Old, incomplete, assumed data calls accuracy into question

On a road trip, current and clear map data is vital for reaching the destination or one might face the fate of the man who nearly married the wrong woman because Google Maps led him to the wrong address.

When it comes to planning the future development of a country, having a set of correct economic and social data is mandatory. The data provided by the Bangladesh Bureau of Statistics (BBS) is used for many kinds of development decisions but it turns out, the data from them is often outdated or incomplete, or worse even, made up and it can lead to confusion at different levels.

Take the case of the rule of thumb the BBS uses to gauge reduction of poverty. It assumes that a 1% uptick in the gross domestic product (GDP) creates 0.25% of new jobs, leading to a fall in poverty rate a bit too. The BBS applied this rule to gauge the reduction in the number of poor people between FY16 and FY19.

It claimed that around 50 lakh poor people came out of poverty during the period as per its assumption on the back of the steady economic growth.

Similarly, without conducting any survey, it claimed that 74 lakh new jobs were created in the first four years of the Seventh Five-Year Plan (FY16-FY20).

Going by the BBS's own rule of thumb, job growth would be less in the pandemic year that saw a 5.24% economic growth down from 8.15% in the previous year. But the official statistical agency could not come up with any data on job growth or loss since it did not run any survey on it.

The BBS is supposed to carry out the Labour Force Survey every three months to provide a snapshot of the country's labour market, but it has not done any such thing since FY17.

Improper surveys yield incorrect economic, health data: WB chief economist

Besides, the last time the state-owned statistical agency conducted the Household Income and Expenditure Survey (HIES) to find out the latest situation of various socio-economic indicators, including per capita poverty rate, per capita income, expenditure, savings, food intake and inequality was in FY16.

The result – the BBS cannot provide data on many socio-economic indicators. Some data it discloses are incomplete and conjectural.

Without conducting the HIES and the Labour Force Survey, the agency determined the number of employment opportunities and the poverty reduction rate, relying on some assumptions and its rules of thumb.

Many independent research agencies claim the number of pandemic-induced job losses is at least three crore, but the government does not accept those estimates.

Against this backdrop, economists have urged the BBS to determine the real number of jobs lost due to Covid-19 through a new survey.

"If a survey cannot be carried out for some reasons, we still can have a rough idea about the trend of employment and poverty based on data of the preceding two years. But we can no way gauge the exact number of job creation or the rate of poverty reduction just on assumption," said Mustafizur Rahman, economist and distinguished fellow at the Centre for Policy Dialogue.

The number of jobs might not go up even if production outside the labour-intensive industry increases. Again, growth may not always be interpreted as reducing poverty at the same rate, he pointed out.

Noting that there is not much match between reality and data based on poor estimates, he said it would be difficult to implement a big plan based on such data.

The latest HIES survey report was scheduled to be released in 2016, but it took an additional three years to finish the task.

While unveiling the final results of the survey in 2019 in a press conference, the agency said the per capita poverty rate came down to 24.3% in FY16 from 31.5% in FY10. At the same event, the BBS also claimed the rate further dropped to 21.8% in FY18.

That the two different poverty rate figures were reached by a single survey had created confusion that is yet to be dispelled.

According to the latest Labour Force Survey conducted in FY17 to highlight the country's job market situation, the unemployment rate was 4.2%, similar to the rate in the preceding year.

The BBS has not conducted the survey to assess the nature of employment and conditions of seasonal workless people in the last four years. Had the quarterly survey been conducted, Covid-19 effects on the labour market would have been known within a few months after the pandemic hit.

Seeking anonymity, a BBS official said poverty reduction rates in the following few years were estimated, considering a positive relationship between per capita income growth and poverty alleviation.

In between FY10 and FY16, poverty was alleviated at an 0.7% rate against a 1% increase in per capita income. The new rate had been calculated based on an assumption that economic growth would alleviate poverty at the same rate in the coming years.

As per the latest BBS data, per capita income marked a growth even in the pandemic period that saw a drastic fall in earnings of most people engaged in the informal sector. But the BBS did not say what impact it had on the poverty situation, although independent research revealed a massive rise in new poor.

In 1974, the then president Bangabandhu Sheikh Mujibur Rahman created the Bangladesh Bureau of Statistics by merging four statistical agencies of the erstwhile provincial and central governments – the Bureau of Statistics, the Bureau of Agricultural Statistics, the Agricultural Census Commission and the Population Census Commission.

The following year, the Statistics Division was formed under the planning ministry to look after the BBS. In 2002, the division was again transformed into a wing of the planning ministry. In 2010, the Statistics Division was reconstituted, which was converted into the Statistics and Informatics Division in 2013.

The BBS has a decentralised network for the implementation of its activities with its head office in Dhaka. At present, it has statistical offices located in each division and district and upazila. The agency's total manpower is 2,234. Besides, there are around 100 employees in the Statistics and Informatics Division.

In the last fiscal year, Tk665.94 crore was allocated for conducting regular activities and development work of the BBS. The government claims that if Tk100 is spent on the organisation, Tk99.40 plays a role in alleviating poverty.

No database of poor people created even in nine years

The National Household Database (NHD) project launched in 2013 to prepare an information registry of all people in the country on the basis of income to quickly identify potential beneficiaries of social safety net programmes and disaster relief activities has not been completed in the last nine years.

The Tk328 crore project, financed by the government and the World Bank, had its deadline extended three times and the cost escalated to Tk728 crore, but its financial progress has been only 20% until June 2021. The fourth extension now awaits approval.

The government had decided to give maternity allowance to six lakh women through a World Bank project with an assumption that the NHD would be ready in 2014. Later, the World Bank has slashed $50 million from the project, with beneficiaries still not identified. The number of beneficiaries picked by alternative means has come down to four lakh.

According to BBS sources, the Management Information System under the project has not been completed yet. Besides, there are various inconsistencies in information collected for the database.

It will not be possible to get the final data on the poverty rate in the next two years, as various BBS data suggest.

Officials at the BBS said the Household Income and Expenditure Survey project has made only 2.5% progress since the beginning of last year.

A tender was floated using the electronic government procurement (e-GP) platform for the purchase of computers under the project, but lack of any response forced a second tender, officials involved in the project said.

A field survey, which was supposed to be conducted last month, has not begun yet because of Covid-19. Yamin Chowdhury, secretary to the Statistics and Informatics Division, at a meeting said there is no progress owing to a lack of farsightedness of those concerned.

Deputy Secretary to the Statistics and Informatics Division Dr Dipankar Roy said there was some delay in starting work because of the pandemic. Besides, a population census is going on now, which will continue till 2022. After that, the new HIES report will be available.

Poverty facts get trimmed

The BBS had completely thrown away some crucial surveys such as the income of the poor, their spending, and their standard of living.

Since independence, the bureau had been regularly publishing a consumer price index for the middle-class people of the capital until the fiscal 1997-98.

The BBS also suspended the agri-price publication featuring the wholesale and retail rates of farm items in FY06 as the huge price gap between the two segments continues  to deprive the consumers and farmers.

The living expenses reports for the Narayanganj, Chattogram and Khulna workers also have been on hold since FY06. Reports featuring the consumer price index and actual wages of industrial workers also met the same fate.

Industrial surveys lag behind

The last economic census on industries and service sectors was carried out in 2013. In the last eight years, the country's GDP size tripled, thousands of business entities mushroomed, and many new sectors joined the economy; but all those developments apparently failed to convince the BBS to initiate a new census.

The last survey on the cottage industry was in 2011 while the ICT sector was surveyed two years later.

Survey reports require extra time

The final report of the 2017 agri-census is yet to see the light although it was supposed to be published in 2018. The initial report of the census was published last year – just before the survey project crossed the deadline. With the project period having been extended, the final report is still in the works.

At a press briefing, the Statistics and Informatics Division put the blame on the agri-wing, saying the wing could not provide the main crops data on time.

After conducting a survey on the manufacturing sector for three years, a one-page summary was published in 2019. The final report is yet to be published though two more years have already passed since then.

The Household Income and Expenditure Survey was conducted in 2016, while the final report – which is still being widely used in economic research – got published three years later in 2019.

Economist Prof Sayema Haque Bidisha said, "If the findings of a survey are revealed this year, we can forecast the next near-future based on the data. As such, every important sector should be surveyed at one-year intervals. Especially, surveys on income, spending and employment should be carried out regularly."

"Most of the vital censuses are not being carried out regularly. On top of that, if the final reports take three to four years to get published, they often are of no use," she added.

Old baselines contributing to data-mismatches

A number of non-governmental organisation surveys and researches suggest 10% of the Dhaka dwellers had to leave the capital for their ancestral villages during the pandemic owing to income erosion.

The rush to rural Bangladesh depressed the annual house rent spike. But the BBS house rent indices claimed housing in Dhaka has become 6.18% costlier in the last one year.

When contacted, BBS officials said they use a 47-year old baseline to gauge the house rent hike. The house rent baseline has not been updated since the fiscal 1973-74.

The bureau tailors the Quantum Index of Small Scale Manufacturing Industries with the fiscal 1995-96 baseline while the Building Materials Price Index in Bangladesh follows the fiscal 1998-99 baseline.

Inflation, GDP, Quantum Index of Industrial Production, and Price Indices of the Domestically Produced Industrial Goods are calculated with the 15-year-old baselines.

As time passed by, economists say many new sectors joined the economy and many sectors got lost. So, the older the base year of the calculation, the greater the inconsistency in data.

Statistics bureau officials said they are working to update FY16 as the base year for GDP growth. It was supposed to be completed within the current fiscal year, but the pandemic delayed the completion.

Recent surveys don't reflect any Covid shocks

The pandemic already dealt a mighty blow to the country's tourism and hospitality sector even before a 2019 survey report on the sector got published in July last year.

The report published during the pandemic used FY18 data – that completely ignored the virus fallout on the tourism sector.

Another BBS report published in July last year on Urban Economic Assessment said 21.25% of urban dwellers worry about food adequacy, while 20.64% cannot afford their preferred meals.

The report reflected on urban employment, food, housing, availability of potable water and sanitation, but failed to mention how the indicators got affected by the Covid-19 outbreak in Bangladesh.

Similarly, the Cost of Migration Survey Bangladesh-2020 did not take the virus fallout into account.

What is BBS doing on Covid fallout?

Since the Covid-19 outbreak in the country, the BBS has conducted two surveys in one and a half years to assess the pandemic shock on the economy but none of the final survey reports has been released yet.

In a summary of the "Covid-19 Bangladesh: Perception Survey over Impacts on Livelihood" based on telephone feedback from more than 1,000 mobile phone users,  the BBS reported last October that all of the workforces, except day labourers, had returned to work by September after the first wave of the pandemic.

But, there are some inconsistencies in the summary of the unemployment rate, income fall and expenditure during the pandemic. It said the unemployment rate was only 2.3% in March last year.

But according to the Labour Force Survey, the unemployment rate was hovering around 4.2% from 2010 to 2017.

In the summary, the BBS said household-level spending fell by only 6.14% during the pandemic although the household income fell by 20.24%.

The BBS conducted another survey on the health sector but is yet to release the report. However, citing the report at a workshop, the secretary of the Statistics and Informatics Division told reporters that the number of fatalities due to cardiac diseases is many times higher than the Covid-19 deaths.

He advised the authorities to think about other health issues alongside the virus infection.

Confusion over GDP growth data

The BBS did not publish the complete GDP growth data in the last two years – leading to confusion over the pandemic shock on Bangladesh economy.

Without considering the movement curbs-led economic crises, the BBS claimed Bangladesh clocked 5.24% GDP growth in the last fiscal year. But the World Bank claimed the growth to be 2.4% only.

The exact growth figure in FY20 still remains elusive though one more year has passed.

The World Bank projects Bangladesh's GDP growth at 3.6% in the current year, while estimates of other development partners are also almost the same. But the government says the growth will be more than 6%.

In the last week of May every year, the BBS publishes an initial growth estimate for the current fiscal year. The final report comes within the next three months at the end of the fiscal year. Breaking with the trend, the final growth report of the last year and the initial growth projection for the current fiscal are being delayed.

Against such a backdrop, economists say the target of more than 7% growth in the next fiscal year is quite irrational.

Dr Zahid Hussain, the former lead economist of the World Bank, said, "The GDP size in the next year would be relatively small if the World Bank growth projections in the last year and current year are taken into account. On the other hand, the GDP size will be bigger if we consider the government target.

"It is important to determine the GDP size on which the next year's growth target will be based," he added.

The confusion over the GDP size has been exacerbated by data released on GDP as per market prices, national income and several other indicators although the actual GDP size has not been ascertained in the last two fiscal years.

According to a report by the BBS presented at the cabinet meeting last month, the final size of GDP stood at Tk27,39,300 crore as per the market value in the last fiscal — that is Tk57,078 crore or 2.04% less than the initial estimate.

If the GDP size excluding inflation falls by 2%, so will the real GDP by more than 1%.

Economist Ahsan H Mansur said it is not difficult to find out the actual GDP after determining the size of GDP in market price.

He said it should be expedited for the sake of transparency and economic policy formulation.

When contacted, Planning Minister MA Mannan said there have always been questions from various quarters about the accuracy of official data. But the government, particularly the planning ministry, has to rely on BBS data, he added.

"The BBS lacks some skills and capabilities, which is nothing new," he said, adding that the government seeks support from the International Monetary Fund (IMF), Statistics Division of the United Nations and developed countries to strengthen the BBS.

The government has also planned to revamp the BBS to ensure data accuracy.

Responding to a question regarding the new poor, he said there are some inconsistencies in independent researches because of flawed methodologies. The actual number of the poor will be gauged through the next survey of the BBS.

"We have not suspended any survey in our tenures. Conducting any survey or stopping it is totally up to the BBS. We do not interfere in its work. We will ask it to reinitiate essential surveys if there are any," said Mannan.