Episode Transcript
[00:00:00] Speaker A: Imagine a justice system built on rigorous evidence, not gut instincts or educated guesses about what works and what doesn't.
More people could access the civil justice they deserve.
The criminal justice system could be smaller, more effective and more humane.
The Access to Justice Lab here at Harvard Law School is producing that needed evidence. And this podcast is about the challenge of transforming law into an evidence based field. Jim I'm your host, Jim Greiner and this is Proof Over Precedent.
Welcome to another edition of Proof Over Precedent, the Access to Justice Lab's podcast.
My name is Jim Greiner and I'm the faculty director of the Access to Justice Lab.
And today we have part two of the podcast that we are recording on the Access to Justice Lab's project on Expungement, Randomized Control Trial on Expungement in Kansas.
The technical title is the Final Stage Reentry Project, referring to reentry from a period of incarceration or a period of criminal justice involvement. And the final stage would be potentially clearing criminal justice records. And so when we say expungement, we mean the, what we mean is the suppression of records. Public view, particularly as we'll discuss employers and potential landlords.
And in part one of the of our podcast series, which you, you may have listened to already, we discussed the logistics of the randomized control trial, the RCT and the partnership that Kansas Legal Services was kind enough to create with the Access to Justice Lab to implement this study in Kansas.
And today what we're going to be discussing is the results of the RCT as they relate to employment outcomes. Because as we'll talk, as we'll discuss that is one of the primary rationales that folks who are seeking expungement cite for seeking expungement. This is one of the things that they want to accomplish is improved employment and improved income.
Today let me, let's get an introduction of who's joining us today.
Marilyn, can you briefly introduce yourself again for folks who may not have caught part one?
[00:02:13] Speaker B: Sure. I am Marilyn Harp, the retired executive director of Kansas Legal Services.
[00:02:19] Speaker A: And you continue to be involved in criminal justice record clearing work in Kansas, is that correct?
[00:02:24] Speaker B: I do. I have the ability now that I didn't have in my prior job to do advocacy at the legislative level. So I work on some things there as well as a volunteer doing a lot of expungements for people.
[00:02:39] Speaker A: Terrific. Renee, who are you?
[00:02:42] Speaker C: I am the non retired researcher at the Access to Peace to slide.
[00:02:48] Speaker A: Terrific. And you were the frontline researcher and the person most involved on the day to Day operations.
[00:02:54] Speaker D: Correct.
[00:02:54] Speaker A: Ryan, who are you?
[00:02:56] Speaker D: I am a data analyst here at the lab and I did a lot of the number counting and data cleaning and processing for this project.
[00:03:05] Speaker A: Terrific. And Patricia, who are you?
[00:03:08] Speaker E: I'm Patricia and I'm also a data analyst at the lab.
[00:03:11] Speaker A: Terrific. Okay. And so we're going to be discussing the findings from a report that we will be turning into academic papers to submit for publication.
The report will be eventually posted on the National Institute of Justice's website. We want to acknowledge that this project was funded by the National Institute of Justice, the Charles Koch foundation, and the Chan Zuckerberg Institute.
The CZI is no longer with us. It's moved on to other, other forms and other things. So. But we are grateful for the funder, all three of our funders. So let's begin. Let's set the stage here a little bit. Marilyn, when clients come to you and say, I'd like your help for criminal justice, this record clearing, what are things, what are reasons that they typically give you for seeking an expungement?
[00:03:58] Speaker B: I think the biggest reason is that and the global reason that is my criminal record is not who I am now. And I don't want it to define me future further.
But when you break it down, many of them have a job or some prospect in the future about something they want to do, foster care, being a foster care provider or a professional license or something like that, that the, that they tell me the things are holding them back from.
[00:04:32] Speaker A: Terrific. And so you do hear, for example, seeking a new job as a frequent rationale, then among the folks that you talk to.
[00:04:40] Speaker B: Absolutely. A better, a better job. Getting a better job.
[00:04:43] Speaker D: Yeah.
[00:04:43] Speaker B: Do doing something either either maybe where they are or in a different field or a different company, but they want to do something that different.
[00:04:55] Speaker A: Terrific. And knowing that this might be one of the rationales, but there might be many others, including the identity rationale that you discussed. Marilyn, we asked participants at enrollment what were the reasons why they were seeking expungements. And Patricia, you took a look at some of those data. What, first of all, what were the. And Ryan, you did as well. Patricia, what were some of the reasons that we allowed folks to check box off. In other words, what was the form of the question that we asked them at enrollment?
[00:05:23] Speaker E: We had to tell us all reasons that apply for either get an expungement. We asked them whether they were applying for a job that they didn't think they could get because of the criminal record to move to an apartment or a home. They don't think they can get the record. The certified for Section 8 housing criminal record isn't who I am now. They were afraid that someone in their life, a friend, a family member, partner would find out about the record. One of the and we also gave them another option which some of our participants could use.
[00:05:48] Speaker A: Terrific. And so Ryan, you crunched some of the numbers here in terms of why people, which ones which boxes people checked off by the way. We did since that was an enrollment. Filling out the enrollment survey was a requirement of enrolling in the study and getting and potentially getting the assistance at Kansas Legal Services for expungements. We did have 100% response rate for this question. And so Ryan, what were some of the two or three most cited reasons for why people said they were seeking record clearing?
[00:06:19] Speaker D: Yeah, overwhelmingly as Marilyn had mentioned, people cited identity reasons. Right. That the that having a criminal record didn't reflect who these people believed they were. And that was at 98% for both for both the self help and the legal representation group. I don't know what that says about the 2%. I suspect they would have more interesting thoughts on why on that question. But the next group down at roughly 84% of respondents was that they were applying to a job. Those two categories, the identity and applying for a job are are overwhelmingly the two highest categories. And there's no difference between the the groups, the self help and legal representation statistically that we can detect. So we know that there wasn't really like differences in reasons between those two groups.
[00:07:06] Speaker A: A third kind of most popular or
[00:07:08] Speaker D: most most selected reason had to do with moving but that was about at around 43% of the sample. And that was that's like you're getting now been significantly down and the other reasons drop even more precipitously.
[00:07:23] Speaker A: We'd have a separate section 8 category that garnered around 8 or 9% as well. Some of those folks may have also checked that want to move. Some of them may not have. It's just anyway that housing was was the third most if you group that together housing was the third most cited reason.
We are going to talk about the identity and the housing outcomes in the third of the three Plan podcast series.
Today we're just going to simply focus on the employment because that is the one again after identity that was most frequently cited and it's the one that is most frequently analyzed in the academic literature and the empirical literature.
So we're going to talk about that. And so let's launch right in and just get the elevator conclusion here. So Ryan if you had to conclude give somebody the 32nd version of the findings from this study in terms of employment and what expungement does vis a vis employment, what would you say in terms of first of all, did people think that expungements were going to give to result in better employment and did it look like from what we can tell that expungements actually did result in better employment for them?
[00:08:31] Speaker D: Yeah. So I given the kind of like possibilities of that question, I feel like we end up in the worst world where yes, it does make people think that they're going to have improved prospects, but those improved prospects do not materialize at all.
[00:08:47] Speaker A: That we can see so far at least. So a reminder that what we did here for everybody that may not may have missed podcast one in our series is that we couldn't randomize expungement in the sense that we have sit in a judge's chambers and say Here, there are 500 people here who are eligible for expungements, randomly give 250 of them expungements and 250 deny them for no legal reason.
But because getting an expungement in Kansas is so difficult from an access to justice point of view, we could randomize whether Kansas legal services offered full representation versus versus self help materials in a conversation and sort of 45 minute guidance conversation.
And again, because it was so difficult, the overwhelming majority of people who got full representation were able to achieve expungement within two years versus the overwhelming majority of people who got self help and the conversation did not achieve expungement within two years. We are writing this report based on two years of data, two years post enrollment for everyone. So the last enrolled people still have two full years of follow up.
We are going to follow people administratively for five years and so stay tuned for a further report. Of course we will issue a further report.
What we can see right now, Ryan, you're characterizing as the worst of all worlds. Basically.
[00:10:09] Speaker D: Yeah, I want to take that back. It's possible that expungement could have made things worse. I don't think anyone believed that it
[00:10:14] Speaker C: was going to happen.
[00:10:15] Speaker D: I didn't see that, so didn't see that. That's technically the worst.
[00:10:18] Speaker A: So we saw and said optimism about better jobs, but the better jobs appeared not to materialize.
And they're the this. We track this, Ryan, is that correct? With a combination of surveys and administrative data, Is that right?
[00:10:34] Speaker D: Yes. So every roughly five weeks or every roughly five weeks, individuals would get a survey and one of Those surveys would inquire about their, about whether or not they had applied, whether or not they had attained, and then what their kind of overall employment situation was. And in addition to that, we also have direct information about wages and benefit information from the Kansas Department of Labor, which we can infer employment situation from.
[00:11:00] Speaker A: And this is formal market wages. Formal labor market wages. So these are. This does not include self employment. It does not include gig economy. Is that correct?
[00:11:09] Speaker D: Yes, yes. So this would. They'd have to be paying directly into the Kansas unemployment systems to be able to get picked up.
[00:11:17] Speaker A: Excellent. And Marilyn, did you want to make a comment?
[00:11:22] Speaker B: Really, all we're saying is that their income did not go up. They may have changed from.
We don't know if they changed employers, for example, or if they went from two jobs to one job at the same wage or if they got benefits without changing the wages. We're really just saying their wages didn't change. Right.
[00:11:44] Speaker A: Actually, I think it's a little different because I think if I'm right, Ryan, we can see changes in employers and actually. And our problem is that the data were not sufficient to allow us to tell whether their wages went up.
So I think it's that, Ryan, we can see changes in employers and because we asked them that, we asked them whether they applied for and whether they got a new job.
[00:12:08] Speaker D: So there's two questions I think that get at Marilyn's point here that of what are we thinking of just beyond wages themselves? One is we did directly just ask if they had gotten into a new position so we could see changes in employers. We also asked them about job satisfaction. So the idea being like, if their job got easier because they are now working one job instead of two. Right. That their employer might not have changed, but feasibly they should have gotten happier about their employment situation. Likewise, if they stayed with the same employer and their wages increased but not enough for us to detect, again, you'd assume their job satisfaction would have gone up. And we also didn't detect any changes in how satisfied they were with their job.
[00:12:46] Speaker A: Although I want to be careful because the intervals, in other words, since it's a randomized control trial and there's statistical variation, we can't rule out small effects. And so it's. We can, we can only say that outside a certain interval is statistically unlikely.
And for some of these employment outcomes, the intervals were extremely large. And so we couldn't, as I was mentioning, we couldn't say whether people had increased formal labor market wages.
We couldn't rule that out because the intervals were too wide. However, we were able to rule out anything other than a small increase in an unemployment, excuse me, a small decrease in unemployment, a small increase in full time employment and a relatively small increase in job satisfaction. We were able to rule those out based on the information that we got. So I just want to make sure that we're not over claiming here and then so actually let's clarify here because we do have some more people to thank. Renee. We obviously were able to administer surveys and had very good luck. We had an over an 80% response rate, which is unheard of in a social science survey. But then we were lucky enough to work with some excellent field partners in the Kansas government. Can you tell, can you remind us who those were?
[00:14:02] Speaker C: Yeah, we, we worked with, as mentioned, the Kansas Department of Labor and the Kansas Department of Revenue.
Those data sources showed us wages as well as unemployment compensation Benefits and industry IDs, industry in which somebody was working. And then of course, Department of Revenue also complemented the wage data in its own kind of unique way. And then we worked with the Kansas Bureau of Investigations, which is the criminal justice record repository and of course the court system and the clerk of courts in all of the jurisdictions in Kansas.
[00:14:40] Speaker A: So most of the data that we'll be discussing today comes from the surveys in the Kansas Department of Labor information. Because we weren't yet able to fully incorporate the Department of Revenues data yet. That turns out to be very difficult to understand. We're still taking a look at that. But certainly also the criminal justice records from the court System and the KBI, because those tell us, along with your hard work, MD searching records over and over again to see if they're visible whether people had gotten expungements. Was there a statistically significant effect on what we call job application deterrence? So we saw a reduction in the, in the deterrence to apply for a new job, which we're translating in to a sense of optimism that you could get a new job, a different job. And again, the phrasing of the question was to, to whether you would apply for a new job.
I might have interpreted that as including, and this is something we're going to return to, Marilyn, I might have interpreted that as either applying for a job within the same company or applying for a job within a different firm or a different employer. Do you have any sense of how people would likely have employed, would it have been both or would they have thought of only one?
[00:15:58] Speaker B: No, I think both is correct.
[00:15:59] Speaker A: Yeah. Okay. So we see then from the surveys that the statistical significance is not exactly slam dunk. And I'm not sure that it would survive something that the statisticians call multiple testing because the p value is 0.049.
And so it's lower than the traditional threshold of 0.05, but just barely.
Nevertheless, we're going to say that potentially there is at least some evidence that there is a reduction in deterrence about applying for a new job.
And the effect is probably somewhere in the kind of 5 percentage point range.
And so that's where you started out, Ryan, in terms of saying this is not the worst of all worlds, but this is a somewhat depressing finding that people are more optimistic.
The question asked, are you, do you, are you deterred from getting a new job? Is your criminal record deterring you from giving a new job?
And the yes response to that question was lower in a statistically significant way with a P value. That's right. Of less than 0.02.
And the suggestion was about a 5 percentage point drop.
Then we asked a series of questions and we looked at a series of administrative data points from the Department of Labor, Ryan, and so let's walk through those. One of the things that we looked at was full time employment, whether there was an increase or a decrease in full time employment and how did we define full time employment?
[00:17:30] Speaker D: So, yeah, we have to be careful because there's two full there. We did two, we did full time employment two different ways. Once from the surveys, because we did ask on the surveys, but then from the Department of Labor data, Patricia actually did the work to aggregate up all the, I believe, week level wage and benefit data that we obtained to quarters. So we could see exactly how much in any particular economic quarter of a year a person had obtained in wages for that quarter had paid into the Kansas, the Kansas Department of Labor systems. And from that we looked at if you, if an individual had made less than $100 in that particular quarter, we categorize them as being unemployed for that quarter.
[00:18:08] Speaker A: So that's the unemployment. And then we said if you are full time employed, how did we define that?
[00:18:14] Speaker D: We calculated what the, what the minimum amount you would need making minimum wage employed at full time during that period.
And if you were over that threshold, which is escaping me at this point, you were declared full time employed.
[00:18:28] Speaker A: We tied it to the minimum wage at that time. Right. And said that if you made above that in a quarter, then you were full time employed. Yeah, it was formal labor market full time employment. And Patricia, before we discuss those which are actually these Results are actually easy to summarize. Can you just tell us the data aggregation that you did? That was because the data sometimes came in the form of weeks, sometimes came in the form of months. Tell us about that.
[00:18:52] Speaker E: Yes, for the Department of Labor, we got two different data sets from them. One of benefit claims and one of wages, and those were simply tracked differently. So for wages, those are reported as a quarterly amount by. We have an ID for each person, we have an ID for their employer, we have an amount of wages. We'll have the year, we'll have the quarter. Because that's how the Department of Labor wages claims are reported in a very different manner, where you become eligible to collect benefit on a certain day and then you are eligible for a number of weeks. And during that time, we know the total amount of benefits you could have collected and the total amount of benefits that people did collect. And so in order to know how much benefit someone is drawing in that same period, a slightly more complex calculation had to be done to first put those weeks into quarters and then figure out an average of how much per week the person would have gotten, because the Department labor just tracked those differently.
[00:19:45] Speaker A: Super.
[00:19:46] Speaker E: So those combined together. Can Cliff refer total income as reported Department of Labor, which obviously we know that not all forms of employment are reported to Department of Labor.
We can't guarantee that someone not appearing in the data set means they are not employed. But for many people it may. And the majority of our participants were employed.
[00:20:04] Speaker A: Yeah, that was my next question. It turned out that a surprising number actually did appear in the Department of Labor data. It was the overwhelming majority of them appeared in the Department of Labor data. So this wasn't an issue of most people being only in the gig economy or only in the self employment category.
Is that correct, Patricia?
[00:20:27] Speaker E: Yes. The majority of your participants defining the data, they are reporting a wage.
We also, the majority of participants are filing a tax return and reporting a wage. And very few of our participants, a tiny percent, are growing benefits point. The only notable period where we have it's still a kind of minority, but big enough to even look at is really only during 2020 where pandemic benefits were available. So the majority of our participants are receiving an income from a job already before they received their study. That took baseline.
[00:20:58] Speaker A: Yep. And this we want to return to because this may explain in part the lack of an effect in terms of expungement people.
The story that people are unable to get any form of employment and they get an expungement and therefore now they become eligible to be employed. That's not what we saw in these data among our participants. And at least I think that it's probably what's going on in terms of people who are able to seek expungements under current law.
It may be different if you employed a different record clearing mechanism, such as a clean slate statute that automatically after a certain period of time clears records. But for people that are able to get their acts together to go through the demanding process of getting an expungement under state law in most states, including Kansas, it may be that they are already employed in many instances. And so Marilyn, this is something we're going to return to ask you about in terms of what you among your clients. Let's go through again, formal employment, full time employment and unemployment. These are all things that we were able to look at from a combination of administrative data and surveys. Ryan, back to you. Was there any statistically significant effect on any of these outcomes?
[00:22:23] Speaker D: Sadly, no.
[00:22:25] Speaker A: These were by then. For those outcomes along with job satisfaction which came from surveys, there is, there is some suggestion that we can rule out just policy relevant effects. At least some policy relevant effects. The intervals were tight enough to allow us to suggest, for example, that for formal unemployment, meaning you were below $100 for a quarter, we can basically rule out a reduction of unemployment of four and a half percentage points or larger.
In full time employment, we can basically rule out an increase in full time employment of 2 and a 2.3 percentage points or larger.
And then for job satisfaction, it's hard to measure because it's a survey based measure there. It's hard to tell because it's based on standard errors of the responses for new job applications. In other words, when we asked people whether they could get new jobs, Ryan, was there a statistically significant effect there?
[00:23:26] Speaker D: Yeah. Despite the optimism that the individuals had, we did not actually see that translate into actual observed new job applications. There was no statistically significant difference between the groups.
[00:23:37] Speaker A: And again, we can rule out an increase there of new job applications of 2.4 percentage points or larger. Again, there couldn't have been a. There should. The data are inconsistent with a larger effect.
There were a series of income measures that we took from the unemployment from the Department of Labor's information and there. Ryan, what happened to the intervals there? At least as far as we can on our current modeling techniques, what happened to the intervals?
Intervals around the effect sizes?
[00:24:07] Speaker D: Yeah. So Inc. Income, they were huge.
[00:24:11] Speaker A: The intervals were huge. Meaning not the income, the intervals. Right.
[00:24:14] Speaker D: Yeah. In part that's because people's incomes are very, are highly variable. And so there's a lot of space over that, over which people's incomes could take on. And so we had looked at regular, just like the, just the raw income data, we had logged income values and we had done some ratios as well to attempt to get those variations down. But regardless of what kind of transformations we applied, the intervals on our treatment effects were still very large.
[00:24:44] Speaker A: It's too large for us to say anything, quite frankly, say anything useful in terms of how much money from there, from this source people were earning. We're still trying to see if there's anything we can do in terms of variance reduction and we're going to try to incorporate other data sources. And again, this is based on two years worth of information to your follow up.
So it's possible that five year follow up may give us more information that might tighten those intervals up or show an effect. But right now we can't rule out the possibility that people made more money in, for, in companies for which they were currently working without thinking that they were applying for new jobs.
And so this is what I, Marilyn, what I was hoping you could talk to us about. Basically it sounds like in order for there to be a story that says no, it's still the case, consistent with the data that this study generated, that people are earning more money even though the Access to Justice Lab and Kansas Legal Services data don't show it. They'd have to basically be earning more money in the job where they already are because they don't think that they're not reporting that they applied for new jobs. And I guess my question is based on your work with clients, do you think that's very plausible that would have happened?
[00:26:01] Speaker B: No.
One of the things that I hear from clients is a gratefulness that somebody took a chance on them and that that could be a reason why an employer hires somebody with a record. Now they don't have the record, but there's still a feeling appreciation for that employer who took the chance on them. So they, they may not be, they may be staying with the company but doing new things and those kind. That would be one thing I hear that might give a window to explain these, explain the, the apparent results.
[00:26:48] Speaker A: But, and that would be a way where it might be possible for them to earn more money in the current job where they are. Um, I guess you will let listeners make their own judgments on whether that's plausible. For me it sounds in terms of a, in terms of happening a lot, it sounds a little far fetched. To me. But again, listeners can make their own decision, make their own determinations. Do you what's your view on whether that would be happening? An awful lot. That they'd be making more money for the same employer, doing essentially the same work, not having a new job.
[00:27:20] Speaker B: So some of my my anecdotal information is the person who was doing very successfully dealing with catering orders for the restaurant but could not move up to the next level of overseeing a regional catering, you know, doing that job regionally because of her record for the person who and I have a client who works at McDonald's and would love to get in the management track there, but can't because of a record.
So it's a bit anecdotal but. Or and the, and the third example is the person who is a CNA but be and because of their record, they can't get a license to be a cma.
[00:28:17] Speaker A: What is a CNA and a CMA for listeners?
[00:28:20] Speaker B: A certified nursing assistant and a certified medication assistant. So they couldn't the employer would let them get people in and out of bed but not give them their medication because of their record.
And but if it's expunged, they can do that that different jobs. So I can think of a few instances where that seems plausible but agree there's not a, it's not everybody by any means.
[00:28:52] Speaker A: Patricia, go ahead.
[00:28:55] Speaker E: In Maryland's example, there is actually reflected in our data in our question of asking systems why they want to receive an expungement. Some of our participants checked other and we did have a few responses specifically saying to get a nursing license and a couple people who said that they worked on a military base and they couldn't get a security clearance. So although these are not a majority of the participants, we do see that in our data that for certain specialized fields, we had participants who were looking for a specific certification that was barred from them.
[00:29:23] Speaker A: Terrific. So this is something that could be happening some. It would just have to happen a lot for there to be an effect that we wouldn't have detected in the, in the questions and the data that we have.
I certainly can't rule it out. So suppose it is the case, Marilyn, Suppose it is the case. That mechanism might be happening for some people, but not for a lot of people.
I at least would call these findings surprising that we're not seeing a lot of people move up to new jobs. We're not seeing a lot of people applying for new jobs and reporting that they got them.
We're not seeing movement in the administrative data if it turns out that's true, that appears to be in tension with the traditional rationale for why people seek expungements. And certainly they reported that they were looking for better employment outcomes. If it is true, then that expungement doesn't support this tradition, that the data don't support this traditional narrative. Why might it be? What could be the problem here that is preventing expungement from doing what we all hoped it did it would do.
[00:30:38] Speaker B: I think we're laying on to this specific situation a rather stag, a rather stagnant employment market anyway. Right. Nobody's getting much higher wages. So expungement people with expungements are kind of getting treated like everybody else.
The interesting and I don't know if we have enough data and I don't understand what Patricia and Ryan do well enough to say so what's going on with those people who didn't complete their, the self represented folks who didn't complete their expungement? Were they and I assume it's what we're comparing this to that that they also had stable employment, stable wages.
[00:31:26] Speaker A: An awful lot. Right. An awful lot of them did. Right. Which is again where we started out that the set the population of people who are able to seek expungement. So who to whom it occurs to do so are able to find Kansas Legal Services, et cetera. And I think this is probably true in many states they were already typically employed five year.
[00:31:47] Speaker B: Your probation officer makes you get a job. Right. It's one of those things that that happens in order to get out of the situation. To start with, you have to have a job and then so you have a job and then you have to get through your life for three to five years before you can get expungement. And that means probably you have a job.
People have found a way.
[00:32:12] Speaker A: Yeah. So this is I was, I wanted to highlight the potential for waiting periods the way the eligibility criteria for getting an expungement, the possibility that there would be waiting periods after though not possibility there are waiting periods. Renee, I'm going to ask you in just a minute what those typically were under expungement law. What they don't generally look like. But the recent scholars in an article that we'll link to in the blog post, there's a recent posting from some excellent economists who have suggested proposed calling this scarring that after a certain period of time living with a record your employment prospects, your housing prospects, whatever it is, they're scarred and that the scar won't heal based on the removal of the criminal justice record because and this is where I was heading basically the year you have to wait so long until you are eligible to get an expungement that as you've just said, Maryland folks learn to live. So this is one, one possibility. Renee, can you remind us generally under Kansas law in terms of expungements was surprisingly generous in terms of what it allowed to have cleared, but it made you wait a while before you could do the clearing, Is that correct?
[00:33:21] Speaker C: Yeah, that's correct. So Kansas law allows clearing of convictions, which is unusual typically, or a lot of record clearing law does not allow for clearing of convictions and limits itself to non convictions.
But to clear the record, you had to wait in a period of three, five or 10 years, depending on the type of record you're trying to clear. And that waiting period doesn't begin to run until you're you've fully completed your sentence. So this is what Marilyn was alluding to just a moment ago.
That could include a completion of your probation post incarceration or parole post incarceration or any period of monitoring that needed to occur after your sentence and which also could include a period of incarceration. So the waiting period doesn't begin until then. I think most of the charges that were being clear that can be cleared in Kansas and were being cleared in this study were a waiting period of three or five years. There's very few that fall into the 10 year.
[00:34:32] Speaker A: So again, three to five years starting after the last day of your probation, if there is one probation not starting until after you're released from a period of incarceration. Incarceration not starting until after your conviction
[00:34:47] Speaker C: or after you're discharged from those. So completion of all, completion of all these stages, all the conditions.
[00:34:53] Speaker A: So again, long after the first set of records are created, which is the first set of records are typically created after arrest and then become visible to the public via the court system after a charge.
And the charges typically follow right on top of the arrest. So a potentially long period to wait.
[00:35:13] Speaker C: Yeah. Arrests are viewable in criminal record searches, but charges do not become. But the case does not become a case on the court record, which is much more accessible until charges are filed.
[00:35:28] Speaker A: Yeah. So the arrest would be available, say from the Kansas Bureau of Investigation, from the KBI record, the chart which you have to pay for. But the charge would be now especially visible from just simply hopping on the Internet.
[00:35:40] Speaker C: Correct.
[00:35:41] Speaker A: Yeah. In summary, I think a little bit surprising finding and Ryan, use your characterization a little bit of A depressing finding, something that we, quite frankly, at the Access to Justice Lab had not hoped for. But we want to know the truth, of course, and to figure out what what's going on. Marilyn, I'm just going to ask everybody for final words, but I'll start with before we close, the second of the free podcasts, Marilyn, any final thoughts here? Obviously, a couple of different mechanisms for why this might not have had the effect that we'd hoped for.
You talked about the fact that the economy, employment economy for people in this wage, wage range is generally sluggish and has been for a long time because many of these folks are not at the upper end of it, where all the dynamic movement has been, you know, in the past 20 years or 30 years, also potentially scarring in the long waiting periods.
Any final thoughts you want to leave us with in terms of the employment related findings for this study?
[00:36:44] Speaker B: The only I guess thought I have is that we haven't really addressed, but I see all the time the participants saying I got by because of a spouse or another person in the house.
And that's certainly something that people end up with as a way to get by. And I think things like satisfaction are not insignificant when we think about somebody who is in a, in a relationship for very true and realistic financial reasons and feel like they have no way out. So I wish we had stronger results, obviously. But we want to know, as you
[00:37:27] Speaker A: said, yep, among other things, if it were, for example, the waiting periods that's causing and that's actually the speculation that I engage in is it's the waiting periods, then maybe the policy response is not saying, oh, expungement's not worth it. It's actually perhaps it could be very effective if we just took chances on people a little bit earlier in the process. We're willing to either clear these records that has its own issues in terms of accountability and accessibility of what the government is doing to the public. But we can talk about those another day.
Renee, final thoughts from you.
[00:38:03] Speaker C: Yeah, I think we were surprised. We definitely expected to see different results on the employment front. I think when we began this study, our hypothesis, I think evidence that and I think what foreshadowed these results were the surprise that I recall seeing when we got the first round of Department of Labor data and seeing just how many of our study participants actually did exist in that data. And Jim, I think you mentioned we got this data for a period of two years prior to enrollment. So we have some historic records and that probably foreshadowed those results at that time, I don't know that we necessarily were thinking that far ahead, but I do recall being surprised by that.
And then if you'll indulge me, I just have one more final comment.
[00:38:52] Speaker A: Absolutely.
[00:38:53] Speaker C: And just to say that in addition to the waiting periods, we have a podcast that was dropped maybe two weeks ago about the difficulty of achieving expungement. And so I would just caution policymakers to think that changing the waiting periods is enough.
Perhaps don't change the waiting periods and increase the effort someone must make to get an expungement, but perhaps couple those two solutions together.
[00:39:24] Speaker A: Yeah, and of course, one way to get around potentially both is to have a clean slate statute that will do this automatically. The issue there, which we can talk about dedicate another podcast to is that the. That it starts to look like ban the box, which is. Which has some fairly disturbing evidence regarding its effects.
So we'll need to have a podcast on that sometime as well. Patricia, final thoughts from you.
[00:39:50] Speaker E: I think one aspect of our data that and Ryan can definitely speak to how we've dealt with this is that our research never takes place in a vacuum. Right. We just happen to be collecting these data during a period of immense economic upheaval.
We had a global pandemic. And depending on where people lived and what field they worked in, even if they were feeling more optimistic about applying to new jobs because of their expungement, the general economic situation may have reduced that and they may have decided to stay where they were because there was a risk in trying to make a change at that time. So that definitely affects our ability to detect something.
My perspective is that it's true that we have not proven that expungement benefit people's employment. But you put it the other way and say is there a proof that making people wait this extremely long time at a record bench society, rather than trying to justify why we shouldn't harm people, do we have a justification for doing it in the first place?
[00:40:45] Speaker A: And I struggle to come up with one quite frankly. But I. But one could, I can hypothesize one. Just not one that I personally believe there are such things. No question about people cite public safety.
The recidivism numbers are their own thing that we can talk about. But there, there could very well be. There are such rationales, just most of them I don't personally find appealing. But Ryan, final thought from you.
[00:41:10] Speaker D: Yeah, I'll try to make this quite digestible as I can, but something that Marilyn referencing reference is the nature of the RCT Means that fundamentally what we're doing is we're trying to make a comparison about what would have happened to a person had they not gotten the expungement in the case of getting the expungement. And our best source of that comparison is the group of people who didn't get expungements. Right. So fundamentally, this is just about differences between the experiences of those two groups.
[00:41:40] Speaker B: Right.
[00:41:40] Speaker D: There's lots of stories that you can fit that can explain why we don't observe differences between these groups for employment. It could very well be the case that an employer is telling people that, oh, I'm not. You're not getting a job because you have a record. But the data is showing us that even if they didn't have a record, the employer would have just found a different reason to not give them a job.
[00:42:01] Speaker E: Or you can flip that on its
[00:42:02] Speaker D: head and say, maybe the employer said, hey, I'm giving you this promotion because your record got cleared, so you're now management tracker. You're now capable of doing this thing. And that. And the story is saying that if you hadn't got the expungement, then that wouldn't have happened. But the data is telling us that the employer just would have found a different reason to promote you. Right. Or would have found an alternative method. Or maybe the people who didn't get expungements are just actually pretty inventive in finding alternative ways to succeed in the same way that people who did get expungements. And that's why we're not seeing those different datas. All right, so there's. I think, yeah, there's still lots to know here about what's specifically going on, and that can fit lots of different things that may be happening.
[00:42:42] Speaker E: Yeah.
[00:42:43] Speaker A: Although we'd hope that the randomization would take care of the inherent inventiveness. It would balance the inherent inventiveness of people in the. The group that got expungements and the group that didn't. But you're. They could just be. Again, we probably think that the people who are able to apply for expungements under current law are probably invented. Right. They're taken. They're seeking remedies that the law provides when many people do not. Because, as we'd covered in the previous podcast nationally, looks like from what people can tell, Only something like 5 to 7% of people who are eligible for expungements ever achieve them. So there is this vast expungement gap that Colleen Chin and J.J. prescott and Sonia Starr have documented previously. We've had J.J. and Sonia on the on Proof over Preston in a prior paper before. So that's going to be it I think for today. On terms of the employment outcomes for this study and the next the third of the three podcast that we're going to record, we will discuss the most common rationale that people cited, which was the identity rationale and whether there was improvement on identities. And then in terms of from getting an expungement as well as housing, as well as whether we can see anything in terms of how movement improvements on housing.
We will ask folks to look forward to that. And in the meantime, I want to thank Patricia, Ryan, Renee and Marilyn for taking the time to record this with us. And we'll look forward to a conversation a third conversation very soon. So thanks so much everyone for participating in Proof Over Precedent.
[00:44:20] Speaker C: Thank you.
[00:44:21] Speaker B: Thank you.
[00:44:22] Speaker A: Proof Over Precedent is a production of the Access to Justice Lab at Harvard Law School.
Views expressed in student podcasts are not necessarily those of the A2J lab.
Thanks for listening. If we piqued your interest, please subscribe wherever you get your podcasts. Even better, leave us a rating or share an episode with a friend or on social media.
Here's a sneak preview of what we'll bring you next week.
And so today we're going to the third of the three. We're going to be discussing the other two primary outcomes that we studied in this research effort. Those are related to housing and identity, to see whether there is any improvement in those two areas.
