* African economies struggle on China, commodity prices
* Investors question quality, independence of statistics
* Ad-hoc, niche information could prove more useful
* Data standards could slip further on funding cuts
By Joe Brock
JOHANNESBURG, Oct 15 (Reuters) - A dip in the fortunes of many African economies has raised doubts about the accuracy of the heady statistics that lured hordes of investors during the "Africa Rising" boom years.
Sub-Saharan Africa has achieved annual growth of more than five percent over the last decade, and foreign investment has more than quadrupled over the same period, as a commodities boom and an increase in consumer spending drove bumper returns.
But a slowdown in China has depressed demand and global prices for the commodities exports that many African nations rely on, prompting the IMF and World Bank to slash economic growth forecasts this month.
This cold dose of reality has raised questions about Africa's ability to cope with weak prices for oil, minerals and farm produce that investors rarely asked during the good times.
Attention is particularly turning to the quality of national statistics and the usefulness of sweeping reports from the likes of development agencies and brokerages which suggested African economies were diversifying and would be resilient to commodity busts.
"Numbers are being scrutinised more closely now things aren't going so well," said Simon Freemantle, senior economist at Johannesburg-based Standard Bank. "When the data is in your favour, why question it?"
Although data collection in Africa has improved significantly in the last decade, statistics offices in most countries remain rundown and underfunded.
Economic growth, one of the most fundamental indicators, is very difficult to track in African countries with weak infrastructure and sprawling rural populations.
"We do our best but the reality is we often have to rely on estimates," one official at a statistics office in a large African country told Reuters.
An exception to the rule is South Africa. The continent's most developed economy has a sophisticated statistics system and supports regional data collection.
The independence of statistics has also been questioned as some studies are produced by institutions that invest in Africa, while many governments are involved in how data is handled.
Population numbers are often difficult to calculate accurately and can be controversial as they influence the distribution of state funds. In Nigeria, Africa's most populous nation, the population is officially evenly split between the largely-Muslim north and mostly-Christian south.
In Ethiopia, the economy tends to perform exactly as the government predicts.
"The manipulation of data is a huge issue and it might blow up in our faces," said Morten Jerven, author of "Poor Numbers: How We Are Misled by African Development Statistics and What to Do about It".
Even where data collection has improved it has stirred suspicion.
Oil exporter Nigeria recalculated the size of its economy last year, a much-needed exercise that almost doubled the size of its GDP and made it the largest in Africa, overtaking South Africa. Ghana performed a similar exercise in 2010, when its economic output officially jumped 60 percent.
The process, known as "rebasing", also changed many assumptions investors had about the countries. Ghana's economy wasn't so much driven by agriculture such as cocoa production; services were at the heart of growth. Nigeria's record on tax collection wasn't a plus point, it was woeful.
Statisticians say rebasing should be carried out every five years to take account of the changing nature of economies. Some African countries haven't done so for 25 years, and only when the exercise is finally carried out are the shortcomings of previous data exposed.
"The rebasing shows you knew nothing. It makes a mockery of everyone staring at their Excel spreadsheets," Jerven said.
There is a further risk that slower growth in many African economies may result in cuts to state funding for statistical agencies, and data collection could get worse.
Poor data not only makes investing riskier, it becomes more difficult to lift millions of people out of poverty.
"During tougher times investment in statistics is one of the first things to be scrapped, which could perpetuate the problem," Charles Leyaka Lufumpa, director of statistics at the African Development Bank, told Reuters.
Faced with contradictory and unreliable data, many investors rely on other indicators to gauge African markets, from the sales of cement, beer and mobile phone credit, to the growth in the number of bank accounts.
Researchers have even taken satellite images of settlements to see the number of corrugated-steel roofs or strength of light emissions, indicators of population and power consumption.
Some investors view the problems with African data as an opportunity to outwit competitors and bag some of the highest returns available in the world.
"A lot of suits sat in boardrooms in New York and London are sold Africa with glossy pamphlets filled with meaningless stats," a private equity investor in Africa told Reuters. "To me what is important is having on-the-ground knowledge and that's where the advantage lies if you know Africa." (Editing by David Stamp)