A Formula for Future Growth

As explained, my research team and I built the formula for future growth from the top down. We started with my concepts of how productivity and indebtedness affect growth, then fleshed these forces out with specific indicators, and then saw how the formula created this way worked. I followed this approach because I believe that one should be able to describe the cause-effect relationships and the logic behind them without looking at the data and that only after doing that should one look at the data to see how well the descriptions square with what happened because otherwise one would be inclined to be blinded by data and not force oneself to objectively test one's understanding of the cause-effect relationships.

如上所述，我的研究团队和自上而下建立了未来增长的公式。我们开始了我的生产力和负债如何影响增长的概念，然后用特定的指标来强化这些力量，然后看到公式如何发挥作用。我遵循这种方法，因为我认为应该能够描述因果关系和背后的逻辑，而不用查看数据，只有在这样做之后，才能看到数据，看看这些描述与什么相符发生了，因为否则会倾向于被数据蒙蔽，而不是强迫自己客观地测试人们对因果关系的理解。

Below is a list of what I have come to learn about these things along with the names of the indices my research team and I created to reflect them. Based on the reasons outlined there, we created 1) a simple logic-weighted index of productivity and 2) a simple logic-weighted index of indebtedness. We used the same set of factors weighed the same way for each gauge across all the countries and across all timeframes. That way there was no fitting the data and our measures for productivity and indebtedness are timeless and universal. We put two- thirds of the weight on productivity and a third on indebtedness.380 After creating these indices, we observed how each predicted the subsequent 10 years’ growth rates for each country (which we measure every 5 years). In other words, we observed rather than fit the data. The table below shows the concepts, their weights, and their correlations with the next 10 years’ per capita growth rates for our universe of 20 countries. Together these indicators were 84% correlated with the countries’ subsequent growth rates. Below we show how well these measures related to future growth across countries and time.381

如上所述，据我所知，约有三分之二的国家的10年增长率将由于生产力而大约三分之一将由于债务。下面的视觉传达了这两个力量。我们的生产力指标旨在衡量生产力增长率随着时间的推移如何陡峭，我们的负债措施旨在衡量债务周期如何影响中期的增长。

These measures of productivity and indebtedness can be used to predict each country’s absolute and relative growth rates over the next ten years, or longer periods. They also can be used by policy makers to indicate what levers they can move to influence future growth. To reiterate, my goal is to get the big picture right—i.e., to reliably be approximately right by focusing on the most important drivers rather than to try to be precise by focusing on the details.

这些生产率和负债的衡量标准可用于预测未来十年或更长时间内各国的绝对和相对增长率。决策者也可以使用它们来表明他们可以采取什么杠杆来影响未来的增长。要重申一下，我的目标就是要把握好大局，即要把重点放在最重要的驱动因素上，而不是通过专注于细节来精确地来确定大致是正确的。

Before looking at the picture we will show you how our aggregate indicator would have predicted growth versus what actually occurred. While staring at the observations helps us ground ourselves in reality and test our logic, we know there is no precision in the specific numbers and what matters most to us is whether our logic is strong. Our examination covers 159 separate observations across 20 different countries over the last 65 years, which provides a wide range of different environments to test our indicator. Along with the correlation of our predictions and what growth actually materialized (shown below), another test is how reliably we predicted something reasonably close to what happened. In our set, our aggregate predictions for a country’s average growth over the next decade were within 1% of the actual about half of the time, and within 2% around 80% of the time.

在查看图片之前，我们将向您展示我们的总体指标如何预测增长与实际发生的相关。在盯着观察结果的同时，我们可以将自己置于现实状态，并测试我们的逻辑，我们知道具体数字没有精确度，而我们最重要的是我们的逻辑是否强大。我们的考试覆盖了过去65年来的20个不同国家的159个独立观察结果，其中提供了广泛的不同环境来测试我们的指标。随着我们的预测与实际实现的增长（如下所示）的相关性，另一个测试是我们预测与发生的事情相当接近的可靠性。在我们的集合中，我们对一个国家在未来十年的平均增长的总体预测在实际的大约一半的时间内不到1％，在80％的时间内在2％以内。

Below we show the same perspective for each of our productivity and indebtedness gauges, comparing what they implied individually for a country’s growth versus what happened. As you can see our measure of productivity is more strongly correlated with each country’s growth than our indebtedness measure is (64% vs. 44%), which makes sense given it is the more important driver over the timeframes tested. Still, each has a fairly good relationship on its own.

下面我们对每个生产率和负债量表都展示了相同的观点，比较了他们对一个国家的增长与发生的事情单独所暗示的含义。正如你可以看到，我们的生产率测度与每个国家的增长率都高于我们的负债量度（64％vs. 44％），这是有道理的，因为它是测试时间框架中更重要的驱动因素。不过，每个人都有一个很好的关系。

Because these are timeless and universal drivers, we expect them to be just as important in developed countries as they are in emerging ones. The type of investment or education that matters may shift, but ultimately whether a country sees productivity growth is still going to be largely a function of the basic building blocks of productivity----whether its workers offer value, whether it is investing in its culture and creating a culture of success----as well as how its indebtedness is evolving. Across the countries we have examined, our aggregate indicator is about as correlated with future growth for developed and emerging countries (69% correlated with the growth in income per worker in developed countries and 81% correlated in emerging countries). Of course, which countries are ‘‘developed’’ or ‘‘emerging’’ changes over very long periods as discussed in ‘‘The Rises and Declines of Economies Over the Last 500 Years.’’ So in the tests shown below, we adjust for that, for example excluding Japan in the 1960s when it was much more like an emerging country.

因为这些是永恒的和普遍的驱动因素，我们期望他们在发达国家和新兴国家一样重要。重要的投资或教育类型可能会发生转变，但最终一个国家是否看到生产力的增长仍将主要是生产力的基本要素的一个功能 - 无论是工人是否提供价值，是否投资于其文化和创造成功的文化以及债务如何演变。在我们审查的国家中，我们的总体指标与发达国家和新兴国家的未来增长率相关（69％与发达国家工人收入增长相关，新兴国家81％相关）。当然，哪些国家在“过去五百年经济崛起和下降”中讨论过的很长一段时期“发展”或“新兴”的变化，所以在下面的测试中，我们调整对于这一点，例如在20世纪60年代不包括日本，它更像是一个新兴国家。

To reiterate, I believe getting to this fundamental level is critical to understanding and predicting the growth of countries. Naïve measures of a country’s future growth, for example just income on its own or a country’s trailing growth, won’t get you much because they won’t help you get at the drivers. They also tend to be much worse predictors than the formula I have described here (about 25% as good by traditional statistical measures). Looking at the economy as a machine and granularly measuring the cause-effect relationships makes all the difference.

要重申，我认为达到这个基本水平对于了解和预测国家的增长至关重要。一个国家未来增长的初步措施，例如只是收入本身或一个国家的拖尾增长，不会让你感到太多，因为他们不会帮助你得到司机。它们的倾向比我在这里描述的公式（通过传统的统计学措施好约25％）更倾向于预测。将经济看作一台机器，并且细粒度地测量因果关系造成了所有的差异。