We wondered whether this discrepancy was caused by our data showing a smaller proportion of dairy farmers being sharemilkers than the Dairy Board’s records. Because they tend to own more land than sharemilkers, farm owners tend to have higher debt. According to the Dairy Board, there are 14 673 dairy farms in New Zealand, of which 5238 are sharemilkers. Because our data show a higher proportion of farm owners (85.6 percent) than the Dairy Board records (64.3 percent), we would expect our calculated liabilities to be greater than the Dairy Board’s. We recalculated average dairy-farm debt for New Zealand, using weighted liabilities for sharemilkers and dairy farm owners. Using the Dairy Board data for total liabilities and the proportion of farms in each ownership category, the average dairy-farm debt is $400,000; using our liabilities per farm and Dairy Board data for the proportion of farms, the figure is $435,000. This reduces the difference between the two estimates, to the point where the Dairy Board average is only 8 percent lower than ours.

There are several other possible reasons for the differences between the Dairy Board data and ours. The Dairy Board used a different sampling strategy from our survey. They derived their sample of farms from Livestock Improvement Corporation lists. They sampled only factory supply farms. Also, they removed farms with fewer than 30 dairy cows, those with less than 50 percent of their gross farm income from dairying, and those with no separate accounts for their dairy enterprise. This would give them a precise average of a "pure" dairy farm, one that concentrated on dairying. For the purposes of our survey, however, New Zealand farming is often not very "pure", and many farms have multiple enterprises. By including such farms we have likely increased accuracy, but at a cost of precision. The Dairy Board was more rigorous than we were in removing liabilities from farmers’ accounts that they considered to be private items, such as cars and extra houses. We accepted the liability figures in the farm accounts; the Dairy Board removed items they considered private. This would decrease the average liabilities in the Dairy Board data, compared with the averages in our data.

The total liabilities for dairy farms supplied by one accounting firm is very similar to our data ($530,224 compared to $502,488); the data on term liabilities supplied by the other accounting firm is for a term debt in excess of our survey figure for total liabilities per dairy farm. Both accounting firms are based in Canterbury, and their figures may not reflect national averages.

We are not aware of any other reliable source of debt data for the horticulture sector.

For cropping, the data provided by the two accounting firms are similar to the data for our survey, although their data show slightly lower debt levels than ours (Table 29). Both accounting firms are based in Canterbury, and their figures may not reflect national averages.

Table 29 Comparison of horticulture and cropping farm average liabilities from our survey with data from other sources