NEW YORK • Art institutions in Britain have faced declining attendance in recent years and museum officials met last year to discuss the issue.
But when a steep drop-off in visitors to the National Portrait Gallery in London was included in its report to Parliament last month, people were alarmed - it was a decrease of more than 40 per cent.
It turns out it was the result of a counting error by Ipsos Retail Performance, the company that tracks visitors to the museum and other establishments.
The number of visitors in the 2017 to 2018 fiscal year, which ended on April 1, had been falsely reported as having dropped to about 1.1 million, from about 1.9 million. The real number of visitors for 2017 to 2018 was nearly 1.7 million, the museum said on Wednesday.
That represented a drop of about 10 per cent from the previous year, not the more alarming decline based on the false information.
The report that there was a more than 40 per cent drop in visitors caused consternation at the museum, The Art Newspaper reported this week. The falloff could make it difficult to attract donations for renovations.
Ipsos, which has been working with the Portrait Gallery for 17 years, told The New York Times in a statement on Wednesday that the miscount at the National Portrait Gallery was believed to have been an isolated incident and that an audit had been completed for the other public spaces for which it provides visitor counting.
The error resulted from a faulty sensor over the main entrance that was initially detected in the spring of last year, a museum representative said.
At that time, an engineer was sent to repair the device, but the device later failed a routine accuracy test in July last year.
But Ipsos said that it passed the test and the museum was never notified of the error.
The issue continued until last month, when the museum noticed a discrepancy between the Ipsos reporting for the main entrance and the Ipsos data for another location there.
The museum raised concerns about the accuracy of the counting more than once before last month, the representative said, but Ipsos assured the museum that the system was working.
The new, adjusted numbers are based on a recalculation by Ipsos.