
Reasons |
Frequency |
A. |
Statistical Offices and Research companies will give the data to third parties. |
53 |
B. |
The industry fears the data will not be used to the advantage of the industry |
41 |
C. |
The Government will intervene in the industry’s operations. |
68 |
D. |
Lack of trust. |
32 |
E. |
Statistical Office and Research companies does not produce correct and accurate statistics from the given data and hence it is a waste of time giving them data. |
60 |
F. |
The data will be manipulated by Government officials to suit the interest of the Government. |
66 |
G. |
The produced statistics does not help the industry. |
32 |
H |
Statistical Offices and Research companies’ staff do not explain well the instructions to complete the questionnaires. |
36 |
I |
The dressing of the Statistical Offices and Research companies’ officials do not warrant being given industry data as they dress like criminals or ordinary people on the street. |
44 |
J |
Time being given to complete the questionnaires is short |
18 |
|
Total |
450 |
![]()
Figure 1. Percentage distribution of types of reasons.
the total 450 sampled establishment have shown that they feared that the government will intervene in the industry’s operations if data is submitted to statistical agencies and producers of statistics. About 14.7 percent of the respondents feared that data will be manipulated by government officials to suit the interest of the State. The above two statements have shown that, there is need for producers of statistics to engage the industry if meaningful statistics is to be gathered. In addition to the above, 13.3 percent of the respondents have indicated that statistical offices and research companies do not produce correct and accurate statistics from the given data and hence it was a waste of time to share data. With this again, there is a need for producers of statistics to engage the industry to alleviate the fears in some industrial captains when dealing with producers of statistics.
In conclusion, about 64.7 percent of the sampled companies fears sharing the data with producers of statistics in Zimbabwe. Therefore, producers of statistics in Zimbabwe must deal with the issue of fear reported by data suppliers in Zimbabwe.
Whilst it is noble to engage data suppliers, producers of statistics also need to improve in the way they collect establishment data in Zimbabwe. From the survey, it was revealed that about 9.8 percent of the respondents feel that the dress code of data collectors needs to improve.
5. Recommendations to Producers of Statistics
From the research, it is recommended that producers of statistics in Southern Africa must recruit suitable personnel with the required skills for data collection for industrial surveys and censuses. The personnel should explain adequately the objectives of the surveys or censuses to industrialists. This is so because about 64% of the responses point out the fact that little is known by industrialists about the objectives of surveys and censuses and hence they fear to share data.
Producers of statistics need to equip data collectors with the necessary tools required for data collection for industrial surveys and censuses. Enumerators need to dress formally when conducting these surveys and censuses. Data collectors should not use public transport or walk on foot when visiting industries.
In addition to the above, regular meetings with industrial captains and introductory meetings before any survey or census is necessary and should be done. Introductory meetings should be done prier to the conducting of each survey or census. At the end of the survey or census, report back meetings are necessary with captains of industry. Publicity which is done on press and other forms of media does not really solve the worries of the captains of industry concerning sharing their data.
It is also recommended that conducting a survey by phone call should only be done after introductory meetings are done with the industry.
In addition, producers of statistics need to gather statistics which are also relevant for the industry in addition to the information needed for policy makers. The industry should see the value of statistics in their production line.
6. Areas for Further Research
Although the industry has pointed out in this research that producers of statistics need to improve in a number of areas, there is still fear in industry. The fear to share data with producers of statistics. It is still not clear whether the fear is emanating from the weakness of producers of statistics in their methods of data collection, which are regarded as poor or is as a result of something which the industry could not disclose at the time of the research and which has nothing to do with the weakness of producers of statistics. From the responses given by industry, there is much fear if the industry is requested to share data with the government. There is also need for further research to find out what causes the fear of sharing data with the State. If the fear is that the State may manipulate the data, the data will be manipulated at whose favour and does the process disadvantage the industry? If the State is to intervene in the operations of the industries, does the industry perceive that it will lose at the end? These are some of the questions which need further research.
Acknowledgements
Special thanks go to my family for the support given during the time of the research and also to the companies who responded earnestly to questionnaires sent to them.
Conflicts of Interest
The author declares no conflicts of interest.