We provide new disaggregated data and stylized facts on firm dynamics of the U.S economy at establishment level by using a state-space method to transform Census yearly data of entry and exit from 1977 to 2013 into quarterly frequency. We select the most significant determinants of these variables by matching Census data with a new database by Federal Reserve Bank. These determinants are extrapolated by using an unobserved factor model whose loadings are estimated via Principal Component Analysis. Alternative sources and their data are also investigated and discussed. We find that (i) Entry is pro-cyclical, coincident and symmetric; (ii) Exit is lagging with a maximum positive correlation with RGDP at lag 5 and asymmetric along the business cycle; (iii) the standard macroeconometric models estimated on our disaggregated series support the recent theoretical literature.

我们通过使用状态空间方法将 1977 年至 2013 年的人口普查年度进入和退出数据转换为季度频率，提供有关美国经济在基层组织层面的公司动态的新分类数据和程式化事实。我们通过将人口普查数据与联邦储备银行的新数据库进行匹配来选择这些变量的最重要决定因素。这些决定因素是通过使用未观察到的因子模型外推的，该模型的载荷通过主成分分析进行估计。还对替代来源及其数据进行了调查和讨论。我们发现 (i) 入场是顺周期的、重合的和对称的；(ii) 退出滞后，与 RGDP 的最大正相关在滞后 5 处且沿商业周期不对称；