4/29/2011 5:13:58 AM By
E.J. Reedy
The
National Establishment Time-Series (NETS) Database, a longitudinal database of linked Dun & Bradstreet records, is one of the standout data sources to emerge in the last half decade for the study of entrepreneurship (and businesses, more generally). As an early proponent of the potential of this data, I wanted to offer some commentary now that comes from grantees that have used the NETS data for different purposes such as matching to other data sets, surveys, or just aggregate examinations of firm dynamics. Every data set has its advantages and disadvantages, NETS is no anomaly in this regard, but one of the unique things about my role is that I get to see some of these trends and connect experiences. These comments emerged as the result of an email inquiry I had made to those listed here for a potential new project. I found the comments so valuable I decided they should be published as blog post so that others could see them. All the commentators were allowed to review and modify their comments before posting. I would encourage others who have experiences that can help to hone the use of NETS and hopefully drive improvements to its core to add them as comments to this post.
“The strengths of the NETS are its historical address information, descriptive data (industry, incorporation codes, and CEO and ownership demographics) and survival measures. These traits seem sufficient to conduct event history analysis using quasi-experimental methods.
If your focus is on other types of performance such as growth, then I'm not convinced this is the best dataset. I found in my research that employment and sales figures in the NETS are often estimates and not actual observations of the levels of these measures. It presents a problem when you are dealing with small firms because you lose variation since so many of these small firms appear to reach stability over short periods of time. I'm still using the data for this purpose but I take the results to be more speculative than definitive.
Also, the lack of information on other attributes of organizations (i.e. invested capital, costs, revenue streams, and the education and experience of CEOs) makes it hard to account for unobservable differences that may be the underlying cause of performance when evaluating a policy outcome. There are statistical methods to address these issues but these also open the door to skepticism and criticism.
I'm working on ways to improve on the value of the NETS by merging it with other datasets to add richer data to what I already use. Depending on the types of firms being studied, one may improve on the types of controls that can be accounted for and on the type of performance being evaluated.”
“I agree with Alejandro that NETS is not the best data to see changes in firm level outcomes. In our work we found that the information on sales and credit scores are actually the best of their outcomes variables, and are updated reasonably well. But especially employment variable are very sticky in NETS and do not seem well covered in NETS.
One of the best data set that you could use is probably LBD data at Census, but this data is difficult to access since you need to apply for access.”
“I am less skeptical of the employment measures than Alejandro and Antoinette, although it depends how you are using the data. At the individual establishment level there is clearly a lot of stickiness. But when you are averaging across many establishments (like in our
EZ paper) this doesn't really matter. I think that if you are matching based on geographic location the NETS are very useful although the geocoded information that comes provided with the data might need to be re-geocoded.”
See select other posts dealing with NETS:
Comparing Business Registers;
youreconomy.org Updated
3/23/2011 10:49:47 AM By
E.J. Reedy
The Census Bureau has released new tabulations for its Business Dynamics Series that show in stark detail how the recession has not only significantly impacted job destruction in the U.S. but also the rate of job creation. 2009 brought job creation rates in the U.S. to their lowest level on record (in 29 years).
Read an overview report or
explore the data (which is available by SIC and state, as well as what is included in the report).
3/7/2011 7:52:55 AM By
E.J. Reedy
This morning we released the
2010 Kauffman Index of Entrepreneurial Activity. The Kauffman Index is a rather unique measure of entrepreneurship as it relies on a quirky design of the U.S. Current Population Survey to come up with a monthly measure of the number of households transitioning into entrepreneurship from other labor force statuses.
Since we are working from a household survey, the Kauffman Index has the advantage of being able to give very detailed demographic information about who is becoming an entrepreneur. Indeed, it's from this perspective that these current year findings arise:
- The immigrant rate of entrepreneurial activity increased substantially – from 0.51 percent in 2009 to 0.62 percent in 2010 – and declined slightly for the native-born. This increase expanded the large positive gap that already existed between immigrant and native-born entrepreneurial activity rates.
- A growing immigrant population and rising entrepreneurship rate contributed to a rise in the share of new entrepreneurs that are immigrant, from 13.4 percent in 1996 to 29.5 percent in 2010.
- Entrepreneurial activity increased slightly for men and decreased slightly for women. For men, the entrepreneurial activity rate increased from 0.43 percent in 2009 to 0.44 percent in 2010. The female entrepreneurship rate decreased from 0.25 percent to 0.24 percent.
- The African-American entrepreneurial activity rate decreased from 0.27 percent in 2009 to 0.24 percent in 2010. The white entrepreneurial activity rate decreased from 0.33 percent to 0.31 percent.
- The entrepreneurship index was highest among the least-educated group, moving from 0.49 percent in 2009 to 0.59 percent in 2010, suggesting an increased number of people entering entrepreneurship out of necessity. The largest decrease in entrepreneurial activity occurred for high school graduates.
On Data Maven, I wanted to specifically point out two new features of this year's report. First, and most importantly, is the expansion of the report to utilize the Business Employment Dynamics (BED) series data from the Bureau of Labor Statistics. One of the limitations of the CPS as a data source is that we are not able to disaggregate the type of businesses whose start is captured in the Kauffman Index. Thus, the Kauffman Index captures both transitions to self-employment as well as the start of larger businesses. In an attempt to add more depth of understanding to the current state of entrepreneurship in the United States, an aggregate measure of new employer establishment starts was computed this year from the BED. The resulting picture, shown below, gives a much more accurate national picture, in my view:
Taking the Kauffman Index and BED-based measures both into account gives the picture of an economy with the highest number of people becoming entrepreneurs on a monthly basis but most of them only going into self-employment or starting lower employment-potential businesses.
The second change that I wanted to highlight this year was something which
I originally wrote about last year in a blog post - a look at the changing look of entrepreneurs, in the aggregate. This takes into account both the proclivity to become an entrepreneur (the Kauffman Index) and also the changing demographic composition of the U.S. to look at changing total numbers of new entrepreneurs. For three categories in particular, this measure is quite telling.
3/7/2011 7:20:08 AM By
E.J. Reedy
On the heals of the
Kauffman Index release, I want to call on the Census Bureau to produce more information on immigrant entrepreneurs. In my other posting I pointed to some of the big shifts which have occurred in the composition of new entrepreneurs over the last decade but I think it's worth repeating that here:
As the Kauffman Index shows
immigrant entrepreneurs made up almost 30 of all new entrepreneurs in 2010, more than doubling over the last decade and a half, and yet this is a group which we know very little about systematically at the national level and especially at the sub-national level. There are competing streams of research (see
Wadwha, et. al and
Hart, et. al), some of which Kauffman has funded, that point alternatively to the importance or the normalcy of immigrant entrepreneurs. Regardless of which stream of research you believe more, the Kauffman Index numbers make it strikingly apparent that we need to know more about rapidly growing population of new entrepreneurs.
I specifically suggest to Census that it consider adding a report to the current production schedule for the
Survey of Business Owners which is the largest survey of small business owners. In 2007 this survey added a question about immigrant status of business owners (a good move!) but currently there are no plans to specifically provide a detailed overview of immigrant entrepreneurs (see
SBO release schedule here). The SBO already provides detailed overviews of Black, Hispanic, Native American, American Indian or Alaska Natives, Native Hawaiian or Pacific Islanders, Asian, Women, and Veteran-owned businesses. The immigrant data will be part of the June release of the report "Characteristics of Business Owners" but today I want to call on Census to consider adding a separate and more detailed Immigrant overview. The SBO could provide detailed sub-national estimates of the type and impact of immigrant-owned businesses and help to understand what has been a seemingly large shift in the composition of new entrepreneurs. The data exists; all that is needed is a recognition of this important group among entrepreneurs.
2/20/2011 11:15:54 PM By
E.J. Reedy
The Wall Street Journal has a
nice post on their blog looking at Google's new tool for searching the text used in printed texts over the last several hundred years,
Books Ngram. Take for example the graphic the WSJ put together examining the use of "entrepreneur, small business, and self employed" in texts over time.
I have done similar comparisons of language used in different settings over the years. With searchable Abobe files it's pretty easy to do now. The one I most remember was a review of the language used in the Economic Report of the President, an annual publication on the economy in the U.S. It was very clear from a review of language there that there had been a shift in language used over time from "small business" to "entrepreneurship" in the late 1990s and then continuing in the 2000s. In that publication there was very little mention of the term self-employment, which the Google tool also found. That said, I am surprised the Google Ngram tool found quite as low of levels for "self employment" as it did in the recent decades as that is a terms still fairly prevalent in some academic literature.
I most recently looked at some of Google's other language-use/charting tools for some work with the
Organization for Economic Co-operation and Development (OECD) where we are looking to develop better indicators of entrepreneurship in culture. This new Ngram tool holds some good potential for further development for this type of indicator but from what I saw on their website there isn't any way to indicate where the publication occurred. That'd really be necessary for use in something like an OECD country-level measure.
2/15/2011 8:32:56 AM By
E.J. Reedy
Is job-lock (the locking of a person into a specific job beyond when they would like to be there) occurring in the U.S.? Are entrepreneurs being forced to stay in wage jobs, working for someone else, because of the fear of losing their health insurance? Below are a few pieces of research/data sources that I've come across recently that are attempting to make research contributions to this question. But before I comment on them and their source data, it has to be said that all of these research efforts are attempting to get at the impact of health insurance provision on entrepreneurial entry in the United States but because of limitations in how the data is collected none of the efforts really gets at the
real issue. Simply put, there is a belief that more potentially innovative nascent entrepreneurs are kept from starting a business (or doing so full-time) because they are tied to a job in which they receive health insurance. None of these data sources/research is able to ask this exact question and realizing the limits of their data to the analysis is critical. Someday, I hope there will be some data that can really inform this question more directly, but I am not currently aware of any such data in development. I certainly am of the crowd that believes this problem to be huge.
- Two sets of authors (Maria Minniti and Yunwei Gai; Ian Michael Breunig) have papers out using the Medical Expenditures Panel Survey at Department of Health and Human Services. This data is rich in that it gives information on where the insurance is coming from and also the work experience of all primary members of the house. The downside of this data in my mind is that it forces the authors to use a self-employment measure for entrepreneurship, which is not typically associated with high-growth firms.
- Rob Fairlie, Kanika Kapur, and Susan Gates in a forthcoming journal article Is Employer-Based Health Insurance a Barrier to Entrepreneurship? take a look at the Current Population Survey and specifically make use of it's unique design to track transitions to self-employment around the age 65. As such the authors are able to infer inhibitions at the time individuals become eligible for government-backed insurance and make attempts to measure job lock more broadly from this population.
- Scott Shane and Alicia Robb have analyzed the Kauffman Firm Survey questions asked of its panel of new businesses started in 2004 and finds little evidence of significant annual change in health insurance provision among the panel.
Past posts related to health insurance:
2/14/2011 5:58:00 AM By
E.J. Reedy
The events of the last month in Egypt and the Middle East have had me thinking some about the forces that can drive change. One of strange parts of about the space that I occupy, not being in a stat agency and not being an academic, is I that get an outsider’s perspective of what is happening within and across these institutions. This is particularly helpful in watching how change happens within these organizations. I wanted to reflect a bit on some observations about the global statistical system, specifically the informal international network of governmental statistical offices and officials that produce the data which we rely upon to understand our economies, our people, and all aspects of our world.
I don’t know what I’d call these musings but I think when looked at together there are some truths here that can be helpful in particular to academics who might over the course of their careers only have an opportunity to work with statistical agencies once or twice. I have talked to a lot of academics (and others) who might want to see improvements in certain data, release of certain data, coordinated collection of surveys, or the like. Most academics are befuddled by national statistical offices and their bureaucratic ways. There is little academic training given to understanding project management, partner coordination, etc. – all things which are critical to getting data collected, produced, etc. Consider this a short primer but also a larger consideration about how the system informally operates.
While we would like to pretend that all countries are playing with an identical deck of power and influence in national statistics, the reality is that three players wield unusual power and influence on the direction of the global statistical system. Innovative measurements and production of new and meaningful series by a relatively small country, say Chile or Belarus, are very unlikely to be replicated or noticed internationally without the attention or interest of one of the three trump statistical players. Recognizing where and how power centers are shifting in the statistical community is important for those of us trying to drive improvements in data and for researchers attempting to see cross-country studies replicated.
The Fading Trump
The first trump that I recognized in my job was the United States. Simply put, in the production of innovation and entrepreneurship data, if equivalent data is not produced on the United States (and if people don’t believe the results), any international comparison is doomed. As a Foundation whose main focus is production of data on these topics in the U.S., we were often approached about funding or supporting U.S. replications of data series and research. While this was a particularly powerful position for the U.S. in the 2000s, post-crisis it seems this trump has diminished. Now, while most efforts would like to have the U.S.’s involvement, countries have tended to have more interest in more timely and meaningful national-level data, with much less importance given to comparing their results to the U.S. benchmark. The U.S. has never fully taken advantage of its potential power as a statistical leader, too often relinquishing involvement in international efforts to gather improved data. Driving the global conversation was a space the Canadians really saw outsized-power with until recently with the gutting of many of their statistics programs.
The Suit Bigger than All the Others
The quiet behemoth of internationally comparable statistics is Eurostat, the European Union’s statistical office. While they have wielded some power since their establishment, my own experience is that just as the European Union’s powers have broadened in policy and other rights, Eurostat’s power to drive the international statistical conversation has increased. In our own experience, the signing-on of Eurostat to the Entrepreneurship Indicators project at the OECD, which we funded, was a significant milestone in driving internationally-replicative data. With the ability to drive statistics coming from 27 member countries, Eurostat is both blessed and cursed. The blessings in terms of potential replicants are easy to see but the curses are a bit more hidden. One significant curse is the limited ability of Eurostat to drive implementations which ALL member countries don’t agree to. Hence, the sovereignty of data production remains at the state-level and can sometimes lead to the dumbing down of mandatory statistics. A lack of microdata access and a mandate that is too inwardly focused have limited their global power.
The Rising Trump
While no expert would currently identify China as a global statistical leader, from what I see in the academic world I have no doubt its power in global statistics is also rising. While researchers doing cross-country studies of entrepreneurship and innovation are a limited bunch, it feels as if almost all the proposals I see for collections in this area involve China and X. And with the OECD’s increased focus on non-member economies, it is certainly trying to get China and other BRIC national statistical offices to join the global conversation. So far, this seems about the only potential international source of power that China has not systematically tried to take advantage of in its development efforts. From my understanding and time living/traveling there it would seem there is still some distrust of official statistics and too much information about the activities within the country. India is a country that has a huge history of measurement and record keeping but still lacks many of the basic infrastructures for modernization.
2/11/2011 6:03:58 AM By
E.J. Reedy
On the heels of my posting earlier this week describing
global trends in small business lending, the U.S. has
new national and state-level data available today from the Small Business Administration. While I would like to say that this data is about "small businesses" I would reminder readers that really these are data about "small loans" (under $1,000,000) and really small loans (under $100,000). While this might be a proxy for small business lending, I don't think anyone can legitimately argue that it's a full representation of the reality. That said, it's the best we have in the U.S. currently. Arg.
The SBA is reporting a smaller decrease in all categories measured than was seen in 2008-9. But a decrease is still a decrease; it seems the mid-tier loans ($100,000-$1,000,000) continued to be the hardest hit. Specifically the study found "the smallest business loans under $100,000 began to stabilize in 2009-2010—the total was down by 1 percent...[overall] small business lending dropped by 6.2 percent, less than the 8.9 percent drop experienced in large firm lending over the 2009-2010 period."
One new feature of the SBA's website this year seems to be better automation of the data tables and information underlying the state-level data. I think this is something that many scholars studying state-level effects will have some interest in. The data still suffers greatly from the bias of not knowing where the businesses were located that received the loans. For example, Bank of America only shows up in the North Carolina data (from those spreadsheets I examined) since it's headquarters is there. Still, I like the increasing accessibility of this data from the SBA.
2/9/2011 8:21:53 AM By
E.J. Reedy
For those of you who are not experts in data dissemination but want a crash course, start with tomorrow's event in Washington, DC.
"Responsible Data Sharing in the 21st Century" will be hosted by University of Chicago NORC and the National Institute for Standards and Technology. Sadly, I can't attend in person, but am happy to report my colleague Alicia Robb will be presenting. She'll be discussing our
private sector experience in disseminating the Kauffman Firm Survey but most of the speakers are from government and will be discussing their various experiences.
I really hope people are coming to learn because the truth is that as a community of data stewards we are still making many more failures in disseminating our data than successes. Here I don't mean failure in the sense of a data breach but rather the science behind what scholars and communities need in addition to data access is still in its infancy. With the Kauffman Firm Survey we've tried to be a real-time laboratory on these activities, testing new modes of dissemination, programs to support outreach, as well as coaching and training. So we sincerely hope others can learn from our successes and failures in data dissemination and we look forward to hearing about similar experiences.
2/8/2011 6:04:56 AM By
E.J. Reedy
I have spent a good deal of time over the last four years working on surveys asking individual companies about how their business activities are financed. It’s been amazing in that process to see how differently the topic must be discussed in Europe and the United States. Specifically, topics like credit card financing, which are a mainstay of U.S. surveys, don’t event make an extended pick-list of financing types used by most European small businesses. I bring this up to illustrate that international comparisons of financing mechanisms for small business are a tricky thing. And yet, we all yearn for more comparable data (or any data at all!). Today I wanted to describe a couple of fairly recent surveys focused on small business financing data internationally.
Asking the Firm about Financing
Completed every six months, the
European Central Bank Access to Finance of SMEs is a large-scale survey of establishments in the EU zone looking at changes in conditions but not gathering nominal values of loans. It has a nice
questionnaire that keeps things simple although the overall effort is a bit limited in that it's bypassing official statistical systems and doesn't have the ability, as I understand it, to link reported credit conditions to firm outcomes in later periods. It's most recent report summarized that “between March and September 2010 the proportion of SMEs reporting a worsening in access to bank loans, at 24%, almost halved compared to the previous survey, when it stood at 42%. At the same time, 12% of SMEs reported an improvement in access to bank loans, compared with 10% in the previous round.” This seems to show that the debt financing options for European SMEs are still getting worse but not as quickly. This is a good example of a survey gathering perceptions from businesses. The recently reviewed
NFIB survey also comes to mind here.
Asking the Regulator (or Bank) about Financing
The other route that can be followed in gathering data is to go to the providers of capital. In most cases, in small business financing this means looking to providers of debt capital. A new paper by the World Bank titled “
Small and Medium Enterprises: A Cross-Country Analysis with a New Data Set” describes an aggregation of SME financing data of this types from multiple countries around the world. This arises from some of the efforts of the World Bank and G-20 to aggregate data from central banks and other administering agencies about the size and quantity of loans to small businesses in their country. Asking regulators about loans is quite different than the above described establishment surveys. The advantage of the second approach is that it’s much cheaper but as the authors outline, there are huge definitional differences across countries in how SMEs are defined.
From
table 2 of “Small and Medium Enterprises: A Cross-Country Analysis with a New Data Set”
The above shows the U.S. definition which centers on loan size, not employees or revenues of the business; I find this to be highly unsatisfactory. So while I applaud this G-20/World Bank effort, I personal still don’t think it’s a full substitute for a quality establishment survey. That said, outlining some of these differences in definitions is the first step in pushing towards further agreement on appropriate definition and comparability. The authors have tried to make global estimates on the size of small business lending, finding that $10 trillion flows annually to small business loans but that this is disproportionately higher in developed economies. Additionally, they believe the differences in definitions are not driving the difference between developed and developing economy levels of small business financing.
A Gripe
While there is much to applaud in the ECB survey, they still have work to do in regard to their sample. Specifically, it does not appear to me that their sample is adequately covering new and young firms. While they have about 2 percent of firms responding that they are under 2 years of age, that is dwarfed by those firms age 2-4 years (8 percent of their sample). This would appear low by all estimates I can think of coming out of the OECD and other international sources. And indeed, in this turbulent time we have shown the importance of tracking new business finance separately as these are the firms most likely first effected by the downturn and any credit tightening. Take a look at Spain’s statistics on new business registrations, as an example, which have halved in the last year. The discussions about small business financings too often are portrayed as the same as new business financing. They are not and more effort needs to be made to have samples which allow for representative studies of credit conditions faced by new businesses.
A Compliment and A Call
But it’s great to be squabbling on the details of the ECB effort while the U.S. remains devoid of meaningful ongoing federal collections at the establishment level. Our own Kauffman Firm Survey is the only longitudinal effort collecting annual debt and equity injection information on U.S. businesses but it’s only one cohort. How we are not demanding more on business financing is baffling to me. Communities everywhere in the U.S. want to know what is happening with their small business credit conditions.