I will pick up on a few points. I agree with David Bell with respect to trade data. We have three different measures that get at trade: the index of manufactured exports, export statistics Scotland and the HMRC data. They all tell us slightly different things, but the best among them is probably the export stats Scotland data.
From looking at the documentation that the Scottish Government has put out, I think that one of the challenges for that concerns the response rates of firms. Firms are not required to complete the information so, as a result, the sample sizes are not as comprehensive as they could be. There is an issue around the fact that particular types of firms tend not to answer them—larger firms or whatever.
In our submission to the committee, we outlined how moving towards a situation in which the Scottish Government can compel firms to complete the information, on the same basis as the Office for National Statistics, which has such a power just now, seems infinitely sensible. That could be one way to broaden the coverage and at the same time improve the sample size and the quality of the data.
It is important to recognise that Scotland is very well served with economic data relative to other parts of the UK, both in terms of quality but also as far as timeliness is concerned. This year, we got GDP data less than 100 days after the end of the quarter for the first time. Typically, we get that data in 110 to 115 days. We will not find out until December what happened to the economy of the north-east of England in 2016. Relative to other parts of the UK, we are pretty well served in terms of timely quality indicators.
There are things that I know the Scottish Government is considering that will drive improvements in some statistics. The Scottish Government is now getting better access to the monthly business survey, which it uses to produce the index of manufactured exports, and that may have spillover effects for GDP data, for instance.
There is a programme of continuous improvement, both on the ONS side and on the Scottish Government side, to try to produce better-quality, more timely data. It is worth stressing that.
I would say from an academic perspective that, when we are thinking about headline indicators, we need a long enough consistent time series. For instance, if someone is trying to forecast gross value added or most other macroeconomic indicators, they need a long enough time series on which to estimate that model. If we are investing in new data series or changing existing data series, we have to consider the compatibility of that over time and relative to other parts of the UK.