Hello,
I am using fractional logit model to understand the impact of employer power on involuntary part-time employment which is a proportion of employed civilian. I am adding occupational (six-digit SOC), geographic (Commuting Zone - CZ), and time (Year-quarter) fixed effects to the model. I created a variable called "marketid" by grouping SOC and CZ. A total of 28,993 marketid was created. When I run the code below, the results show that "characteristics contents too long.The maximum value of the content is 67,784." I think it's because of the added fixed effects of marketid. How can I increase the limit for fixed effects?
When I add SOC and CZ fixed effects separately, (i.e., i.cz i.soc instead of i.marketid), the model runs successfully.
Also, if you know any books that I can learn about fractional logit model, please let me know!! I would greatly appreciate it.
Thank you!
I am using fractional logit model to understand the impact of employer power on involuntary part-time employment which is a proportion of employed civilian. I am adding occupational (six-digit SOC), geographic (Commuting Zone - CZ), and time (Year-quarter) fixed effects to the model. I created a variable called "marketid" by grouping SOC and CZ. A total of 28,993 marketid was created. When I run the code below, the results show that "characteristics contents too long.The maximum value of the content is 67,784." I think it's because of the added fixed effects of marketid. How can I increase the limit for fixed effects?
When I add SOC and CZ fixed effects separately, (i.e., i.cz i.soc instead of i.marketid), the model runs successfully.
Code:
xi: fracreg logit unemp hhi_log total_vacancies1_log demo_cat* i.marketid i.yq[iw=vacancy_weight], vce(cluster marketid)
Also, if you know any books that I can learn about fractional logit model, please let me know!! I would greatly appreciate it.
Thank you!
