The researchers developed a new method to estimate data in short panels that have cross-sectional dependence and dynamic heterogeneity. Their method is better at reducing bias and error compared to existing methods. By applying this method to sector-level data on labor productivity and hours worked growth in several countries, they found that not considering cross-sectional dependence can lead to misleading results.