MODERATOR: Good morning. This is the technical briefing for the corrected version of the Crop Production Report. I am USDA Press Secretary Keith Williams.
On the call for USDA this morning is:
Carol House, of the National Agricultural Statistics Service known as NASS. She is deputy administrator for Programs and Products.
Dr. Gerald Bange. He is the chairperson of the World Agricultural Outlook Board.
And Dr. Thomas Hofeller, Farm Service Agency, associate administrator for Operations and Management.
We'll have statements and then go to your questions. A USDA news release is online at USDA.gov
Two terms you'll be hearing this morning:
One term is "data mart" - an internal data system we'll explain to you. Carol House will speak to you about the Crop Production Report.
Also, Dr. Gerald Bange will be mentioning another report which is known as "WAS-DEE" for its acronym W-A-S-D-E, the World Agricultural Supply and Demand Estimates report.
And now, Carol House, the NASS deputy administrator for Programs and Products.
DEPUTY ADMINISTRATOR CAROL HOUSE: Thank you very much. NASS annually estimates acreage, both planted acres, harvested acres, and production for a major crop. We have a Crop Production Report that goes out monthly. Our October Crop Production Report that we issued on October 10, we updated the planted acres for several field crops. After that report was released, we found out from our colleagues at the Farm Service Agency that a major database, called the "data mart," or that we refer to as the data mart, that we use to go in and actually get information, had some data discrepancies in it.
What NASS uses to determine what acres have been planted: we use farmer surveys, satellite imagery, but also this time of year it's very important for us to have information from the Farm Service Agency about how many acres farmers have signed up under different USDA programs. So that becomes a very important part. So we drew that data in for our deliberations.
We found out after the October 10 release, just about a week ago in fact, that there were the data discrepancies there. We took a look at what those differences were and decided that we needed to go and do corrections to our acreage, planted acres. That then corrects - means there will be changes in acres for harvest and also production.
So we released that corrected report this morning. We actually made changes in planted acres - and therefore production - for corn, soybeans, canola, sunflowers, dried beans and sorghum. The two significant ones in terms of crops were corn and soybeans.
Corn-planted acres in this report are down about a million acres from what we estimated in the release on October 10. The production is down about the same amount, 1-point-4 percent. Soybeans, again very similar: planted acres down about 1-point-1 million acres; and production, that's about 1-point-4 percent decrease in acres and about a 1-point-5 percent decrease in production.
WILLIAMS: Dr. Gerald Bange is chairperson of the World Agricultural Outlook Board. Dr. Bange?
DR. GERALD BANGE: Good morning. We had just made some relatively small changes in our balance sheets this morning. Spinning off of the October 10 report we have revised only three tables: one for the U.S. soybean supply and demand, one for U.S. corn and one for U.S. sorghum. Basically what we did is account for the 1-point-1 million less acres of soybeans harvested, approximately one million acres less of corn harvested; and we made a slight adjustment associated with the increase in sorghum acres that NASS reported this morning.
At this point I will assume you have the numbers in front of you. I won't go through them right now but we are available to talk about any changes in the balance sheet that we made. Suffice it to say on corn and soybeans, we basically tightened up on the supply and made a fairly slight increase in our price forecast for the 2008 and '09 marketing year.
Okay. Thank you.
WILLIAMS: And reminding you that we also have from the Farm Service Agency, FSA, Dr. Tom Hofeller here to take your questions, too.
Operator, we are ready for the first caller, please.
OPERATOR: Our first question comes from Charles Abbott (ThomsonReuters).
REPORTER: Hi. Good morning. I appreciate you making yourselves available. I'm hoping to cover a couple of things here. First off, perhaps Ms. House can explain to us what brought the discrepancies to light. I mean, I understand there's at least one data users conference at which USDA was questioned about its acreage figures. Anyway, what brought this discrepancy to light?
What would CFTC and SEC be doing because you alerted them to this problem?
And thirdly, maybe somebody can describe a little bit of the security precautions in a Crop Report. And I understand - Perhaps as a footnote, I'm curious: is the Crop Report in the federal world regarded as a major economic indicator?
So there you go - simple questions.
DEPUTY ADMINISTRATOR HOUSE: What brought the discrepancy to light and sort of how did that relate to the timing of when we were holding a data user meeting is Chicago, which was the 21st of October, and the 20th and 21st. The 20th was actually our data user meeting. At that point at that data user meeting I can tell you that I did not know that there was a discrepancy in the data. And so we were not talking about problems. I did not know that there was any problem at all at that point.
We did find out the day after that there was in fact this discrepancy. It was discovered because FSA analysts and some of the NASS analysts began to talk, in particular an FSA analyst asked the NASS analyst, "Why was our relationship, historical relationship with the farmer-reported acreage data, so different than it had been in the past?"
And we said, "It's not. We were using it the same way." But the FSA analyst was pulling data from a different FSA data source than what we traditionally report from. So they began to see that there were some differences between what FSA refers to as their "mainframe files" and the data mart which is a queriable that NASS always queries for our information.
We began to suspect that there were differences. FSA actually went to their computer specialists to get some definitive look at whether there were differences between those. Once we realized there were differences, we set in process the process of updating these acres, as quickly as we could.
-- Other parts of the question: Yes, the acreage and production numbers are a principal economic indicator of the United States. And the Commodity Futures Trading Commission and SEC (Securities and Exchange Commission) - the USDA notified them because it was appropriate to do that.
OPERATOR: Our next question comes from Alan Bjerga (Bloomberg).
REPORTER: Yes. In your numbers you also have changes to imports, exports. There's a slight change in the corn yields. How was it determined that those numbers are going to be changed in the context of what you were doing with acreage?
DR. BANGE: Well, the change in the corn yield reflects where the states are. I mean this is basically the way it's averaged by state, to arrive at a corn yield, and it showed a slight difference of one-tenth of a bushel per what NASS informed us this morning.
As far as the changes in the trade data, for instance we're looking at exports of corn to be a little bit lower. Well, we've got a little bit tighter supply. As a result of not having those one million acres harvested, we've lost about 167 million bushels, a portion of the supply. And of course we would have a little bit less to export.
Also, tightened up on the ending stocks forecast a little bit. It took that down 66 million bushels for corn. All of this resulted in a slightly higher price for corn, which one would expect to be a little bit lower exports as a result of that.
On the soybean side, a similar thing: We had 1.1 million acres less. We calculated we had 45 million bushels less production. And looking at exports it would be a little bit less than we had previously estimated. Because of that, we took that number down 30 million bushels as well as taking ending stocks down 15 (million). We also increased our price forecast just slightly 10 cents to 10.45, somewhere between 9.70 and 11.20. So we went up 10 cents on both ends and 10 cents at the midpoint to the 10.45.
WILLIAMS: Thank you. And that was Dr. Gerald Bange of the World Agricultural Outlook for the chairperson. Operator?
OPERATOR: Our next question comes from Christopher Doering (ThomsonReuters).
REPORTER: Thank you. I was wondering if you could just - related to this, but how does the average Crop Report work in terms of time you spend preparing for it? When do the USDA folks come to USDA the night before to begin putting it together? And if you could just kind of give me some basics?
DEPUTY ADMINISTRATOR HOUSE: Most of our work for an average Crop Report will begin in the end of the month somewhere around the 23rd, 24th of the month, as we begin data collection. In particular for yield we are going out into corn, soybeans, cotton fields, and we're actually making measurements of the crops in randomly selected fields and randomly selected plots.
We're also, during that timeframe, a few days before the first of the month and a few days after the first of the month, are calling a random sample of farmers and asking them what their yield is.
All of that information is gathered first in our state offices where they have the first look at that data, they are editing that data, and they are trying to make sure that it in fact is consistent. They have several days to do that. They also do the first analysis of the data - yield, acreage and production for their particular state. The staff in those state offices are not allowed to talk to each other but only to look at their own state data during that time period.
We then bring in some of the data into Washington: the acreage data; the nonspeculative yield data - and by that what I mean is from states that are not the top producing states. And our headquarters staff look at that information for several days before the Crop Report is due.
Then the night before the Crop Report we lock up in a locked area. And Chris I know you have been in there as a press member a number of times. We lock up sometime at night at between 9:00 at night and 3:00 in the morning, and we hold the Agricultural Statistics Boards which look at the yield for the speculative states, the states that produce the most of that commodity. It's usually someplace between -- those states would have 85 to 90 percent of the total production for that commodity.
Once we're in a locked environment like that, we unencrypt information that our state offices on their analysis of that data - the speculative states again - that they have done. And that's the first time that all of the data for the Crop Report comes together. We work through the night on the yield and production numbers. At 3:00 in the morning we turn those over to the World Agricultural Statistics Board which is locked up with us at that time. And that's the first time that they have a look at that information.
We are then printing the report, preparing the tables, and we publish at 8:30 in the morning, at which point we release to everyone at the same time.
OPERATOR: Our next question comes from Stewart Doan (AgriPulse).
REPORTER: A question for anyone who wants to take a stab at it. I gather the data I heard earlier off the spring planted crops - I didn't hear, there were no operations made to cotton or rice. Should that tell us then that the data was insignificant to move the estimates for those crops? Or is it the data from a producer set did not include cotton and rice producers?
The second question would be: What procedures have FSA, NASS implemented to try to avert this from occurring in the future again?
DEPUTY ADMINISTRATOR HOUSE: I'll take the first part of the question. NASS updated the acres for cotton and rice in the September report. And we used the FSA data that was available to us in September. Those program crops have an earlier sign-up date with FSA. And so they are more stable and complete a month earlier than the later, the other field crops we were talking about.
So we looked back at the data that we had gotten from FSA for the September crop and compared it with the data they had in their data mart at this point. And we were convinced that we did not have the same anomalies when we actually produced the September report.
Normally the September updated acres for cotton, rice and peanuts are just carried forward into the October report. So we made no changes to those planted acres based on the problematic data.
WILLIAMS: And this will be Dr. Tom Hofeller of Farm Service Agency, associate administrator for Operations and Management.
DR. TOM HOFELLER: Good morning. We're taking essentially a two-pronged approach. One, we've frozen the databases that were in question, and we'll be looking through the software and the procedures which transfer the data from our mainframe - which is the authoritative source - to the data mart; and trying to discover what went wrong.
But more importantly, we're also having meetings between our database managers and database experts from the department and from outside sources to look at this whole process and make sure we true it up and set in proper control procedures to ensure that the data mart is in sync with the mainframe database.
WILLIAMS: Thank you. Operator?
OPERATOR: At this time there are no further questions.
WILLIAMS: Are there any follow-ups? Any reporters with follow-up questions?
OPERATOR: We have an additional question from Stewart Doan (AgriPulse).
REPORTER: And thank you. I wanted to follow up. Has anyone gone back -- Is there a reason to believe that this could have happened in prior reports in prior years, or at this point are you confident that this was a one-time glitch?
DR. HOFELLER: We're fairly confident that this is something that has come up lately. We did go back and look at the '07 data and the figures that came from both the mainframe computer and the data mart. And they were as we would have expected with those two databases.
MS. HOUSE: I would just reemphasize that we did go back in September and looked at cotton, rice and peanuts. And we did not see a problem.
WILLIAMS: And Operator, anyone else with a question?
REPORTER: - (Chuck Abbott - ThomsonReuters) I was wondering what the Inspector General is going to do, because you've notified him as well as SEC and CFTC? Thank you.
WILLIAMS: Sorry, but we didn't get the first part of that. But I believe the question was regarding informing the Office of Inspector General Phyllis Fong within USDA. I believe there was just a question about that she has been notified?
DEPUTY ADMINISTRATOR HOUSE: Again, I think that USDA was just using due diligence in making sure that the Inspector General, the CFTC and the SEC were all notified. I do not know of any action that they are planning to take or not planning to take.
WILLIAMS: And Operator, we'll ask for a final round if there is any other reporter with a question on the line.
OPERATOR: Our next question comes from Bill Tomson (Dow Jones).
REPORTER: Hello. Can you hear me?
WILLIAMS: Yes. We hear you. Go ahead.
REPORTER: Lucky you - No. Just a minute ago somebody asked about whether or not this could have happened earlier. It was not Carol House that answered. The person didn't identify himself so I don't know how to quote it.
DR. HOFELLER: Tom Hofeller.
REPORTER: Okay. Thank you.
DR. HOFELLER: Yes. You're welcome.
WILLIAMS: Operator, are there any other questions?
OPERATOR: There are no further questions at this time.
WILLIAMS: All right. Thank you very much. The Crop Production Report, the abbreviated WASDE report, a news release from USDA are on-line already. A transcript of this call will be available on USDA.GOV.
Also, reporters with other follow-up questions may call Ellen Dougherty at 202-690-8122 or me, Keith Williams, at 202-720-4623.
We appreciate your time and interest. Thank you, and good morning.
