Page 138 - AJWEP-22-4
P. 138

Wang, et al.

                 Table 2. (Continued)
                 Explained variable                      Digital 1                                Digital 2
                                                 (1)                 (2)                  (3)                  (4)
                 Wet j, t                     0.674***             0.228**              0.021***            0.013***
                                               (0.086)             (0.101)              (0.003)              (0.004)
                 Wet                          0.278***              0.071               0.015***              0.005
                    j, t-1
                                               (0.090)             (0.099)              (0.003)              (0.004)
                 Sun j, t                     17.308***            6.300**              0.536***             0.307**
                                               (2.781)             (3.125)              (0.110)              (0.120)
                 Sun j, t-1                    4.881*               1.278               0.244**               0.116
                                               (2.566)             (2.727)              (0.096)              (0.101)
                 Industry-fixed effects          No                  Yes                  No                  Yes
                 Year-fixed effects              No                  Yes                  No                  Yes
                 Observations                  12,908               12,908               12,908              12,908
                 R-squared                      0.179               0.223                0.165                0.340
                 Notes: Standard errors in parentheses; *p<0.1, **p<0.05, ***p<0.01.

                 Table 3. Extreme temperatures and enterprise digital transformation: robustness test

                 Explained variable: Digital 2     (1)               (2)            (3)           (4)          (5)
                                             Replacing digital   Changing the     Relative      Delete      Variables
                                              transformation     measurement     threshold     Zhejiang      lagged
                 Extreme                          0.009*           0.018***       0.006**      0.026***     0.030***
                       j, t
                                                  (0.005)          (0.007)         (0.003)      (0.009)      (0.008)
                 Extreme                           Yes               Yes            Yes          Yes          Yes
                       j, t-1
                 Control variables                 Yes               Yes            Yes          Yes          Yes
                 Other climate variables           Yes               Yes            Yes          Yes          Yes
                 Industry-fixed effects            Yes               Yes            Yes          Yes          Yes
                 Year-fixed effects                Yes               Yes            Yes          Yes          Yes
                 R-squared                        0.248             0.313          0.248         0.248        0.328


                rate, etc., constructed the logarithmic form of the Cobb–  and  higher  TFP  will  weaken  the  effect  of  extreme
                Douglas production function to estimate the TFP of the   temperature  on digital  transformation.  We, therefore,
                enterprise,  and  tested  model  (II)  and  the  production   focus on the interactivity coefficient.
                function (III) as follows:
                                                                    lnY  = µ  + µ lnK  + µ lnL  + µ lnM  + ε     (III)
                                                                                1
                                                                                                 3
                                                                                    i,t
                                                                                         2
                                                                                             i,t
                                                                                                      i,t
                                                                                                          i,t
                                                                       i,t
                                                                            0
                Digital    ± Extreme   ±Extreme       M
                      ij t,,  0      j t,  1      j t, 1  2           Where  Y represents the output of the enterprise,
                  MExtreme      X    X     W    W          measured by the company’s main business income for

                                            '
                  3           j,tt  1  i t,  2  j t,  0  jt,  1  jt1,
                                                               (II)  the year; K represents the level of input of capital factors,
                   ,,
                  ij t                                              measured by the company’s investment in fixed assets
                  Where  M is the mechanism  variable,  including:   for the year;  L represents the level of input of labor
                employee  pay share (Wage ), executive  pay share   factors, measured by the company’s cash flow statement
                                          1
                (Wage ), cost growth rate (Cost), and TFP (TPF_OLS,   for the year, “cash paid to and for employees;” and M
                     2
                TPF_LP). If the hypothesis is valid, then high employee   represents the enterprise’s intermediate  products and
                salary share and high cost growth rate will strengthen the   raw material  input, measured by the company’s cash
                effect of extreme temperature on digital transformation;   flow statement for the year, “cash paid for purchasing
                on the  contrary, high  executive  compensation  share   goods and receiving labor.”
                Volume 22 Issue 4 (2025)                       130                           doi: 10.36922/AJWEP025210166
   133   134   135   136   137   138   139   140   141   142   143