 So good morning, good morning, everyone. My name is Laura Schmarmann. I'm a PhD student in the Department of City and Regional Planning here at UC Berkeley. And I'm also a graduate student researcher within the Center for Community Innovation. And I'll talk more about the Center and the Urban Displacement Project shortly. So today I'm presenting along with undergraduate students Hannah and Rachel on a research project which we are undertaking which maps the vulnerability of BIPOC owned businesses across the Bay Area in the wake of the COVID-19 pandemic. We have been working on this project since March of this year. The project is ongoing and I would like to just emphasize that we're presenting the preliminary findings. The outputs of the project are not yet finalized or publicly available. So I'll talk through the Center for Community Innovation and UDP and provide an overview of the project we are working on. Rachel will talk through the Neighborhood Vulnerability Map. Hannah will talk about business database and associated map and we'll wrap up with some findings and plenty of time for questions and answers. So if you have any questions, please hold them till the end. You're welcome to post them in the Zoom chat and we'll get to them at the end and we'll also welcome questions from the audience here as well. So the Center for Community Innovation at UC Berkeley has a mission to nurture effective solutions that expand economic opportunity, diversify housing options and strengthen connection to place. The Center houses three projects, one of which is the Urban Displacement Project. So the Urban Displacement Project conducts community centered data driven applied research towards more equitable and inclusive futures for cities. So our research aims to understand and describe the nature of gentrification, displacement and exclusion and also generate knowledge on how policy interventions and investment can respond and support more equitable development. So I'd encourage you at a later date to have a look at UDP's website and explore more of the kind of data and mapping tools that we've produced. So now turn to this specific project. So many small BIPOC owned businesses are already vulnerable and are at risk of closing permanently due to the ongoing impacts of the COVID-19 pandemic. While research has been undertaken at a national level, there has been limited research on BIPOC owned businesses in the Bay Area. Various stakeholders we have spoken to including business advocacy organizations, chambers of commerce and cities have undertaken surveys of businesses to understand how the COVID-19 pandemic has been impacting their business. However, the information is rarely publicly available and it's not consolidated. So we are working with the Asset Funders Network to undertake the study of BIPOC owned businesses across the Bay Area. The Asset Funders Network is a national grantmaker membership organization and they focus on advancing equitable wealth building and economic mobility. So the study which we are undertaking aligns closely with the mission of that organization. The project will bring together a range of data sources to create a database on minority owned businesses across six cities in the Bay Area and we'll explain more about those specific locations shortly. So now I'm just going to provide a bit of context in terms of the study. So in March of 2020, as I'm sure you're aware, stay-at-home orders were introduced across the Bay Area in response to the increasing number of COVID-19 cases. As a result, restaurants, bars, salons, gyms and other what are termed like non-essential businesses were required to close or alter their operations. So 18 months later or over 18 months later now, many businesses have either temporary or permanently closed and others have reopened with various restrictions or reduced capacities. So while restrictions have not been the sole driver of operational change, they have had a significant impact. A survey done by the Federal Reserve called the Small Business Credit Survey has identified various other reasons for impacting operations of businesses and reasons for closure. So one of these is the nature of people's lives during the pandemic and probably impacted change in demand for products and services. So you can imagine that people would have had reduced income and that would impact their ability to purchase products or services. Government mandates affected various businesses and their clients businesses compared to previously. This includes restrictions on operations such as restaurants. Other reasons include the need to adapt to health and safety guidelines, which can be quite costly for companies, businesses, worker availability, supply chain disruptions, global supply chains in particular have been impacted. And the final one is owner's personal or family obligations. So an example of this would be that many people had to childcare responsibilities with homeschooling that would have impacted their ability to run their business as well or staff their business. So the drivers of operational change we expect would vary depending on the location of the businesses as well. Businesses operating in downtown commercial districts, for example, would have experienced reduced demand associated with the shift to working from home. Office workers were probably a significant source of demand and with businesses closing up that demand would have reduced. So an example of this would be in San Francisco where there's a lot of offices. So the types of financial challenges faced by businesses have also varied including paying operating expenses such as labor and rent, making payments on debt, purchasing inventory or supplies to fulfill contracts. And this can be difficult when there's uncertainty about demand and also accessing credit. So reports have suggested that various forms of government support have helped to prevent dramatically higher business closures. However, this may not be a solution for businesses in the long term as they may have only been able to defer payments and they would eventually become due. Businesses who are provided with the paycheck protection program loans across the US were required to retain employees in order to meet the requirements of these loans, although some of these requirements were eventually relaxed. The first round of PPP loans only provided for up to two and a half months of payroll costs and the pandemic has obviously continued beyond two and a half months. Businesses with high labor costs or fixed capital costs would also be particularly vulnerable as they might not be able to shift these costs as well. So the US Census Bureau previously undertook a survey of business owners which included questions regarding race and ethnicity. The last survey unfortunately was back in 2012, so there hasn't been one more recent than that. This chart on the screen here shows the results of that survey. So in 2012, 29% of businesses across the US were minority owned. And that's shown by the black line across the chart. The sectors where there is an overrepresentation of minority ownership include transport and warehousing, other services, healthcare, admin and support services and accommodation and food services, which have about between 39 and 44% minority ownership. We've also prepared updated statistics based on the current population survey. The current population survey is undertaken monthly by the US Census Bureau and the Bureau of Labor Statistics, and it's the primary source of labor for statistics across the US. The survey uses a representative sample so we can extrapolate that out as an indication for minority ownership across the whole United States. We reviewed the industry of employment for self-employed respondents, so those who own a business. And according to that overall 26% of businesses are BIPOC owned, which is a slightly different statistic compared to that 2012 source. The industries with the highest BIPOC ownership are transport and warehousing, accommodation, administration, personal services and food services, and that's a consistent pattern with the 2012 survey. So McKinsey's produced this chart which identifies the most vulnerable industries based on both financial and COVID related challenges. So smaller businesses and minority owned businesses are more vulnerable. The industries which are particularly vulnerable, which are shown in the top right quadrant, accommodation and food services, arts, entertainment, other services, mining, transport and warehousing, wholesale trade. And we can see that these are the industries here with higher minority or BIPOC ownership as well. So you can start to see how there's an kind of cumulative impact of vulnerability among these groups. So for our study we are looking at how different characteristics of businesses impact their vulnerability across kind of three areas. The first is business ownership characteristics. So minority owned businesses have reported greater financial strains during the COVID-19 pandemic. Financial impacts vary depending on the minority group as well. So various reports have indicated that black owned firms credit availability is a top expected challenge. While Asian owned firms have cited weak demand and some of this is due to discrimination associated with the COVID-19 pandemic. In terms of industry sectors, accommodation and food services and personal services have been hit hardest due to the restrictions placed on these operations and the time it took for them to be able to open up. Businesses where workers were able to work from home have obviously not been impacted as severely. We are also looking at geographies, so the spatial distribution of these impacts. Businesses operating in lower income neighborhoods or in commercial districts or locations which have had more restrictive guidelines about business operation are expected to have experienced much greater impacts on reduced demand as well. So if you think about California's had more restrictions compared to other states, so businesses in California might have been more severely impacted from a purely economic perspective. So our study focuses on the industries which are most vulnerable to displacement or closure and those which have higher minority ownership. So we're looking at retail, personal services, food and beverage services, accommodation, arts and childcare. And we're also focusing in on black owned, Latinx owned and Asian owned businesses. And so the analysis which we are undertaking, particularly the business database, it has the potential to be expanded to incorporate other industries in the future, but we have to constrain it just to kind of limit the amount of analysis we could undertake in the short time frame. But we see the benefit of expanding it in the future. So our study has three kind of stages to the data component. The first is a neighborhood vulnerability layer and for this analysis we used the national level data on BIPOC ownership to identify minority ownership by industry and clusters of this across the nine County Bay Area region. And so this is a top down approach using data at the national level and identifying concentrations at the local level. And Rachel will talk about this shortly. The second component is the business vulnerability layer. And this covers six Bay Area cities so it covers Oakland, Richmond, Redwood City, San Francisco, San Jose and South San Francisco. So we've identified businesses which fall within the industry categories we're looking at and the minority ownership groups which I mentioned. And the map provides a more comprehensive picture at the local level because we're using local local level data. We've also undertaken some site visits to certain areas to kind of sense check what we have found as well. And Hannah will talk through this in more detail a bit later. The final component is the property ownership layer and we're using tax assessor data on property ownership to determine the share of businesses that own their own property. This is based on the assumption that businesses who own their own property are less vulnerable or at less risk of displacement than those who rent their property. We're still in the process of undertaking this analysis so I'm not able to present on this specifically today. In addition to the data analysis and mapping, we've also been undertaking outreach with various stakeholder groups across the Bay Area and we'll be developing a policy brief which brings together the findings of both the quantitative and qualitative analysis. So I'm now going to hand over to Rachel who's going to talk through the neighborhood vulnerability mapping. Hello. My name is Rachel McCarty. I am a junior studying data science at UC Berkeley. And this past semester and summer I was working on this project helping to create the database and helping to create the map that you see here. So perfect. So basically what I'm going to talk about today is we basically identified concentrations of BIPOC owned businesses across all zip codes in the nine county Bay Area region. So we made this database based on the national level estimates. I'm going to talk about this a little bit later. A BIPOC ownership from the current population survey and we did the number of businesses in each zip code from the zip code business patterns data. So one of the conclusions we got was that across the Bay Area there are about 61,000 businesses that are BIPOC owned across all industries. Okay. So now I'm going to talk a little bit about the different surveys, different census surveys that we use to actually create the database. So the first thing we use is a zip code business database from the Bay Area. So this is a survey conducted by the census. It's annual and it has a bunch of data about businesses by the industry and it's only businesses with paid employees. So basically this we use the survey to get the counts of total businesses in the Bay Area. So for example, we could look at some zip code, some zip code in like Berkeley and see the number of hair salons like in that industry. So we just got that using this database we got the total number of businesses in that industry. Then we also use the current population survey. It's a monthly US wide survey and it's a representative sample as Laura was saying. It's also conducted by the census and this is how we found the actual percentages that we applied to our database. So basically this would be like, oh, there are X percent of hair care like at a national level. And we applied those percentages to our original zip code business database to get the counts like estimated counts of BIPOC businesses in the Bay Area. So like I said, we combine these two databases and then this is how we actually created a map that shows the proportion of BIPOC businesses in the Bay Area. Okay. So here's an example of one of the maps we made. We can see here that we have a map and there are different sections in the different colored sections are each zip code that we actually looked at. So we can see like a little bit of a range. We see like lighter colors representing that there are less BIPOC businesses and darker colors representing their more BIPOC businesses. So we can see I just like this is interactive maps. So for the screenshot, I just like scrolled over one of our zip codes. So nine, four, five, 10. And you can see we have the percentage of BIPOC businesses on there. The total number of BIPOC owned businesses, the BIPOC owned businesses per 1000 people and the BIPOC owned businesses per acre. So we did this for everyone in this map is actually representing the total number of BIPOC owned businesses, not the percentage. So all the different colors are representing how many BIPOC owned businesses are there. Okay. So here's another data table that we created. So this is a table representing each county in the Bay Area, the number of BIPOC owned businesses. This is calculated by using both the percentage we got from the CPS survey and the total number of businesses using the zip code business patterns data. The total number of businesses is just from the zip code business population survey and then the percentage BIPOC owned is just a ratio. So we can see here that Solano, San Mateo and Alameda have the highest concentration of BIPOC owned businesses with 30.9, 30.4 and 30.3 respectively. And we see that surprisingly San Francisco and Marin have less BIPOC owned businesses with 29.7 and 28.7%. So you might be thinking these percentages are really similar, like the biggest difference is 30.9 to 28.7, which is not that big of a difference. But you have to keep in mind we applied these percentages from the CPS survey. So it's a uniformly, we uniformly applied them to every business or every, sorry, every county. So these percentages are very similar, but it does make the differences very significant because of the difference between Solano and Marin is actually very different considering that we applied these uniform percentages. So yeah. So I wanted to compare two maps side by side and it's a little bit hard to see. Sorry about that. But so on the left we have the total number of BIPOC owned businesses. This is the same map I showed a little bit earlier. And we see that in like Northern San Francisco we see some concentrations. We see some concentration in the San Jose. These are again places where there's a lot of higher quantity of BIPOC owned businesses. And then on the right we see a definitely more uniform distribution of percentage of BIPOC owned businesses. This map is the colored based on the proportion of BIPOC owned businesses. So like I said, we do have a very uniform spread because we applied those uniform percentages. But yeah, we don't see that. I think it's interesting that we don't see that concentration really in San Francisco anymore. And we kind of theorize that's because San Francisco has a lot of businesses, but the proportion of BIPOC owned businesses is not as high. So here are like two concentrations I wanted to look at a little bit. So we can see that even though it's really hard to see the coloring on the screen on the screenshot, but in North Northern San Francisco we don't see any like higher concentrations. But in South San Francisco we see a little pocket that has a higher concentration of BIPOC owned businesses. Oh, and I'm sorry, these two maps are about the percentage. And then on the right we see Oakland and we I just highlighted that one zip code in Oakland that has a higher percentage than a lot of its like neighbors. That is 32.66% of BIPOC owned businesses. So that's just how we identified like concentrations both by looking at the table and both by looking at the maps. So kind of summarizing what I talked about, we saw a high number of BIPOC owned businesses in San Francisco. But when we actually looked at the percentage of BIPOC owned businesses, it was actually very uniform and very standard to what we saw everywhere else. The percentages of minority owned businesses in our categories are pretty nor pretty uniform, like I said, because we applied those uniform percentages, but because of that are the differences were very significant. And we saw higher concentrations of BIPOC owned businesses in Solano, San Mateo and Alameda. So I want to talk a little bit about like what I learned about doing this because I'm an undergrad student and I like I learned a lot of things for the first time. So it was kind of difficult to create this database I've never worked with census data before. So I had to learn how it worked and I didn't realize how much census data there would be. I think when we actually filtered by the zip codes we wanted there about 150,000 to 200,000 like records in the database. So it was really interesting getting to work with them seeing like what each column which each record actually represented because it was pretty. It was not as clear as I thought it would be. But it was really interesting and it was lots of trial and error with queries and trying to figure out how exactly we get just the data that we want because these surveys have a lot a lot of data like not even just records a lot of just like different things that they keep track of. And it was actually really cool though and I'm really glad I got to work with that. And maps actually were really interesting too I've never worked with maps like creating maps before. It was actually pretty easy to learn and I really liked working with maps because it was a lot of information that we put in. Because first we have the layer that's just like the Bay Area that we just have like this back layer. Then you have to add a layer for each zip code or each county like depending what we're looking at. Then we have to add a layer for each metric, like for example our map kept track of the total number of BIPOC owned businesses, the percentage of BIPOC owned businesses. A thousand people on the BIPOC owned businesses per acre. So it was a lot of information that we encoded those maps. And part of the reason I liked them so much was looking at the tables it was harder to see like what was happening. It's definitely hard like you can sort things but it's harder to actually get a clear picture of what's happening. But using the maps it's really intuitive and really easy to understand and we're going to show the maps like how they actually are in a little bit. But it's really helpful to actually see and get to play around with the maps a lot. And after the project I feel a lot more comfortable working at census data, like working with maps, creating maps and driving databases. I'm really excited I got to work on this project. It was really interesting and I'm hoping it has positive impacts. But thank you. Hello. My name is Hannah. I'm a senior studying data science. I worked on this project starting from this June till now. I worked in with a group of other undergraduates including Rachel. We went through like minority owned businesses throughout the six cities. I primarily worked on Richmond and Redwood City. And then when all that work was done I compiled it into the database that I'm going to talk about today. So the data sources that we used Orbus is through available through the UC Berkeley library. It's a massive, massive database that has information on like over 200 million public and private privately owned businesses. There is an indicator for minority ownership information. It does not have information about the group. So we use that as sort of a way to flag a business that we might want to look at later. The payment protection program data, which was mentioned earlier is data that was it's a publicly available data set. And it has information on business owner names, addresses, race and ethnicity category, which we also use to identify minority owned businesses. The Caltrans DBE SMBE data is another publicly available data set. Unfortunately, it doesn't have as many industries that are we're looking at in the study, but it is. It was like one of the most complete sources. So that was nice to work with. We also used publicly available lists. So pre aggregated lists of minority owned businesses in the Bay Area, and we went through them one by one and added them as we went along. We also had lists provided by ethnic business chambers, primarily from Oakland and San Francisco. Again, we went through those out of them one by one. And then we also used ArcGIS Geocoder and Census Bureau shape files to do the mapping. Okay. So this is an example of some of the websites that we used. There's do the Bay, which had lists for like black owned businesses, AAPI owned businesses, Latin X owned businesses. And we went through these one by one by the city that we were working on and like search to see if they were still open operating and then added them to the database. This is another pre compiled source of black owned businesses across the entire Bay Area. Again, we went through this one by one to confirm that they were still open and operating and like which industry that they're a part of. This is another pre compiled source for black owned businesses that we went through business by business and added it to our database. This one. So these lists were nice because they were already sort of like separated by industry so we didn't have to like go through and see like what industry they were a part of like they were restaurants we knew that they would be in that category. So this is the street survey that I went to with Tara, who was a graduate student that was also in the group. We did this as a way to kind of like sense check the data that we had and to, because the Orbis data is also very. There were some businesses that were like out of date or like no longer in operation. The street survey allowed us to identify what was still in business and what wasn't. And we also found like some signage on businesses like storefronts that identified them as like black owned businesses or Latinx owned businesses, which was really helpful because almost all of the businesses that we found this way we didn't find online. So that was like a really nice source that we never would have had had this project just strictly been like online. Okay, and then so putting all of these sources together. So once all of the work was done and we had compiled the information across the six cities. The, we had to identify what was a unique business. So the fields that we use were name address NAICS which is the four digit code that can be mapped to the industries that we were working on. And of course minority ownership group. So the database was assembled by doing an outer merge on the Orbis database, the payment protection program data, Keltrans. And of course what we have manually added through like what we have found online or found on the street. And then so once everything had been merged together into like one big data frame. I went through it and looked at duplicates for everything except for one category. And the reason for doing this is because the NAICS code, so a business would have the same information. Same address, same minority ownership group, but they would have different NAICS codes. And so it's still the same business, but maybe Orbis has them listed as like 7223 a restaurant, but they're listed as 7225 catering and PPP data, but it's still the same business and we don't want to over report that information. So I went through that and then pick those out and this was also how I was able to find like businesses where the owner was part of two or more minority groups, because there wasn't a way to indicate that in the PPP application, but the owner might belong to more than one of the groups that were studying. And so I wanted to make sure that we were able to like capture that information. And then once all of those exact matches have been filtered out, then I did geocoding with the 2020 census data. And I did this so that we could do the mapping and ultimately these were all going to be aggregated by like census tract minority ownership group and industry group. So along the way, I had noticed that there were a ton of like partial duplicates. So for instance, a business would have an ampersand in one source, but they would be spelled out A and D in another source. And if I was trying to look for an exact match, those would never come up as being an exact match, even though in reality they are. But anyway, once everything was geocoded, I could group them by their census tract industry and minority ownership group. And then within those groups, I could search for like partial matches and exact matches by name. The exact matches could be dropped immediately, the partial matches I exported and we looked over them to manually review them. So that way, like, we wouldn't drop any false positives or like lose any information. Oh, okay. So the rest of the Bay Area is almost entirely a pale blue. This is primarily because there's so many like large census tracts with few businesses in them. The only areas that really did have any variation in like the business density was San Francisco and Oakland. And it sort of lines up with what Laura had mentioned earlier about businesses being primarily located in like commercial districts, which you can kind of see are the census tracts on the map that are bold blue. And so like I said before, there is a very, very large number of low density business tracts. So I wanted to kind of explore what the data looked like by city. And when I had just the raw numbers, it was just a bunch of like very, very concentrated columns around zero with like a rapid descent and then a few outliers at the very end. And to make it more presentable I did a log transformation so there is not actually like negative to minority owned businesses per squared mile. Those are the census tracts where there's like between zero and one. So those are the very, very low density tracts. But you can kind of see like San Francisco is and Oakland and San Jose are the cities that tend to have like, they're really the only ones that have anything going on unlike the right side of the graph which is where the high density tracts would be. And you have places like Richmond and South San Francisco and Redwood City, which have like larger, lower density tracts that are kind of concentrated towards the middle and the left side. And then, okay, so I wanted to share this as well of like the same screenshot, because it, I think that it kind of shows like an interesting difference between like the number of businesses in a tract versus like the density and how many are like concentrated in a small space. So there's a triangle there in San Francisco that has something like 80 businesses in that tract. But if you look at the density. It's in one of the like lower categories because it's a much larger tract compared to the smaller census tracts that are like in like, I guess, more to the west of it. So the same case with Oakland as well how there's like a very attract that has a lot of businesses, sort of like south and to the east of it, but the higher density tracts that were lighting up on the graph before were like more towards the west. But you see in reality they don't actually have anywhere near as many businesses as that larger tract does towards the east. And so, because there wasn't as many like extreme outliers in this case I did not do a log transformation. So the numbers that you see here are the actual number of minority owned businesses per tract. You can kind of see like the far outliers that I had mentioned before with Oakland and San Francisco. And again, redwood city Richmond and south San Francisco are have like a more uniform distribution. One thing that I did think that was kind of interesting that I haven't really figured out what this is yet is why redwood city has like such a empty part in the middle and why there's like such a concentration of towards the left and then that one single spike around like 50 or so. But okay. So now, these are the results by industry. You can see that the most popular of the industries are the ones that we were originally looking at which are retail food and beverage services and personal services, arts and accommodation and child gate daycare services were added afterwards but they still, they're still like a pretty decent number of them. And then again by minority ownership group. You can see that almost over half the businesses that we found our Asian owned, followed by about a quarter or black owned and about a fifth or Latin X owned, and the remainders are either still in specified or in the other minority group category. And so again you can kind of see the issue of the extreme outliers with most businesses hovering around like the median of like eight when it comes to the count or four when it comes to the density. And then there are two tracks that kind of pull the mean up farther than away from the median. And that's what we saw in Oakland in San Francisco. So what I learned on working on this project. I learned like a direct application of like time complexity which is something that I had learned back in like my data structures class, but it's, it's the issue of when you have to do a lot of computations over and over and over again on the same thing. So for instance, when we were looking for duplicates to do that you have to compare a business name to all the other business names in the city. And over and over again, if I had done this early on before geocoding, this would have required I think something like 11 million comparisons, which is like, it took a while to run and it's not the best way to do things. But, and there's really no way, like, using like Python libraries to make that easier. But if you have things geocoded first, you can group them by census tract industry group and minority ownership group. And then that turns. So now you have like, about 600 different census tracks in the Bay Area. And you can see at most there's 89 businesses per tract. So even in the worst case when you're doing 89 comparisons 600 times, that's like 55,000. And that is far, far, far less than 11 million. And like we already know that the 89 is an extreme outlier so you're really only making closer to like eight comparisons 600 times, which is it was a lot faster than doing it the original way where it was grouped by city. The last thing that I learned is that not having any skeleton code or any like pre existing project or database to like go off of meant that like, I had to create this just myself, and that opened a lot of problems when it came to the fact that like we had to expand the study or add new data sources so I spent a day of just like rewriting all of my code and made it to where anybody else could run this if you know they had like the Python libraries installed, but it's now it's just like six scripts so the output of one is the input the next one. And just kind of in the interest of like looking forward and like making this easier and making this that some thing that somebody else could use and like make use out of. So yeah. It's interesting to look at all the technical complexities of study because it emphasizes why this probably hasn't been done before and how kind of time consuming it has been but we've learned a lot through this process. So I, I'm kind of going to talk through some of our kind of emerging findings because this study is still ongoing. So we've identified concentrations of BIPOC owned businesses in Oakland and San Francisco, and this reflects the concentrations of businesses generally within these locations. And also the known cost us so we know that there are black owned businesses concentrated in Oakland. We know there's Asian owned businesses concentrated in the China towns of San Francisco and Oakland. And we know that there are Latinx businesses in the mission and also around Fruitvale and Oakland. So the mapping has kind of confirmed those things that we already knew. Through some of the qualitative work businesses have identified the need for small business support systems. They've also identified the need to improve access to capital and also mechanisms to prevent property displacement. We're currently looking at policy solutions to address these needs and we'll be working with stakeholders in the coming weeks to have sort of focus group type webinars to develop up these policy recommendations. This is a substantial amount of data data out there, but it's often difficult to access, or there are a large number of gaps. So some organizations we have spoken to have access to data but they do not necessarily have the resources to manage it so they might only be one person who's employed. And that's not their sole purpose, but they have this database. So accessing it and keeping it updated is quite complicated. And that has been a constraint that we've come across. Also data availability and quality is varied across different types of businesses and locations. So we found that some locations we were able to access a much larger quantity of data than some of the others. So an example is Oakland. We found that we were able to access a lot more data sources for Oakland. Outreach has been very critical in accessing data and understanding more about the local context. And that's been an invaluable step throughout the process and it continues to help refine the analysis and plug gaps in the database. The mapping reinforces what we already know, and this is important in telling the story and using this to kind of promote policy interventions, but it's not the only story and qualitative research is important to this process too. And so this project is only the beginning and this potential for the database to be expanded in the future. And we're also looking at ways that the database can be kind of handed over to a key stakeholder to be kind of regularly updated and made available for public policy purposes. So we have time for questions. I thought maybe before that I might just show the actual maps so you can see how they work. So I'm just going to, so this is the business database map. And this shows the concentrations in terms of numbers of businesses. And at the moment you can kind of click on to a census track, see the total minority on businesses and number per square mile. You can see the breakdown by type of minority groups. So there's a concentration of Asian-owned businesses. This is around Chinatown in San Francisco. So that makes sense. There's also some black owned and Latinx owned as well. Can move over to Oakland and see concentrations as well. I believe that's also around Chinatown. And then the other layer we have on here is the density. So you can also see the different illustrations there. The other map, which is the one that Rachel spoke through. This is the early neighborhood concentrations map. It's the same sort of thing you can scroll through and click on a zip code and find out more about it. So I'll leave it on this one. Are there any questions here? Yes. So I have a question on, at the end you talked about adding qualitative data to the map. I'm just curious like what kind of, if you know what kind of qualitative data you're looking for or how you care exactly in terms of this project. I'm curious to see what kind of information or how you would just incorporate it into this database. So one way that you can kind of feed through is through the street surveys, which I guess is somewhat of a qualitative research method that when teams went out, then we filtered that information back through to say like these businesses are closed or operational. The more qualitative policy recommendations will be, we anticipate will be kind of a policy brief type report that will be used in conjunction with the database, but it won't necessarily be mapped physically onto the database, but it'd be that we would have hopefully a website where they're both there together so that people can read the policy recommendations in conjunction with viewing the map. Thank you. This is an amazing project and we have done a ton of work. And so I just want to commend you on this. This is a really important tool that you're creating. And I just thought I asked a little bit more about what you have planned for it. In watching your presentation, it reminded me of another mapping tool that I'm familiar with that is designed by one of our research centers here, Berkeley Interdisciplinary Migration Initiative, and they designed a tool to identify where immigrant populations are and where the services for those populations are. And what they showed was a spatial mismatch in terms of if the resources are here, the populations are here. Is that possible for your tool as well? Is that something you're thinking about in terms of identifying maybe a spatial mismatch or an opportunity for if we had a small business association, you know, if we're going to add one, it makes sense that it goes here where there's the highest concentration. I mean, are those things that you're thinking about? So that's one question. Then another opportunity here is, and I just want to clarify, did you identify businesses that received PPP paycheck protection plan? Okay. Do you have that data? Could you show it to us? Because that is something that our students have been also working on. You're talking about qualitative research. A lot of our projects involve students talking to business owners and getting their stories about why they did or did not receive PPP. And it would be really important, I think, to add another layer, which is the PPP layer. Okay, here are the concentrations. Who got the PPP and who didn't get the PPP or it could be California Relief Grants or something else. Is that also in the works as well? And I guess I'll just stop there. Thank you. So the first question about mapping kind of spatial mismatch, so that's definitely something we can look at. And we have various sources or different layers we can kind of produce around where vulnerable populations are located and where these businesses are as well. So that's something we can definitely look into as well. In terms of the PPP data, we have that. We haven't mapped it specifically, but we do have that in the database. That's something we could interrogate a bit more to look at that variation and whether there's particular types of businesses in terms of industries, types of minority groups or types of physical locations so we can kind of break it down and compare. So that that will definitely be an additional insight that that will be important. So it's a good suggestion. Thank you. We do have. Oh, yes. Sorry. Hello. It's fun to have the mic. This is super interesting. Last year we spent eight weeks talking about vulnerable populations and COVID-19 and consulting like a lot of the same databases and there's a lot of work around that and like imagining what a vulnerable person would look like. And then I kind of took that same research and started looking at a more qualitative approach to like certain businesses in the Bay Area. And one thing that I found particularly interesting was the story of Chinatown. I'm from the Bay Area and like I've lived across the East Bay and a lot of the cities that you guys are studying. And I think what's really interesting about San Francisco and the Bay Area in general is that we have two really prominent Chinatowns. And I'm wondering what it was like researching what happened to those areas specifically. I know I understand that there was some qualitative research done there. I'm wondering if like the influx of hate crimes and just also like a lot of negative media and how that translates to a business owner or just an individual living in that area, but particularly a business owner, how that may show up or like any further insight like those businesses in Chinatown and Oakland and San Francisco. Yes. So we haven't spent too much time in looking at that specific issue yet. But I think through the policy discussions that we're going to be having with key stakeholders who represent some of those locations we'll be able to understand more about that and make sure that's integrated in the policy brief. And we're very aware of that issue, which is why we made sure we incorporated the Asian-owned businesses in the study because we thought that they've been particularly impacted by COVID-19 and associated discrimination. So a very, very important point and something we're very conscious of looking at. Yes. I wonder if it's possible for you to follow the businesses that exit your data set. So we often create arguments about what helps a successful business by surveying existing businesses, but we know very little about those that die. Yes. And you have this amazing opportunity to, as they leave your data set with all this information, who's more likely to die, where they're more likely to die, what sort of areas. And then if you could continue that conversation, you can perhaps even survey them about why, et cetera, what were the challenges. That would also, instead of just cleaning the data, so allow you to get a nice data set, but also give you just this invaluable piece of information of folks that sort of exit. Yes. Yes. So through what we're developing at the moment, we do have some indication of business closures based on some of the information provided through the kind of Google Maps and Yelp and some of those places, but it's not consistent. And that's definitely something moving forward that we could make sure is a kind of key outcome for when the database is updated. It's something that I'm looking at in my PhD research, which is focused on something different, but it's very difficult to track businesses once they get physically displaced because you don't know where they end up, let alone what they've closed altogether because then it's hard to contact them. So databases such as this that can kind of start to track them in a longitudinal way are important. So that's something we can definitely look to incorporate into how this is managed moving forward. And that's something we're trying to work out is who takes ownership of this database and continues to update it in the future, because it takes a lot of time and resources to do that. Yes. I have one more question if that's okay. So how did you decide on, because as you were scrolling over the map, you had different minority groups. How did you decide which ones you had picked? Because then there was a section for other minority groups. Did you go off like census categories? Because I know like with census, there's a lot of controversy around, you know, certain Middle Eastern groups being counted as white, even though they're not just like historically being counted as white, even though the experience different, they have different experiences, especially even in business than like white business owners. So I'm just kind of curious like how you decided which ones to categorize and then which ones to put under other. And like if that was related to the, I guess, census groups, because then that would provide context. So other often includes whether it's two or more minority groups or that's the main source. We didn't specifically include indigenous yet because there's very limited data available. So we didn't want to suggest that we were capturing all of that because we know we're not capturing all of that yet. Asian, I believe, includes like all of the continent of Asia, those sorts of groups. So it would include kind of Middle Eastern, I believe, but that's something we can definitely kind of make clear when we're reporting on this. And we had to kind of constrain the study and recognize, we do recognize we're not capturing all. And that's a significant constraint from doing the studies. There's a lot of businesses that don't identify as minority even though they are for various reasons. So even in the PPP data, there's a lot of businesses that are unclassified. So they're not necessarily because they're not minority and they just choose not to identify. And there's various reasons for that. And that's something that we're aware of. And we're trying to capture as many businesses as we can, but some businesses we know will not be captured in the study. Yes, so there's a question in the Q&A thing. What is the total number of black land businesses in baby hunters 0.94124? We won't. This is not at the zip code level. This is at the census tract level. I do not know where Bayview is. San Francisco? Is it around here? I'm not sure if anyone knows the exact census tract number for Bayview. I mean, this is quite a large sense. Here's a large costar. I'm not sure if this is a specific place. But we've identified 32 in this specific census tract. If we go to the zip code level map, I might be able to find that's the zip code there. This is not the actual number. This is our estimated number based on the national level statistics for that zip code. So we think that there would be 31%. Something we can do is aggregate our business database to the zip code level and we can compare. We just haven't gotten to that stage yet. But that would be useful in comparing if there are zip codes where we've actually found more than what the national data would suggest and some that is a lower as well. Any final questions before we wrap up? Thank you very much for your time and also for the great questions because this study is still ongoing. That helps with forming some of the ideas of how we finalize this study. So I appreciate all those questions. Thank you.