(specially concurring):
I concur with the result and the analysis in Judge Goldberg’s opinion with this comment: as I understand the holding in United States v. Winthrop, 417 F.2d 905 (5th Cir. 1969), and Judge Goldberg’s opinion here, the Court does not deviate from strict application of the statute to determine whether property is excluded from capital asset — capital gains treatment because it is “property held by the taxpayer primarily for sale to customers in the ordinary course of his trade or business.” 26 U.S.C.A. § 1221(1). As I understand the various factors set forth in Withrop and discussed in Judge Goldberg’s opinion here, the statutory test is still paramount and the factors serve only to aid in the determination of whether the statutory characteristics of non-capital assets are present. Under § 1221(1) it must first be determined that the property is held “primarily for sale;” and if so, that the sales are made in the ordinary course of taxpayer’s trade or business. This statutory standard is strictly followed in both Winthrop and in this case. The various indicia which courts have delineated are useful in applying the § 1221(1) test to the facts of a case, but should not be regarded as a substitute for the statutory test.
GEE, Circuit Judge, with whom BELL, COLEMAN, AINSWORTH and DYER, Circuit Judges, join, dissenting.Viewing as incorrect the en banc majority’s restatement of facts and law, I must respectfully dissent. I would ad*425here to the panel opinion, reported at 509 F.2d 171, which attempted to apply existing, controlling precedent in our circuit to the facts of this very close case as they were found by the district court. To obtain a different result, the majority has found it necessary to revise the law and refind the facts in important respects, as though the obtaining of capital gains treatment by this taxpayer in the three years in question were a catastrophe to be avoided at all costs.1
First, in setting out the facts of this case, the majority summarily discounts a critical trial court factfinding without ruling it clearly erroneous.2 The majority rejects the district court’s finding that taxpayer “is still farming a large part of the land,” 356 F.Supp. at 1336,3 refinding the facts as being that taxpayer’s tenant farmer “may” presently be farming the land, supra at 411, and that neither plaintiff nor the lower court claimed any dual purpose, id. at 423 n. 43. But the record affirmatively shows that the land has been and continues to be farmed,4 and the lower court specifically found multi-purpose use of the land.5
Second, although insisting at one point that it only “resummarizes” the relevant case law, id. at 415 n. 23, the majority revises the old test by placing pre-eminent emphasis on sales activity and improvements, effectively eliminating the other factors enunciated in United States v. Winthrop, 417 F.2d 905 (5th Cir. 1969).6 It is true that, to justify its emphasis on sales, the majority claims to detect in our earlier cases an overriding tendency toward finding ordinary income as the frequency and substantiality of sales increase. But that the strength of one factor, if it is a factor, can affect the outcome is self-evident, and as the panel opinion recognized in discussing a different one: “[I]f this factor is to be included in the seven other than as a sham there must be instances imaginable in which it will be critical, perhaps dis-positive.” 509 F.2d at 175. To bolster its reasoning, the majority points to Thompson v. Commissioner of Internal Revenue, 322 F.2d 122 (5th Cir. 1963), which held against the taxpayer, and asserts that the other sales indicia favored the taxpayer more in that case than here. Not so: in Thompson the taxpay*426er earned 48.8% of his income from the real estate sales in question, as compared to 11.1% here. More significantly, once Thompson’s sales began, “there is no evidence that he thereafter held the lots for any purpose other than the sale to prospective purchasers.” Id. at 128 (emphasis in original). Here, however, taxpayer still farms the land.
To explain its emphasis on improvements, the majority stresses the similarity between taxpayer’s improvements and those present in Winthrop. Supra at 417. But this comparison, too, is mistaken. The Winthrop court emphasized that property does not cease to be a capital asset merely because its increase in value was due in part to the taxpayer’s efforts in making improvements, 417 F.2d at 907—09, quoting the same portion of Barrios’ Estate v. Commissioner of Internal Revenue, 265 F.2d 517 (5th Cir. 1959), that the panel in this case cited:
The idea of selling a large tract of land in lots embraces necessarily the construction of streets for access to them, the provision of drainage and the furnishing of access to such a necessity as water. It is hardly conceivable that taxpayer could have sold a lot without doing these things.
Id. at 520, quoted with approval in Winthrop, 417 F.2d at 909, and Biedenharn, 509 F.2d at 174 (panel opinion). Furthermore, the Winthrop situation was shot through with facts not here present supporting the conclusion that the land was “property held by the taxpayer primarily for sale to customers in the ordinary course of his trade or business”: (1) taxpayer used the land for nothing but subdividing and selling, and “[t]here were, therefore, no multiple, dual, or changes of purpose during the relevant years . . . . The taxpayer, long before the tax years in question, had as his sole motivation the sale of [the property],” 417 F.2d at 911 (emphasis added)— i.e., the property was held “primarily for sale”; (2) as in Thompson, the taxpayer’s primary activity was substantial and continuous real estate sales, from which she earned 52.4% of her income — i.e., this activity was a “business”7; and (3) the taxpayer began selling shortly after he acquired the land, never using it for any other purpose8 — i.e., the sales were “ordinary in the course of this business.”
Thus, to the extent that I agree with the majority that “[o]ne need not go beyond Winthrop in order to decide the present dispute,” supra at 423 n. 46, I think Winthrop, in thé form in which it stood at the time this taxpayer was presumably arranging its affairs in reliance on that opinion requires a different result in this case — that reached by the panel, 509 F.2d 171 (1975).
In the course of “resummarizing” the relevant test, the majority adds a wholly new element, volition or the lack of it, in cases such as this involving an initial investment purpose followed by substantial and continuous sales activity. But to permit a finding of continuing investment purpose in the face of major sales activity only (or as the majority says, .supra at 421, “most generally”) when that activity “results from unanticipated, externally induced factors which make impossible the continued pre-existing use of the realty,” id. at 421, makes impossible any voluntary disposition of investment property at capital gains rates, exalts one factor — sales activity — over all others, and changes the law as de*427clared in Thompson and Winthrop, if not their results. This drcuit’s longstanding indusion of factors other than sales activity in the capital gains/ordinary income calculus implicitly recognized that an investor could engage in a voluntary liquidation and begin selling property without a change of purpose. Holding that property is not part of a business only so long as it is sold in large blocs, but not if it is sold in small parcels, discriminates irrationally against an investor who decides on liquidation but cannot locate purchasers interested in large acquisitions.9
And so, under the guise of emphasiz-' ing certain factors, the majority appears to have completely changed the Winthrop test to eliminate all factors but one. And while I entirely agree with the majority’s redundant warning that “once an investment does not mean always an investment,” id. at 423, I would also suggest that once a sale does not mean always a business.
. The majority opinion correctly notes, supra at 413, n. 17, that the panel opinion errs insofar as it suggests that all sales were by commission brokers during the tax years in issue. A very substantial majority were, however; and this circumstance, though not devoid of significance, is not central to the reasoning of either the panel opinion or the majority opinion here, use of brokers being consistent with either a liquidation or a business.
. While I agree with my brethren, supra at 416 n. 25, that this court must review ultimate legal determinations de novo, we must remember that fact determinations, which admittedly affect our legal conclusions, should be overturned only after careful scrutiny under the strict “clearly erroneous” standard.
. See also id. at 1335 (“substantial parts of [taxpayer’s property] are still farmed”).
. See Transcript at 8, 50; Deposition of B. W. Biedenham at 37-38. In fact, many potential customers were rejected when they requested land that taxpayer intended to farm. Id. at 22.
. The district court found the following;
Sales of lots were not the sole object of taxpayer’s business, nor did they approximate even a substantial part of taxpayer’s activities. There was no singleness of purpose here . . Here the property has been used for other purposes and still is being used for such.
356 F.Supp. at 1336 (emphasis added).
. Indeed, at another point the majority appears to say that Winthrop will no longer be followed at all, except in cases of “forced change of purpose.” Supra at 422. If so, we are not told what the new test is to be, unless it is — as the opinion comes perilously close to saying — that unless there is forced change of purpose there can be no capital gain. But the quoted language may mean no more than that the majority proposes to emphasize one Winthrop factor in one sort of case and not in another. If so, one can only pity the poor taxpayer, confronted now not only by Winthrop but by the court’s announcement that dispositive nuances of weight attach to some factors in some cases. If this were the law before today, there had been few hints of it; and with this announcement we all but lay down as law that ordinary income is whatever we say it is.
. The majority attempts to discount the significance of the low portion of taxpayer’s income represented by real estate sales (11.1%) by pointing out that the profit figure is high in absolute terms and low as a percentage only because taxpayer earns high profits in other businesses. Supra at 419. I fail to discern the significance of either explanation, especially in light of the importance placed on a high relative percentage in both Thompson and Winthrop, unless the opinion means to say that taxpayers with high incomes are less likely to have capital gains than those with low incomes.
. As the Winthrop court recognized, 417 F.2d at 912, this singleness of purpose, rather than any such “forced change of purpose” as is introduced by the majority here, supra at 418, distinguishes Winthrop from Barrios’ Estate.
. As the majority admits, supra at 412 n. 10, the record does not clearly indicate the number of sales, as opposed to the number of lots sold, and in some years we know that single sales accounted for the large numbers of lots disposed of in those years, see id. at 412 nn. 11 & 12.