<p></p><p></p><p>In fact, many of the movements in both export-sales and import-sales ratios are short term in nature. Short term influences include (for example) cyclical factors operating at the macro-level and movements in the exchange rate. The first task is to therefore to extract from the available data the long-run cross industry pattern of trade. </p><p></p><p></p><p>If we assume that short-run influences on the pattern of trade specialisation reflect macro-economic influences – both domestic demand and foreign demand and the exchange rate, -  then these may be picked up by assuming that these act in a similar way across industries, so that we can augment the 80 fixed effects by a set of time dummies. A further modification allows for first order serial correlation among the residuals. </p><p></p><p></p><p>The idea is to distinguish between temporary and long run (permanent) differences between sectors. Your technique is very similar to just taking an average of the esport-sales ratio which may be fluctuating from year to year. The time dummies get rid of all common 'macro' influences. When 'world GDP' goes up all UK industries will tend to export more. The time dummy gets rid of this effect. When these kinds of short run influences including the AR1 process are removed, you are left with the 'long run' fixed effects. </p><p>首先是说export-sales and import-sales ratios 是有短期特征的（第一段）。然后我们假设短期影响在贸易分工方面也反映了宏观影响（第2段）。。。然后为什么又说剔出宏观的影响（第3段）</p><p></p><p></p><p>我真的不是特别理解,请大家帮帮我，謝謝!! </p>

[此贴子已经被作者于2009-2-25 22:59:16编辑过]