Skip to main content

Table 2 Unemployment decomposition

From: The importance of frequency in estimating labour market transition rates

  Monthly survey Quarterly survey
  Original Corrected Original Corrected
   Continuous   Monthly Weekly Continuous
  n m \({\hat {\lambda }_{m}}\) n q \({\hat {n}_{m}}\) \(\hat {n}_{w}^{\ast }\) \(\hat {\lambda }_{q}\)
E → U 0.312 0.301 0.412 0.332 0.293 0.247
E → I -0.001 -0.001 -0.002 -0.004 -0.004 0.001
U → E 0.335 0.340 0.277 0.387 0.442 0.392
U → I 0.105 0.107 0.073 0.109 0.125 0.106
I → E 0.086 0.087 0.062 0.059 0.055 0.097
I → U 0.164 0.166 0.177 0.117 0.090 0.158
  1. Note: n m and n q are calculated directly from CPS from 1976:2 to 2011:6. \(\boldsymbol {\hat {\lambda }_{m}}\) and \(\boldsymbol {\hat {\lambda }_{q}}\) are calculated using equations ( 3 ) and ( 4 ). Discrete weekly correction calculated \(\boldsymbol {\hat {n}_{w}= p_{q} \mu _{q}^{1/12} p_{q}^{-1}}\) . The unemployment decomposition is based on Shimer ( 2012 ). I report the \(\boldsymbol {\frac {Cov\left (\tilde {u}_{t,i},\tilde {u}^{AB}_{t,i}\right)}{Var\left (\tilde {u}_{t,i}\right)}}\) , normalised to add up to 1.