Jack S. Futterman Oral History

Part II- Working In The Candler Building

Coming To Work In The Candler Building


So that was my beginning. I came into the Candler Building, which was not a very clean building. It was a warehouse building that was made for heavy industry. Not all the nine floors, except the ones we were on, even had wood flooring. The ones that did had a kind of flooring with sand under the flooring on top of the heavy cement floor. People used to complain about sand fleas and things like that. We chose the building because a consultant, I'm sure, persons like Elwood J. Way, who was very much used as an consultant for office things, knew that we were going to have a lot of heavy steel cabinets. I am guessing about this, I was not in on this decision, it took place before I first reported on November 16, 1936. After awhile, it didn't take very long, we started putting the filing cabinets two high. They were very heavy cabinets and they double-decked them. And eventually they had to have ladders that we rolled up so that you could look into and use them. So those floors had to be able to safely sustain that kind of load. As I say, it was constructed to bear very heavy floor loadings.

Q: What floors were we on? Did we have two or three floors?


We had the 3rd, 4th, and maybe the 5th and 9th. There were several manufacturing plants there. One was Hershey Pants, they made slacks. They had a reputation for being one of the finest slacks then made. Then there was some outfit that brewed "Witches Brew." That's all that I can remember about it. It had a noxious order, it was a chemical, a medicinal kind of thing, and the odor would come up the elevator shafts and was very strong. There were not too many other occupants. We did not occupy the whole building, but we had a major part of it.

On the bottom floor there was a restaurant, a cafeteria, called the "Alderney," as I recall. And there was a bank, a store-front bank, which was a predecessor of the Nations Bank, which took over the Maryland National Bank. That bank, as I recall, was then named the Bank of Baltimore, and I have been using it since 1936. After several mergers it was most recently taken over by Nations Bank.

Q: Now this cafeteria, was this an SSA cafeteria?


No it was a commercial one.

Q: Did we have a cafeteria of our own in those day?


No, not in those earliest days. Far from it. Eventually, an enterprising middle-aged fellow we called "Smitty" was allowed to come through our space with a basket on his arm containing bologna or American cheese sandwiches that he prepared and sell them for something like ten cents.

Q: Now how did we organize the Division of Accounting Operations (DAO). The impression that I have of it is that different functions were on different floors. Is that right?


That is right. There were large groups, you see, they had to be brought together. I can't give you the original breakdown, but I can indicate what it was like. In those days the organization had to be very fluid. Very flexible and very fluid. Why? What was the work that was to be done? Well, the first thing was to take these applications for Social Security numbers that had been given to the Post Office and the cards to which the Social Security card had been attached.

Q: A receipt?


No, the Social Security card and the office record were one form, I believe, and the Social Security card was a tear-off section of it. The Social Security card was as it is today, basically, with a name and a pre-printed number.

The work was the initial job of enumeration that had begun in the Post Office. Using the Post Offices was the only way that it could be done. We couldn't set up offices, we had to use already existing offices. So they enlisted the Post Offices. Accounting Operations got from the Post Office the application form, called the SS-5 application form, and also the office record card, that I just described. The office record and the SS-5 form gave us the name of the individual, their address, if it was a woman, her maiden name, also date of birth, place of birth, father's name and mother's maiden name, and date of issuance. The first job we had was to code the record. Are you familiar with the code?

Q: Are you talking about the Soundex Code?


Soundex code. That was the first job that we had to do.

Q: Is that what you did when you started out, you were a coder?


I was a coder. I'll have to come back to that, because I'm trying to set the stage. We set up our records using the Office Records and SS-5 applications we received from the Post Offices. The scheme of operation was, obviously, not formulated in a significant detail. It was highly generalized, not detailed. We improvised a great deal. The whole organization was operating often without a roadmap. Some "experts" had predicted that the record system just wouldn't work. Life-time earnings records had not been done before. So in the earliest days they made it up as they went along.

Q: That's an interesting point to me. What we were trying to do here, this record keeping operation, was really unprecedented and we came here with a lot of skepticism by people.


I'll go back to that, or otherwise we will leave a lot of loose ends.

Q: Alright.


So, skipping most of the details, the whole thing was first to set up records, and then secondly to begin to take the data that was flowing in terms of earnings in employment and post them to that record and then eventually somebody would have to take claims, and etc, etc. But that was in the distant future, the immediate job was to confront these 30 million accounts--I think it was 26 million originally, and then increased to 30 million very quickly--30 million accounts and set them up. And I just spent a little time on it, to give you the flavor for the kind of people that worked in this circus.

The first thing was to code the records. And coding the last name consisted of using the Soundex code to produce the "name code." The Soundex code is comprised of the first letter of the last name, and a series of three numbers. The six groups of similar-sounding consonants are each represented by a number from one to six. You skipped over the vowels-they were not encoded. I'm not as expert as I used to be, but as I remember my name was "F," the "T" was a three (you didn't code both T's) and the "R" was a six, and I think the "M" was a five. So F365 was the Soundex code for Futterman. DeWitt would be D300, if I remember correctly.

Now it may not make a lot of sense, in the case of DeWitt, but it sure made a lot of sense in the case of possible spelling variations of names like Smith, Schmitt, Smythe, and all the other many variations. And the reason why a file in Soundex code order made sense was that our alphabetic file of office records (the portion left after the account number card was issued) was an inefficient and ineffective record arrangement to use to locate particular records falling in a large group having a common last name or one which often was spelled differently. The application form was filed by the account number. So we had an account number file and then they had the alphabetic file. If you looked in an alphabetic file for a Smith's account number, the Smythe records could be maybe 20 cabinets away and some other variations of Smith might not even be on the same aisle. The virtue of filing names by Soundex code was to throw similar sounding variations of a name together. The sequence was first the code, then the first name and initial and date of birth. In such a file one can find most people's records in seconds and the workers in the so-called "Flexoline File" could usually do it very quickly.

So Smythe, Smith, and Smothers all those were all put together. The next break would be to file them and sort them by first name. A name like Aaron would be number 1 on the list--double A. And all the Aaron's would be together, and then the date of birth. First the year, than the month and the day. So with the date of birth and the person's name we could find the right record--even though there might have been problems: people changing or modifying their name; it's not at all uncommon for people to do this. But if the name variant was within that same code we would be able to quickly locate them in the file.

Eventually after several years of maintaining the alphabetic file, which we found required much more effort than it was worth, we discarded the whole file and relied wholly on the Flexoline File which was arranged in Soundex code order.

Q: Went just to the account number?


No, account number and the Flexoline File. Do you know about the Flexoline File?

Q: I know what it is, but tell us a bit.


The Flexoline File, we already had the SS-5, which was a form that people filled out, they still fill out. It was that form we coded with the Soundex code, and then we punched up cards with that information, that comprised the master card accounting machine file. The master card file, by account number, had all the information that was on that form, plus other things that we added like the account number. Well the account number was on the card form, with the date of issuance and things like that. We used that card to machine-prepare the "flexoline" strips, one line deep, which you could separate. The initial run was in Soundex order so that required no inter-filing to establish th file. All later filings for Social Security account numbers resulted in flexoline strips which needed to be individually separated from the rolls of continuous strips and manually inserted at its appropriate location in the file.

On the first run setting up the file you needed to plan how to minimize the future problems of expanding the file as its size grew. You obviously needed to spread that first batch so that there would be room on each panel of strips for later additions. Expanding a file like that, or an alphabetic file was the pits-- as we soon found out. So you always left enormous room for expansion, hoping to defer expansion for a long time. "Expansion" was the term used to describe the need when the stands on which the panels were hung and the panels themselves became full. But I am going back now, you asked me about the flexoline file.

And so that's what we had. The process, the initial process of handling the work was first to code the SS-5. Code them and then and after you coded them, review them to make sure that the coding was right. And also, as a part of that process, we had to indicate where the coding stopped, since we were only going to use on the punch card record so much of the first name and so much of the middle name--I think only one line. We had to indicate where the punch machine operators should stop. Make a mark with a crayon pencil, a red mark at the end. We would also, as I recollect, use that red pencil to show the date of birth, which was like 01-16-42, or something like that. All of this was to ease and speed the work of the card punch machine operator.

The coding was done before the block of work went to the punching room. It was reviewed in the review section. So we had the coders on one floor and the reviewers on another floor. And the review work was hard because it's hard to keep doing that all day long. Coding involved writing and that proved to be helpful in keeping one focused on the work.

Anyway, the block went through review and then it to punching to prepare the cards and machine verify them. Thus, every card was handled twice. A verifying machine does everything the punching machine does, but it doesn't punch holes. If the verifier did not press the same holes as the punch had, the machine would stop and the card would be set up on end in the tray and subsequently re-punched. With the punch card record we were then ready to set up the remaining machine-prepared "permanent" records.

I'm not going to try to give you the whole process, but the permanent records consisted of the account number application, the office record, the flexoline file, and the ledger sheet. The ledger sheet was a page of heavy paper stock and it had the name and account number of the individual. The idea was that one would post the wages, starting at the first line after the last previously posted line. Each newly posted ledger sheet would then be re-filed to be available subsequently when other reports of earnings were received later.

I only want to give you the general nature of what the central job was. The initial job was to set the permanent records, not to receive the wage reports.

We had a production control unit, I don't think that they were quite operative in the establishment process, we were establishing the records, they became operative when the earnings reports came in.

And if you think about setting up the initial records it was sort of an assembly-line process. The applications would come in and for a while only those at Step 1 had work to do. Obviously, in the early months as the initial load of 30 million records to be set up passed through steps A to Z staff had to be constantly reassigned. Then later on when the work went from one operation to the other; from locations on the left-hand side of the building to the right, the building swayed.

Setting up the initial records took quite a while and long before the initial major workload was completed the second major workload-earnings reports-began to be received. I'm talking about the earnings reports that were for the first six months of 1937, and I guess that they started coming through maybe around the Fall of 1937.

The process of maintaining a life-time record of earnings for each worker started with the preparation of a punched card for each item on the employer's regular report. The individual cards were verified as correct, the employer's group balanced and the individual card sorted by the area code of the SSN and the cards filed by area in the area holding file. The reason this accumulation was important was that you couldn't take the reports that you would get in block by block and post them, either by machine or by hand, because you would in effect have a density that was enormously to your disadvantage. If you had let's say a block of 1,000 wage cards, and a file of 30,000,000 master cards, you'd have to collate all your master cards just to process a block of 1,000 and then the next block the same way.

Joe Kreps had the job of heading the machine work of the record keeping--that vast, mass-production machine work, which DAO was pioneering. In the process, there were many shortcuts developed, improvements, that came as a result of the ingenuity of our own staff. IBM was working with us very closely, it was really learning the guts of how to make its machines useful to the business world. With our help, they developed the collating machine; with our help in processing all this work, we were inventing the systems that employers could use and they could sell. It was from that date, in 1936, that IBM's importance as a business systems operator, and being the largest one, took off.

We always had the highest priorities on new equipment, new ideas. They (IBM) were glad to have our laboratory to use, to test their ideas. Of course, many of the lead ideas that they had came from us, as to which direction to move to get improvements, where the improvements would be most helpful. So we had an extremely close connection with IBM operating officials. I didn't talk with the corporate heads, although I'm sure Joe Fay had quick access to them when we had an emergency. And I know, sort of secondhand, that Joe Fay had occasion to call upon the head of IBM or people prominently at the top level to recognize any special need that Social Security had. And IBM would divert equipment that was meant for somebody else to us, or they would speed up certain developments; or would undertake to solve SSA's problems.

Thus Social Security played a big part in developing one of the industries that became so important, and is so important today, in the United States. And I can't over-stress the fact that Social Security was, if not the only certainly the best, organization for IBM to work with to do this kind of innovative development.

We've been discussing the "machine operations" for which the Mechanical Section was responsible. Another major section of the DAO was the Wage Records Adjustments Section. Lou Baker handled that part, and I was his assistant. Lou dealt with the top staff of the Division and I handled all the other stuff, which were not the glamorous things: complaints, keeping people happy, etc. People wanted to know all kinds of things: what happened to their identification numbers about which they wrote to us five months ago, or they had gotten information from an office three months ago, or three weeks ago. We were only just beginning to have some files established. In many, many cases the folders were bare and we did not have anything to go on except our knowledge of the total operation. As quickly as we could we set up employee-employer correspondence files; employer number files, the flexoline files, and the employer earnings report files, and numerous internal work files, like suspensions, and reinstatements, and numerical registers, that we worked with.

We did some experiments with the operations early on. We would work ad hoc. Then procedures would come out telling how to go further, but also spelling out how to avoid some of the things that they didn't want to happen.

Q: Was there a procedures manual of any kind?


Sure. That's what I'm trying to say. They put out procedure manuals for operation 4 or operation 9 and so on-sometimes after operations had begun. They did the best they could, but like every one else they were confronted by the need to get things done "yesterday." Obviously, they tried to produce uniform and efficient ways of doing the work.

But I'd have to say that for the most part, they were not yet particularly skilled at that kind of thing. In fact, I don't know that any of them came with academic backgrounds or major work experience that particularly qualified them to optimize a particular procedures or systems.

Innovations Introduced in DAO


I would like to give you a feel for the ingenuity exercised by the employees and the way they adapted.

Consider the simple job of alphabetizing the office records. The records came in blocks of one thousand. The objective was to end up with an alphabetical file of 30 million.

How do you alphabetize? Take a bunch of cards and spread them out and try to get A's together and B's together, etc. And we soon saw that this is not going to work-- 26 letters, 26 tables, that's not going to work. You have to put them together. You had to put them into 5 or 6 groups, A , B, C, as a group, etc. But after a while you see that isn't going to work either. Why isn't it going to work? Some letters, like maybe "S," had a large volume, and others a small volume. To be more efficient we tried to figure out what groups to set up so that they were more even. And it worked out after a little while. And they did all kinds of innovations.

I guess what we did first was to break them down into groups and then eventually into A's and B's and C's and so forth and then give somebody A's. After awhile we had a board, this was self-invention, a piece of either cardboard or wood, and we would take the punch cards, take a lot of punch cards, and staple a punch card on this board. An "A" in front, than a little further we put the "B" and then "C", "D", "E". And then we'd hold that over our stomach and use it in sorting the cards. And people were free to make up whatever they thought was effective.

Another problem was the ledger sheets stripping. The ledger sheets had pin holes, they were pin-hole fed. And before they were put into files the pin holes had to be stripped off. This was a big, big job. And how would you do it? And trying to figure out what's the most manageable number that you could do at one time. And then realize "I'm getting all cut up." And you begin to wonder, if you had a pair of pliers, would it help? And so on, down the line. Everybody tried to come up with better ways to do things.

At one point we had a regional setup, that I'll explain to you--remind me to talk more about it, because that is important. Let me just mention it now. I was in Region 2. It was around 1939, that it was set up. The whole idea was we were to try out a regional organizational arrangement for 12 regions, in the Candler Building, and then we would try to disseminate it in various parts of the country and bring the operations out to those various places. And there were a lot of people in those days trying to get into a region they wanted to get into.

I was in the first region which was set up--the New York region. I was the Assistant Chief. Actually, I pretty much did all of the work. Leo Snyder was the Chief. He was an old timer who had been a football player, and he was loved by everybody, particularly me. But Leo I think had taken a few too many hits on the helmet, or whatever. Nice guy, but he was just in another world. But they were carrying him. I was assigned as his Assistant to set-up the first Region. So I carried the ball. And Jeff Chesler, a guy who had a similar position in Region 3, that was Philadelphia, invented this one machine that I want to describe, that was used in the stripping procedure.

Jeff was no intellectual genius, but he must have been mechanically inclined, because he eventually designed this machine, which had a very simple principle. If you have sheets together, and if you could put pressure on them as you pull them along, the pressure will tend to burst the sheets. So he thought of a bursting machine. Before this we tried everything: tearing them with fingers, with pliers, with special cutting machines, with slicing machines, we did everything. So he put these sheets on rollers, and he set the rollers at a certain place, at a slight angle which exerted just the right pressure and the pressure would burst the connection between each sheet. And as they went through the machine there were a couple of knifes out on the end that would strip away the pin-hole feeds. That machine was patented by the U.S. Government, and I don't think that he ever got anything out of it.

One more thing that occurred to me. In the operation in those days we didn't have a computer, we didn't even have a developed punch card system. The existing punch card systems couldn't do the job that we needed to have done.

Contemplate this. We had 30 million sheets like this in the file, and the sheet has the account number, and now we get earnings reports six months later. And then we had to do all of the work necessary to bring it up to a point of posting it to the ledger sheet. The operator had this big piece of tabulating machinery, and he has to take the sheet out of the file and put it in a roller, like putting it in a typewriter, position it manually to the place where he wanted it posted. In order to be able to do that he first had to have a listing of these cards. He had to have a listing of the cards to be posted. So that was a posting list. That meant that somebody had to take that posting list and pull the ledger sheet. Because you were not posting them in order, you had to pull out the ones that you were getting ready to post, maybe one per drawer or something like that. They had a drawer with the ledger sheets, either pulled out or offset in some way. It had to be a one-to-one correspondence between the next card that is coming out and the one on the sheet, and you had to take this and put it in there, wind up the works, push a button, and only one card would go through, sometimes two, because they were employed by two employers. It was controlled by account number obviously, and that was a very tedious thing. Now visualize the next six months, or calendar quarter, you had to do the same all over again. You could have done it three times by hand in the time it took to do it with this process! So we thought, why not take it right to the ledger sheets and eliminate all these cumbersome operations.

So there were various ways in which the posting operations were speeded up. One thing we did was use a light beam, and if you break the beam you know that something is wrong. On the ledger sheets we punched new holes so that we could use a light beam to position the sheet so that you didn't have to manually adjust it on the roller. As long as there was a hole for the light to could go through the sheet would move, and it would stop at the next line. So we developed various automated machines, and then that went on and on and eventually the process became highly improved--although it remained essentially a manual process. These were some of the things that enabled Social Security to overcome the inherent problems of a job that had never been done before.

I touched on this just a little while ago. Another problem with the system we had was we didn't post directly to the ledger sheets because the record that was "machinable," that the mechanical computers could read, was on punchcards, not on the ledger sheets. This was before scanning, so the mechanical computers could not read ledger sheets. And there was a master card of earnings statements. You were gambling purely on that card, a lifetime history of earnings would be posted to that card.

In those years, the first couple of years, I think, they were only accumulating annually until quarters came to be used as the reporting period, which happened later, maybe in 1950. But the base record couldn't be used to go all the way back, to give quarters of coverage etc. After the law changed in 1950 you had a choice: either you could go back to the 1950 record or back to the 1937 record. And so they had to adapt the master record so it would carry a lifetime history including all of the important things like your number of quarters, etc. So this record was labor-intensive. First, it was very difficult to maintain and it was not a very handy kind of record to base your operation on. It become important that we devised a master card to be able to be the vessel containing the benefit information that would be needed later on. And so there were a whole lot of things as far as Social Security was concerned that we had to do to handle these workloads of millions of items.

I was talking about machinery, the effect of machinery. One the problems we had was the need to match names with numbers. Say there was an employee named Lawrence J. DeBach, and it was spelled differently in different earnings reports, so that it didn't match exactly. We needed to know which one matched exactly. The final step in that operation was to match the account number and "X" letters of the first name, and "X" letters of the last name, they had to match what was in the master record. You needed a machine after you punched up the card to match them against the master card and the machine would either inter-file it or reject it, called "rejects."

Q: And that became the Suspense File eventually?


Yes. You couldn't do it then, reinstatement of earnings that failed to post. Let me give you the process and then you can understand. The process was to compare it with the file. Now we had another file which was the account number register. Do you have books in the Archives of a hundred on a page, of a register?

Q: I think I have that.


That was one of the things that we did. Once we produced the account we made the flexoline strips. We also printed a page on a standard form in big books. And that was the fastest way to lookup who had an account number, rather than go to the file or anything like that. And that was a very convenient thing. It gave you the name and gave you all of the information, full name. So we didn't have a machine that would do that. IBM didn't make those machines. They had to invent it for us, and this was the machine that really put them into the business of business machines. Up to that point they were just filers and things like that but they were not pursuing any other worlds.

Q: Were they computers?


No they weren't. We worked with IBM to come up with such a machine, and it was called a Collator. And that machine had a board in the back, an electronic board, with wires you would plug in, and it did certain tasks. It was what was known as the Social Security Wiring Device, something like that. But it was prewired for Social Security. For all other customers it wasn't. It worked out that they would have to wire it for their particular needs. And what we wanted was to have it done in a certain way. I don't know whether you are familiar with those kinds of machines, but they had two stacks of cards. There was one stack here with the master file, and another stack with new data. The master file card was fed then there would be a read on the input file. It there was a match on the account number then this would be further read to see if the name matches. If the name did match, the master file would precede into a slot and the other one which matched it would be filed. If the master file and the input card did not match then the card would be rejected.

Then the manual process was to take the suspended items or rejects, and they were arranged in account number order--that was the sequence in which they were in, give them to a clerk, who would take it to the numerical register, and look at the name as in the register, and if the name was very similar, and there were specific rules, they were obviously the same person. For example, it might be both a maiden name and a married name and in the register you could see both the maiden name and the married name and so you would know it was the same person. And we would make changes on that register so that when a woman got married, we would hand-write the change into that register. So there were some reinstatements from the rejects, and then they would at that point be in suspense if we could not reinstate them. The rejects then became suspense files and what they were subjected to was to go back to the employer report to see if there was any additional information on the report that would enable you to reinstate it. After awhile you developed other ingenious checks. For instance, you could look to see if this is a person that has been working for this employer for some time. And you would be able to deduce that this was the right person.

And then we had what we called the "suspenses." This would be a job that we took a lot of time to do. We tried to find the right number through every source that we could logically use and then we made a list on the IBM machine, of the suspended items, and we would go back to the employers and say, "this didn't match up, please give us the correct number, you made a mistake, or whatever." We used to reinstate a large proportion of the numbers without going the employer. And then we would get them back and we would make the corrections. So IBM did the job for us of developing the machine that enabled us to do that. Incidently, they then took off and we became for many years their test spot, and we had visitors, IBM visitors, from business schools for years and years in Social Security where we were their model site.

Q: We were the state-of-the-art.


Well we were the eye-opener. And that was well prior to the electronic computer.

Regionalization of Accounting Operations

I probably ought to digress and make sure that you understand about regions. We prepared to receive the first reports from employers of employment, presuming the records were set up. We were already getting lots of correspondence before the records were even established. People wanted to know what happened to their I.D. number and lots of inquiries into what their earnings were and we had no records established. We had to answer these people. And this volume of maintenance work was already on hand and we were still trying to establish records.

The thought of those who were planning out the scheme of record maintenance, was that we would try to decentralize that whole operation. Because, keep in mind that Harry Haupf, a consultant, who was hired in the early days had said that the system would fall flat on its face. Or if he didn't, five others did. They said that it was just an impossible task.

Now where was I?

Q: You were talking about regions. Why you set up Regions.


Oh, in the Division of Accounting Operations. The thought then was that this is so vast an undertaking that we should try to break up our operations in the Candler Building into 12 regions. The first one would be the New England area, the 2nd one would be New York, Philadelphia would be number 3, and Pennsylvania, Delaware, and Maryland would be 4, and so on around the country until California was 12. And the first one to be set up was the biggest, New York.

As I indicated to you the other day, Leo Snyder, was one of the staff people around Joe Fay, who came over from the three A's. He was an administrator from the three A's, the Agricultural Adjustment Administration. In DAO many of the original top staff came from the AAA. McDonald came from there. Lew Shield came from there and Glen Duey and Charlie Taylor. And I think that O'Beirne might of come from there and a lot of others. Joe Kreps came from there and John Grogan came from there. And so they had this thing, they were machine people, machine-oriented people.

And so they set up this first region. Leo was designated Chief of Region 2 and I was his assistant. I was still a Grade 2 and the reason for that was that in 1938 or 1939 there was a freeze. Very terrible times as you know from history, and nobody got any promotions. And I don't think that anybody got a pay raise from 1939 or 1940 or something. So there was a freeze. And besides the plan of organization had never really been approved, because the organization itself was changing all of the time to take on these new functions. New, never before performed functions.

Region 2 was set-up and had a Files Unit. It had some kind of an Adjustment Unit in which correspondence was processed. And it had a Machine Unit, with punched cards and it had sorting machines and tabulating equipment that did the New York work that came in. We handled the New York correspondence which was tremendous. We had the largest portion of all the work that was done. Denver was a handful really, Denver, was Region 11 or 9, I forget. And Minnesota was 12, later reduced to 9 or 8 over the years. So that was the reason for setting up those regions.

And I became Assistant Chief under Leo Synder who relied on me to do most of the work. He handled the front office. I eventually had three shifts. I first established the day shift. We had an afternoon shift, which worked until midnight. And we had the midnight shift, because we ran three shifts around the clock. I spanned those shifts. I stayed in that position until 1938, or thereabouts, I think it was 1938, I was moved out of that position. Well I was still a Grade 2.


Abandoning Regionalization

Those regions, you can see how they would not work. They were assigned on the basis of original point of issue. Like new England, if I recollect off hand right now, was then the area where account numbers starting with 001 to 050; New York was area numbers 051 to 120; Philadelphia took off from that. And these were preprinted cards. Now when we set up the regions, employers did not send in their reports by area number, and only send the appropriate areas to us. That would have put a burden on the employer which would have led to a revolution before we even got started. But, conversely, this process of regionalizing into 12 regions would produce a situation like this: New York would get whatever clients existed in New York or whatever location, file the report, and they might file reports only for a local client, or for the entire company like General Motors or Ford. Wherever they were located or however they had worked out their reporting function, rather than for the account numbers, the area codes, assigned to a particular region. So what you now had introduced after these reports had been processed and all the incoming earnings data had been balanced so to make sure you hadn't made any errors, then you accumulate a lot of punchcards for other areas which you would then have to sort and send to another region so they could consolidate it. Because if you were breaking down the ledger sheets, you'd only have the ledger sheets, and so there would be a hell of a lot of shifting all around.

If I had been a member of that staff, which I was not, and this was a long time before I had any kind of influence, I would never have tried it. It was obviously flawed from the beginning. I've never uttered that judgment, that adverse comment, until now. But there was so much, everything was new. Obviously, a lot of it was ad hoc, and some of the things that should not have been ad hoc, unfortunately were. Maybe they had too much pressure upon them to produce a decentralized organization; you know, things like that get fashionable in certain times. And maybe even against their judgment, they gave it a try. But obviously, that could not last; it would not be efficient; it would have introduced all kinds of problems in the process. Again, I'm just underscoring what happens after that is that computers doubly emphasized that there was a need for maintaining records by region. It was then not at all a feasible operation, it was just waste, duplication of effort. As computers took over more and more, geographic location became less and less important in certain functions.

That did not mean that decentralization with respect to other functions, like in the district offices, needed to go the way of record keeping. Obviously to me, record keeping was, by nature, best done even in the early days by a centralized organization, that was the most efficient operation.

But before we abandoned regionalization in 1939, there were a number of consolidations. The problem was that the workloads were not evenly distributed in the regions. New York was certainly not 8 percent of the total.

Q: Much more?


It was more like 40 percent. And Chicago was a large quantity, but not that much. And Philadelphia, which when I say Philadelphia, it wasn't just the city; in some cases, it was a block of States, like Boston was all of New England; New York was only New York; Philadelphia was Pennsylvania and New Jersey, and so on--there were 12. There were consolidations. I remember that Region IV which included D.C. and Region X, which was Texas, were combined under George Moriarity, IV and X. Just on the basis of convenient sizes. I wasn't privy to any of that, I'm just reporting what my recollection is. And also, Region I, which was Boston, was combined first with Minneapolis, which was the whole State not just the city and it was 1, 9, and 11. Denver was the smallest, the smallest population. So there was, prior to this movement in 1939, a functional organization, which brought these 12 individual regions which were broken down, first by region, and then by function--they were brought together when they were consolidated. In effect, we ended up due to consolidations with exactly what we had in Region I, and the other Regions, which was a files organization, which was the lowest level to maintain the files, and that was under Review and Adjustment. And Lou Baker became chief.

Q: So this was a files organization that went across all the regions?


That's right; all of them, they were consolidated. And what you had was a machine section that became Joe Kreps' responsibility. And the wage and adjustment the correspondence and the files, all went under Lou Baker. And adjustments and general files, which is similar to the function I had in Region II. And Lou Baker was the regional chief of Chicago, Region VI, he took over that responsibility.

Leo Synder went upstairs and he got kind of swallowed up. I think at that time, he was beginning to suffer some marked loss of ability, physiological conditions. And that was in 1939.

And again, when you look at the entirety of the split, Joe Kreps had all the machine functions. They were tangible, he could deal with them and he knew just what he was supposed to do. It was a mass operation, not very complicated in its overall function.

Personnel in the Candler Building

When I reported on November 16th, 1936 there were only a relative handful of people in the building. There were some people, like J.J. O'Beirne, a long time DAO employee, who eventually succeeded me in the Control Branch. He was also the Assistant Chief in some interim period. J.J. O'Beirne had been hired even earlier in 1936, close to the time when they were operating without any money--Altmeyer used his funds from the Labor Department. Esther Scholl was there. Jim Murray was a messenger in Washington when he came over, I think that he came over, maybe he never got to Baltimore. He was a messenger in Washington. Jerry Tall was a native of Baltimore and they were hired because they needed people to get started and these were people who did not have status careers, they were temporary workers.

J.J. O'Beirne

Esther Scholl

Jim Murray
Jim Murray

And then they started coming in big waves. People from various parts of the country. It was truly amazing to see, I don't know, it was interesting, if not spectacular, to see these people coming in and come together as a group that sort of pulled together, all anxious, all needing work, all glad to have a job. Qualifications varied a lot. High school graduates to PHD's, and people from the East, particularly New York, many I guess reflecting the economic conditions. They were unemployed, highly educated, and I can even today name quite a lot of people who came in, in those days, who became the "framework" of operations in Baltimore and later went over to Program Analysis and other places.

Something about people, they didn't have any money and they didn't want the expenses. We couldn't have. One helped the other. They were strangers and they met in the employment office of Social Security where they reported for work. And then, by necessity, they needed to find a place and someone already had a place, and said, "You can move in with me." And four or five would find a place and share it, and one or two would run out of money by that time and they'd share. And they were having fun too after a while. They would get together socially. I remember attending dances at the Knights of Columbus and having other social affairs. There was a solidarity of people and they intermixed. My future wife was there from Washington--she was born and raised there. One of the few in those days, most people who came to Washington, were not from Washington. But she was one of the few in those days and she worked for me, she started a week later.

So people from all over the country mixed with different backgrounds and they got along very well together. And they helped each other. In my own case, I came in and they directed me up to a floor. Somebody would sit down and count off six people as one group and say "That's it" and then they would start another group. And then we didn't have any more people.

Then somebody came around with these cards and somebody gave out telephone books so that we could get practice with the coding. Some of the names weren't exactly good for that purpose because these telephone books were in alphabetical order so you needed to mix them up. Eventually somebody made up a collection of random names that we could use.

In my group, on like the third day I was there, I encountered a new character from personnel who had already worked in the Government and had come in as a Supervisor, as a Grade 5. I was a Grade 2. He walked up and down the aisles obviously looking at me and three or four other people. And then he came over to me and said "you're a Group Leader", and that was the third day, I think the 3rd day,

Q: Just like that?


All I can tell you is that the appearance of people differed greatly, some more obviously were ragged or whatever. I reported in a blue serge suit. I am only guessing, I don't think that was the criterion alone, but it could have been one that called some attention. Possibly they might have looked at me and said, "That's a smart guy." I don't know. I don't know. Because there were about 125 to 150 people in that first group and some became Group Leaders. One was W. B. Ruth who was from Baltimore. What a guy! He and I became fast friends. And there was a lady who later worked for me and I can't recall the lady's name right now. Ruth quit Social Security eventually. Anyway, we became Group Leaders the next week--five of us. We established five groups, and we were coding groups.

Q: How big were these groups?


Groups of 20. And it was really strictly production. We were people that were somewhat young. They were college age and were not people who had been bounced around. This was their real first job for the most part--but not exclusively. The thing that happened was it turned out to be a production race.

Q: You mean that there was a race between the groups to speed up work production?


Sort of. There were certain groups that were naturally faster. And I am not sexist, but in those days they were predominantly men who tended to try to show their macho-ism and work faster and all of that. So one who had a group of mostly men, tried to maintain their lead. Production went up very rapidly and it became very discouraging for the others. They couldn't say anything in those days or you wouldn't have a job.

My particular group, were not speedsters. Let's face it, my wife was certainly not a speedster. And I had a number of women who would do the level of work which you should reasonably expect. And then we had a couple of people who were more knowledgeable than most and I have in mind Robert Mayne, you may not know the name, but Bob Mayne eventually became an Area Chief. And Joe Carmody who became a District Office Operations Manager. They both were in my group in the early days. Joe Carmody and Bob Mayne worked as a team and their objective was to be good average coders. Now as far as I was concerned I wanted to be sure that my people would not be so drawn into this that they would refrain from going to the john when they needed to. And so when they left I'd sit down and do the coding for them.

Q: You mean when they took a break?


Yes. They took a break frequently! And I'd sit down, I was a pretty fast coder. So they did better when they went to the john! But my group managed to stay, not on the top, but comfortably in the middle, exemplified by Joe Carmody and Bob Mayne.

Q: Let me ask you about the percentage of employees who were women at that time. Today if you look at SSA 72 percent of the agency are women. I always wondered how early did that start? Was there a high percentage of women back in DAO in those early days? Did that happen when the War came along? What do you remember about that?


I would say that it was roughly even, men and women. Maybe a few more women than men. And as to what happened, I think what happened probably, was given a boost in the DAO by the War. I went into the service, I was gone four years. A lot of people went into the service, so they tried to get people from where they could in those days. Women were not employed to the degree they are today. I think they were natural sources of supply. But I was gone when this took place.

Other Key Personnel

Some names that come to me in that process are John Harrison, who was the first chief of Region I. John Harrison was a young English fellow, with glasses, very studious, serious, well-respected amongst the people generally, well-respected as somebody knowledgeable who would give you easy, off-the-cuff answers, that could stand up. You could rely on what John had to say. And he was, I would say, able to produce with the idea of focusing, to some extent, on the most efficient method.

Conversely, and it is conversely, Jimmy Gray was a guy who had a detailed knowledge of tabulating equipment, especially wiring that magic board for the collators and dealing with certain problems, machine problems. He really wasn't given any task except go see Jimmy if you had a problem, a machine problem. And I think he performed a very, very useful function. But it was a function that increasingly over the years, became less and less important. It was important because, we knew so little about tabulating equipment. We had very few people, some who could run it, there was not much to run a sorter, just push a button and watch the cards go whizzing by. The same way with tabulators and punching and verifying; there was nothing very abstruse to understand about machine processes. But how to manipulate these things, and how to wire electronic boards, electric boards, with circuits that were in the part of the collator to achieve certain results, was Jimmy Gray's special function, he did it all. And he would do it and commune with himself, he didn't talk very much and he was a very introverted person. I think people liked him in general. But you didn't get to know him very much, because he was just by himself, so much focused on that.

And a third person, and these don't comprise the totality of the Methods Branch, was Jimmy Haggerty. Jimmy Haggerty was sort of a wild man. He knew a lot about machines. But he was excitable. He was prepared to take risks, and sometimes he had insights, and sometimes he had blowups. It never was a blow for him; he'd come back punching the next day or whatever. Jimmy Haggerty was useful, very useful on particular occasions, but not very often. He would explore and take the time with the staff to pursue or consider some idea that turned out not to be very practical, but he would pursue it with vigor. So that gives you some picture.

Warren Irons was at the top of that. Warren Irons, as I indicated to you, he was a smart guy. He had some connection with the business world when he came in. He was hired as an expert on the Soundex system. At least that was common gossip when he came in. He didn't even know what the Soundex system was. I gave him a card and we never talked about it. But I'm not being critical about that because knowing the Soundex System was not important, anybody could have learned it in ten minutes. So as a pre-requirement to achieve in Methods, it was really to me, nothing, a nothing factor. Although I was not a direct observer, I was a user. I realized that they did some work that was good in the early days, it was helpful in the early days.

We operated in large part like Lewis and Clark-exploring as we went along. And Warren Irons and his henchmen came along as soon as they could. It was a very messy sort of thing.

Well, I could spend more time on the inventions, and the creative things that took place with Lou Baker and myself in our little place, but I'll pass on.

Discrimination in the Candler Building

We at Social Security were not entirely free of discrimination. Because in the Candler Building, in the Wage Records Section, and I'm being as factual as I can without making judgments, we did discriminate. The only blacks we had were two groups. One, unlettered, uneducated black people, some of whom went to char-force, which meant cleaning the toilets. And they also had a CU-1 grade, custodial and labor; the men were laborers. We had these big file cabinets which we were moving from one location to another all the time, and the laborers did this work.

The other group of blacks, were in the Flexoline Files. They were under a gentleman named Bill McClure; he was an older gentleman, bald-headed with glasses, very serious, he had an authoritative presence. And he had only black people, and they were attending the Flexoline Files, at night, mostly. If you stayed behind at night, on that particular floor where the Flexoline Files were, you'd see these lights coming down from the ceiling, and a black person working on them with a lot of these strips, interfiling, it was very hard to explain to you how that had to be done, it was a very tedious job. And they had the supervisor, and some of these ladies that were doing that work, and men, some of whom were graduates of black colleges. And some of the supervisors were people of that kind of education, well-qualified to do the same kind of work that we did.

Furthermore, and I'll just touch on this, we followed so-called "customs." Joe Fay was particularly concerned that we shouldn't get ahead of conditions in Baltimore. That would cause trouble in the community, etc. And I'm talking about use of the toilets. And this was illustrative of the fact that we were not free of discriminating. In the Candler Building, we did not have enough toilets to meet State standards for the number of toilets per work space. And these toilets were already established in these floors when they were built, which were intended for a much less dense population than we had. So we had far, far fewer toilet facilities than was ever contemplated to service the kind of dense population that we had; we had a couple of thousand people. At most, I think the building was designed for 300 or 400, at most, on all the 9 floors. So when these work operations moved around, we had to have toilets for whites and blacks. We didn't have unisex and we didn't have unicolor.

Q: So you had colored toilets and white toilets?


Four toilets, we had.

Q: You had four toilets?


Yes. Black Men, Black Women, White Men, and White Women. And imagine trying to get four toilets on a floor that maybe had only two or three. I don't mean individual toilets; maybe they had seven urinals in one, but only two or three restrooms. And it was not infrequent when you went into the toilet to encounter that the toilet assignments had changed, and the men might end up with a restroom without urinals. And on occasion, I had wandered absentmindedly, because they had made a change and I wasn't aware of it, went to the wrong designated toilet. There were four possibilities and I think I may have at least enjoyed two! That's just a little sidelight.

Q: While we are on this subject, did you see evidence of anti-Semitic attitudes in SSA in those early days?


I would not go that far. I will objectively say this. At the time, I did not accuse, in my own mind, anybody of anti-Semitism, never. I always felt it was more easily explained by people in such positions starting out with a difficult task. So you relied on people who you knew or had a feeling for. And so, the Irish, in the early days in the Candler Building, had a natural advantage; they were in a group associated with Tom McDonald and Joe Fay. In other words, something to do with bringing together initial cadres. I would have to say that in 1938, Joe Kreps, who was Jewish, had been a member of this group; he had been sort of taken in. So they were not actively, in my judgment, anti-Semitic. I can't say absolutely that it didn't enter in. I think, I'd rather put it the other way, when they had somebody they knew or were familiar with or felt at ease with, and were not uncertain about, who had approximately the qualifications for a job, the other fellow who was competing who was not a part of it, had to be a very clearly better, or very clearly the person who they would have benefit from having someone like that. And that's about as honest an answer as I can come up with.

I'm not reluctant to bare my soul on this. But I personally have been irritated by people who too easily jump to that kind of conclusion. And I am today when the same kind of thing is done by blacks about racism. I'm not denying that racism exists still today, but I know it's a very complicated matter. And charging a group or an individual with racism doesn't help, it hinders progress. And one ought to be pretty damn sure that is an accurate assessment before one adopts it as a point of view. It interferes with progress. And it makes it difficult; it just antagonizes people that may learn to overcome what inherited prejudices they have. So I hope I've answered that.


Relationships In DAO

We're describing the functional organization in 1939 or thereabouts. I touched on Lou Baker being the head of Review and Adjustment and General Files. That was the largest, least mechanical and highest-graded outfit. The low-graded organization was the Division of Accounting Operations. We had unseemly low grades at that level. It took a long time to make some corrections. And then there was the Registration Branch. Now these were called branches, Review Adjustment Branch and the Files Branch, and the rest of them were called branches. Review and Registration Branch. Glen Duey was in charge of that in those early stages. Then we had a Message Branch. Warren Irons was the chief of that branch. The last, there was a minor branch called Production and Control. Then later when I came back from the Service in 1945, I assumed the head and chief of the Production and Control Branch.

It was interesting and I have some interesting things about that. A really funny story, too. In the Office of the Director, there were just a few secretaries. Tom McDonald, let me say now, was not an intellectual genius, and he never pretended to be. If there was one man that was the heart and soul of the Division of Accounting Operations, it was Tom McDonald. As I say, he was not an intellectual genius, and he never pretended to be. I'm not preoccupied with things Irish, I married a half-Irish girl, but Tom McDonald is the essence of the happy Irish. He had the Irish guile with a song in his heart. He was a smoocher of the first order. He'd put his arm around you and you were three steps down the ladder or something, and he'd say: "Jack, old boy, you're just the man for this job. I don't know anybody who could do it as well as you." And then he'd let you do it. He wouldn't give you instructions. One of the reasons why he didn't give you instructions was because I don't really think he knew how to do it. He knew that he didn't know how to do it. I'm not trying to make him incompetent. I think that what he gave was a sense of importance to almost everybody he came in contact with. Against him Joe Fay was a frozen-faced Lace-Curtain Irish, it was the term they used then--not outgoing. He had what I think was then the current demeanor for an administrator, a walk above the clouds. When Tom McDonald and Joe Fay would walk down the aisles, Tom McDonald would be bowing and smiling and saying "Hello, Hello" and Joe would walk straight down. Eventually, Joe learned to do this a little, but he could never do that much.

Joe Fay
Joe Fay

Tom McDonald

I learned a lot from Tom, and I loved him, as a lot of other people did. People on the staff used to make excuses to go up to the 9th floor, and shmooze a little with Tom McDonald. That was the kind of guy he was. He solved many a problem before it came to Joe Fay's desk. It was always not on the basis of substance, this expressed confidence in people and real liking. He couldn't have simulated all this. I spent many a happy Sunday enjoying his family, just visiting. I had nothing, unmarried and I was living in Baltimore. He had a million brothers, at least eight or nine, and six or seven children. They were all around on Sunday. Some of them were getting into trouble, and in those days it wasn't much. One brother worked for IBM. Tom was devoted to Joe Fay. He was as loyal as anybody could be. On the other hand, he was the one who handled all the problems, staffing, etc., and then carried out the dirty work. This is needed in every organization, to separate the tasks that your subordinates do, what they do best and what they can contribute the best. So there was Joe and Tom McDonald.

Did I say I was going to give you some idea about some personalities?

Warren Irons was hired as an expert, my recollection, in 1936. As he came in, he was hired as a Soundex expert. He was a redheaded, aggressive type. Aggressive compared to the laid-back McDonald. I don't want to over stress the Irishness, but this was an Irish staff. You had Barney Nolan, and Glen Duey was not Irish, nor was Lew Shield, but there were a lot. There was a lot of Irish in this. There was Joe Kreps whose real name was Sol Kreps. Everybody called him Joe, and he was Jewish, but not very. J. J. O'Beirne was very close; there was an affinity there. Warren Irons was not really one of them. He had a certain starch about him that didn't fit in all that easily with the easygoing nature and fun-loving character of all of the Irishmen.


Warren Irons

Warren Irons was hired, as I understood it then, as a soundex expert. We were undertaking to do soundex, so-called flexoline files which I described to you. And I remember him coming to me and he said: "You got one of those Soundex cards?" The Soundex card was coded. They didn't have desks. They all shared tables. That's how we worked in those days. In this early fluid time, he met the test they had on the 9th floor, he was one of that group. From what I could tell, he was pretty sharp. He tried to stay ahead of the problems of how to process these things.

We were really given the tasks, but how to do it and specifics and all that were not written down. So we were really like Lewis and Clark, seeing if the world dropped off to nowhere, and if it didn't we just charted it away up there and then we looked over the hills to see where it was. And it was unchartered territory. And Warren was the head of the right group. Leo Synder went back to Warren Irons's branch. He later wound up as the night supervisor. John Aimison who used to, in effect, take Joe Fay's place after Joe went home, which was that night shift. It was dark, the lights were all out. He and his secretary had no lights. Nobody would ever hear from him except if there was some emergency and he needed someone to make a decision. John Aimison ostensibly was the guy, I don't know what kind of decisions he would make. He was an old experienced person that they had come to know sometime in their past they had been together, he was inoffensive. Betty Quinn worked for him as a secretary.

Q: I know Betty, and I've been thinking about interviewing her.


She worked with him.

So that was Warren Irons. He ran this little shop.

I think that Lou Baker had been like Steiner was, the Chief of Region II. Lou Baker was the chief of Region VI which was Chicago; I think the second biggest. And Jerry Tall was my alter ego in Region VI. Then after Joe Kreps came and I was reduced--reduced on the paper, I had never been promoted on paper--no loss. I'll tell you a little more about Baker. There was Warren Irons. Glen Duey was sort of an easy guy. He was a relaxed kind of fellow. He was no brilliant star, but steady.

Some of the employees got in a little trouble in those early days, I'm not pointing at anybody, but that was not uncommon for part-time employees. A lot of people were away from home and away from their families, if they had any. There was some trouble with some of them, but no more than one would expect in those circumstances.

Glen Duey was Registration and Joe Kreps, stable guy, old-fashioned machine man who had a lot of experience with machines, he ran the Machine Section. He just died recently. He was the other man with practical solutions and just an easy going guy. I know he was divorced. I would answer to him: "If you don't agree with what I'm doing, then tell me and we'll do something else. Take responsibility for your own judgment. Don't try to convince me when I'm well past the point that I'm not going to be convinced by what you say. I'm not going to change it unless you direct me to." I explained that he didn't need to go through all this stuff to try to convince me to do what I thought was right. We got along, especially in later years, and we had a pretty good relationship. That was the machine section with Joe Kreps.


Joe (Sol) Kreps

Lou Baker was the opposite of a happy-go-lucky Irish type. He was a super-serious guy who found it very difficult to laugh. He had a very deep voice, and he was always smoking on a ten-cent cigar, or a nickel cigar I guess in those days, and he was inseparable from them. As I say, he was very methodical. He was sober, super-serious. You wondered sometimes when you were saying a joke whether he understood. He was no dummy. He was an able man. But his outlook on life was different. I became his deputy, his assistant, full assistant, in the Wage Records Branch in 1939. I liked working with him.

photo of Baker looking down

Lou Baker

The same thing happened with Joe Kreps. I funneled all my suggestions and ideas through them and they would go up to the front office and eventually to Dick Branham. They could have been more generous in having me recognized. But neither Joe nor Lou I think were overly generous in that, judging from the results. I was a little bit of the outsider then, as I think about that, Lou Baker I'm sure was too because he wasn't in that original group. Not that they were obnoxious, but they just naturally leaned on one another. They knew each other and trusted one another and were suspicious of a character from New York, you know, who was a little bit different. I don't fault them for that.

Lou Baker and I were both outsiders. He didn't suffer all the disabilities that I had with respect to being an outsider. Now Lou Baker and I each had two assistants. This is a peculiar arrangement with Lou Baker and Jack Futterman as assistant of the Review and Adjustment Section. I worked basically on the night shift, and he handled the day shift. I did all of the memos and all that stuff for the front office which he took up and then had signed off, and so forth. He had as his assistant--but they were only shift assistants--George Leibowitz, who had been my assistant when I was Chief in Region II. As a matter of fact, I bumped him and he went to the night shift and that characterized our relationship for quite a long time.

George was sort of in my wake. During the War he took over my position in Readjustment and General Files, he took it over. And then when Baker moved up to replace Warren Irons, Leibowitz became, on paper, my boss while I was in the service. We had this policy of promotion in absentia which was obviously not working that well. I was actually not promoted. When I got back, I got this job that they brushed off on me. As I said, we had two assistants. Lou Baker on the day shift had George Leibowitz, who handled the Review and Adjustment, and Jerry Tall would handle the files. George Leibowitz later worked in Program Analysis and then he went to ORS where he had contact with Herb Fischer.

When I came back from the service 1945, I was given the job of chief of the Production Control Unit, that small unit. The job had been vacant for about three years. What they had was a subordinate unit. The guy that last had that job was a fellow by the name of Ed Whitney. I think he left in about 1938. And under him was a fellow by the name of Roy something, and he had just a very strong unit. His function was to establish accounting control of the work. In broad detail, these few people were responsible for taking employer reports and bunching them so that they made up a block of work. I forget now, whether a block of work was roughly a thousand cards. The idea was that as the work processed, you would bring up the accounting control of how much earnings had been received and posted, and how much went into suspense and how it would be reinstated and so forth.

Part III- Wartime Experiences