Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=

# Cinderella No More: Night Shift Bans and Women’s Employment in India

Authors: Anshika Arora, Sarvnipun Kaur, Prakarsh Singh  
Discussant: None  
Video: https://www.youtube.com/watch?v=0mE9GNhwdY8&t=9142s


## Talk (02:32:22 – 02:47:57)

[02:32:22] will not conclude but I think that's it. [applause] >> Thanks Lucas. Um, our next two presenters are uh, excuse me,  
[02:32:37] [clears throat] going to be presenting uh, shorter 15minute egg timer presentations. So, unfortunately, hold your burning questions until lunchtime.  
[02:32:46] Um, we're very happy to have uh, Sarvnapun here to talk uh, about a policy that I know a number of people in the room are interested in learning more about the night shift ban on women's employment in India.  
[02:33:09] Um, all right. I'm going to start by thanking the organizers for having me here, but also for organizing this session for the past three years. Every year I go back to the schedule of the  
[02:33:22] gender in the economy session to look at what's new in the in the field and I feel so lucky to be in this room today.  
[02:33:30] Um I'm going to be presenting joint work with Anika Aurora and Prakash Singh and we essentially try to draw out lessons from this deregulation of the night shift ban exercise that happens um in  
[02:33:43] India. A little context about uh night shift restrictions. uh these are bans or prohibitions on the employment of women in the during the evening shifts um such that they're not employed on the same  
[02:33:58] terms as men. Um no surprises here. Um this isn't a problem specific to India.  
[02:34:05] uh close to 81 economies of 190 documented by the women business and law report um encoded uh night shift restrictions in some form or the other in their laws in the 1970s. Um by 2024  
[02:34:20] close to two 20 economies continued to house those restrictions. Um India was uh one of them. Um why we think this is important is because we have room to  
[02:34:32] grow. um our uh Indian employment to population ratio is significantly below the world average and this coupled with the fact that we have a very large female population uh just puts into  
[02:34:46] perspective uh what increases in employment outcomes can do uh both for India and for the world.  
[02:34:54] Um I'm going to give you some context about the legal system or labor regulation system in India. Um so we have different regulations kind of uh governing different sectors in India. So  
[02:35:08] the manufacturing sector is governed by the factories act. The commercial agriculture sector is governed by um the plantation workers act. Agency work is governed by the contract labor act. Each of them in some form or the other encode  
[02:35:22] a night shift restriction um in their in in their provisions. We are going to double down on an act called the shops and establishments act which looks at the uh which which essentially governs  
[02:35:36] services sector in India. Um this is a colonial era regulation and uh it was introduced at the time to regulate uh small uh shops, eeries, theaters etc.  
[02:35:48] But as India has come to grow, it has come to cover uh large multinational corporations, IT sector establishments, um hotel chains, uh the works. uh now  
[02:36:02] then the shops and establishments act uh also has a night shift restriction encoded into its provisions and we model this night shift prohibition as a kind of a hard constraint on women's  
[02:36:16] employment in India and we believe that there are gainful employment opportunities lost because of this restriction uh both for the employers and the uh prospective female employees uh different Indian states have started  
[02:36:31] to deregulate uh this night shift prohibition. Um what I mean by deregulation is that uh there are two forms in which this is deregulated. The first of which is rules-based. A firm is  
[02:36:46] told to comply with X number of regulations such as provision of safety um uh uh services uh presence of a security guard in the on the premises, provision of cab services and then  
[02:36:59] they're allowed to employ women in the night. There is another form of deregulation where a firm has to actively go and seek um uh permission from a civil servant in order to employ women. We are going to be focusing on  
[02:37:13] the first one but I will eventually show you um how effects kind of differ by the way this law was deregulated as well. Um one sort of disclaimer before um before  
[02:37:26] I start is that um we use the periodic labor force survey for our analysis. The periodic labor force survey was initiated in the year 2017. There are three states which actually deregulated  
[02:37:40] night shifts but they've been excluded from our analysis because they did it in 2017. So we did not have a large enough prep period to include them in the analysis. Um all the other states are um  
[02:37:52] essentially our control group. And then um we exclude union territories for uh purposes of simplicity and um there are two states uh Tamil Nadu and Goa which never had a night shift restriction to  
[02:38:07] start with. So these were never treated states. These are also excluded uh from the analysis. Now we believe that um this deregulation exercise the effects  
[02:38:19] of it are ambiguous uh because even though um there there are gains that employers would experience in terms of greater scheduleuling flexibility on women's arts uh but at the same time they have to comply with a set of  
[02:38:33] regulations and these could impose uh compliance costs. So we would see an increase in service sector employment only and only if the shed the gains from theuling flexibility  
[02:38:47] um for an employer by employing a woman exceed um the costs that they have to incur in order to be able to uh introduce night shift um in their in their working schedule. So this  
[02:39:01] motivates three questions for us. Uh the first is if deregulation coupled with compliance cost would actually improve uh women's employment outcomes. Um the second is if government can actually  
[02:39:15] safely outsource um women's safety without compromising on employment outcomes of women. And then uh finally if the legal architecture of how this deregulation  
[02:39:28] takes place will have an will show up in terms of economic effect or not.  
[02:39:35] [snorts] Um we use the periodic labor force survey. Um we use 5 years of of this data and uh to kind of encode our uh legal amendments we used we mined  
[02:39:47] state gazettes and kind of triangulated uh our findings with another report which had systematically documented instances of um legal discrimination or  
[02:40:00] or cases where men and women weren't treated equally by by law in the labor laws.  
[02:40:07] uh we use two identification strategies. Our baseline identification strategy is a simple um linear probability model. We calculate two-way fixed effects. Um we control for uh ear fixed effects and  
[02:40:21] state fixed effects. And uh we also try to control for individual attributes like the age of the woman, the religion, the marital status. Um, finally, it's also possible that some of our results may have been driven by an increase in  
[02:40:36] service sector demand in the treatment states. And to kind of account for that, we control for state time uh varying uh things like state um service sector GVA  
[02:40:49] and u netstate domestic product. um the law was deregulated um in different states at different  
[02:41:00] times and so um we kind of we use the calves antenna estimates to ensure that uh we we've debiased the coefficients  
[02:41:11] and we don't uh we devate it for um late versus early regressions. Uh we aren't the first people to do this. Surprise surprise. uh there is a host of literature some of these are the are the  
[02:41:24] latest kind of uh findings most of which have found that when uh women's some of which have found that when women's uh labor hours are regulated uh it it's detrimental to their employment  
[02:41:39] outcomes. Some have found that actually um regulation of working hours of women improve um women's employment outcomes including this uh 2024 Hadad and Kadan Katan study which actually found out  
[02:41:54] that when night shifts were restricted in the United States uh women perceived them as welfare enhancing and actually um started to work more. So we use that as our starting point and look at if uh  
[02:42:08] introduction of a night shift deregulation would uh reduce women's labor force participation rate in some form or the other. Uh we actually don't find um that's the case. We also don't  
[02:42:21] find that uh night shift deregulation increases aggregate employment of women which means that it's not that unemployed women um started to find  
[02:42:33] employment in the labor market. What we do actually find is that women move from nonservices sector to services sector and particularly they move from  
[02:42:48] self-employment and non-services sector which as my um as the previous uh presenter here also mentioned is is relatively informal. So they move from uh self-employment and non-services to  
[02:43:03] wage employment and services where they're provided formal contracts, social security etc. So it's more white collared work in the services sector from um um self-employment and non-services  
[02:43:18] sector. The second thing uh that we kind of find out is that um the effect of this deregulation is not consistent across the board. It heavily depends on  
[02:43:30] the safety context of women. So um if the government kind of asks a private employer to uh provide a cab service or a um or a guard security guard, it's is  
[02:43:44] that enough when there is not enough public safety in in the woman's environment? And we actually find out that that is not the case. the effect in relatively safer uh districts where  
[02:43:58] which had uh lower crimes against women reported uh is much higher uh than the effect that you would find in a in a relatively unsafe environment for women.  
[02:44:11] So um you know private enterprise can provide a cab and a guard but if uh the streets are not safe it still may be the case that actually women don't find uh  
[02:44:25] um employment opportunities lucrative enough to start joining the uh joining gainful employment. [snorts] Finally, um  
[02:44:35] I also uh look at um if the regulatory design has a role to play in how um employment outcomes are kind of affected. Uh so I look at uh whether or  
[02:44:50] not rules-based versus discretionbased uh deregulation would affect employment outcomes. Uh this means that if you require an employer to uh employ women by complying  
[02:45:04] with X number of regulations, would that be different from if states required employer to uh apply for a um apply for a license in order to be able to employ  
[02:45:18] women in the night shifts. And actually we find out that uh when the deregulation is rules-based where an employer needs to meet x number of conditions in order to be able to employ  
[02:45:30] women um the effect is much larger and it is attenuated if the regulation is um um uh is essentially  
[02:45:44] not rules based but actually discretion uh based. Uh now uh to conclude and I'm way early so maybe I can talk a little bit about uh some of the other things  
[02:45:57] that we try to do here. Um we find out that actually the reform did not expand employment opportunities for women but it did shift the composition of existing uh employment. Uh those women who were  
[02:46:11] in the reforming states they became much more likely to work in the services sector. Um we believe that according to our framework this would mean that the compliance costs that the deregulation  
[02:46:24] imposed were lower than the gains that um um firm owners would have experienced because of uh greateruling flexibility over women's working hours. Um the wage  
[02:46:37] employment and services increased and self-employment and non-services declined. And then um this wasn't uniform across the board but it depended heavily on the safety context of women.  
[02:46:49] Uh regulatory design had a role to play in uh the size of the effect of the deregulation. One thing that I would also like to um state here now that I have time is that uh there are since I  
[02:47:03] mentioned that there are different types of regulations that um that regulate different sectors and there is a factories act that regulates the manufacturing sector. There is a shops and establishments act that regulates  
[02:47:17] the services sector. Simultaneously as the services sector was getting deregulated, factories act was also getting um deregulated at the s same time and this may have uh had an effect  
[02:47:32] on on the results that that we we saw. So we control for that and we still uh see that the results uh hold. Um  
[02:47:45] that's about it. uh I'm happy to have any conversations or um discussions in and I'm I'm really grateful for the opportunity. Thank you.