It seems that firms with very low productivity (relative to others in their industry) can survive in NZ. Aaron picked that factual gem from Roger's(pdf) analysis, as Eric noted.
I agree that weak competition is a prime suspect in this puzzle but I wonder if there are also some measurement issues. 2 things.
- productivity is measured as the ratio of the value of output to the volume of input. There are lots of things that could affect this ratio
- we're puzzled about low values of this ratio - firms with low output value compared to input volume.
The denominator is just hours of human work whereas in the real world labour quality varies. It turns out(pdf) that using the wage bill as a proxy for labour quality explains some but not all of the puzzle.
Another factor could be tax evasion. Suppose a firm is under-declaring revenue for tax purposes. On the productivity measures, it would disclose a high volume of inputs relative to its declared revenue or profit.
Tax evasion is easiest in sectors with direct personal contact between buyer and seller, so its interesting to note (pdf) that the services sector has "greater productivity dispersion" than manufacturing or agriculture where scope for evasion is lower.