J. & E. Stevens: Following the Money Behind America's Cast-Iron Toy Industry

An Industrial Enterprise Behind the Toys

The mechanical banks produced by J. & E. Stevens are among the most recognizable American cast-iron toys of the nineteenth century. Collectors know the Tammany bank, William Tell, Creedmoor, the Kicking Mule, and dozens of other ingenious creations in which coins activate miniature mechanical performances. Their invention, design, and decoration have received considerable attention. The economics of their manufacture have received rather less.

That omission is unfortunate because an unusually rich collection of company records survives. Transcriptions of original ledgers provide evidence of the firm's financial operations from the 1880s into the late 1890s, while an invoice copybook covering 1895–1896 records the quantities and prices of individual products. These sources allow us to examine J. & E. Stevens not simply as a maker of collectible toys, but as a business that employed workers, purchased materials, developed intellectual property, distributed merchandise, and rewarded its investors.

The central question is whether Stevens was a marginal manufacturer surviving on pennies or a genuinely profitable concern. The distinction matters. Its elaborate mechanical banks were labor-intensive products, but the company also sold inexpensive cap pistols, still banks, miniature safes, wheels, tools, and a remarkably broad assortment of cast-iron goods. The ledgers suggest that this diversity, combined with established production methods and nationwide distribution, produced a business capable of paying substantial dividends without continuously enlarging its factory.

The records do not establish every dollar of net profit. They do, however, permit a considerably more informed assessment than has generally been available to collectors.

Four Shareholders and a $140,000 Corporation

John and Elisha Stevens established their Connecticut ironworks in 1843. Initially, the business manufactured household hardware, tools, and utilitarian castings alongside a limited assortment of toys. Russel Frisbie, an accomplished pattern maker, designer, and businessman, joined the enterprise in 1866 after working for W. & B. Douglas.

A contemporary company history describes Frisbie's purchase of a one-quarter partnership interest for $3,000, paid equally to three existing partners. At the time, the business reportedly generated approximately $75,000 in annual sales. Three years later, J. & E. Stevens was incorporated with stated capital of $140,000, represented by 1,400 shares valued at $100 apiece. [1]

The original share allocation was straightforward.

Joseph Waters had died following a carriage accident in December 1868. His original quarter interest was divided among his widow, Sarah, who received 119 shares, and three children, Charles, Julia, and Mary, who each received 77 shares. The allocation is attributed to Corporation Record Book, Volume III, pages 18–22. [1]

This was neither a small family workshop nor a corporation assembled through a broad public stock subscription. Ownership remained concentrated among a few individuals and families. The company possessed valuable equipment, manufacturing patterns, inventories, and an established commercial operation before the mechanical bank business reached its maturity.

The shares' $100 par value should not be confused with their market value. Whether an outside buyer in 1869 would have paid $100 per share, more, or less cannot be established from the capitalization alone. What matters for the subsequent financial history is that the shares provided a clear framework for calculating dividends and, indirectly, reconstructing ownership.

What the Company Sold

The invoice copybook covering September 1895 through August 1896 presents a remarkable picture of Stevens's business. It includes 1,605 invoices and records more than 1.2 million cap pistols, approximately 17,000 mechanical banks, more than 72,000 toy safes, and more than 50,000 still banks. [2]

The figures are revealing. Cap pistols accounted for approximately 30 percent of recorded sales, mechanical banks 11 percent, toy safes 5 percent, still banks 4 percent, and other merchandise the remaining 50 percent.

A mechanical bank brought the manufacturer an average of approximately 54 cents, while a cap pistol averaged only two cents. Stevens had to sell about 27 average cap pistols to obtain the invoiced revenue of one average mechanical bank.

The quantities explain why inexpensive products could produce substantial revenue, but they do not establish which products were most profitable. A cap pistol required less material and finishing than an elaborate mechanical bank, although the margin on an item selling for a penny or two at the factory was necessarily measured in fractions of a cent.

The more consequential discovery is the importance of merchandise outside these two familiar categories. The invoice transcription identifies household hardware, miniature furniture, wheeled toys, trains, tools, wheels, and other goods. Mechanical banks may have been among Stevens's most distinctive and prestigious products, but the company was supported by a far wider manufacturing operation.

This diversity may have been an important source of financial resilience. A foundry equipped to produce a mechanical bank could use many of the same employees, furnaces, patterns, and distribution channels to manufacture less elaborate articles.

The Cost of Making a Mechanical Bank

The foundry payroll provides valuable evidence of the direct labor required to manufacture individual bank components.

Workers were paid for the production of acceptable castings, and the surviving piecework figures have been used to calculate approximate casting labor costs for several models.

These figures did not represent the complete manufacturing cost. Iron, fuel, molding materials, casting losses, cleaning, polishing, mechanical assembly, painting, packaging, and factory overhead still had to be considered. [2]

The variation is important. The Kicking Mule could command a substantially higher invoice price than certain simpler banks while requiring comparatively little casting labor. The elaborate mechanism did not necessarily translate into correspondingly high foundry piecework costs.

Consider a mechanical bank invoiced at 54 cents. If casting labor consumed seven cents, other direct labor thirteen cents, and materials and consumables six cents, the factory would have incurred an illustrative direct manufacturing cost of 26 cents. That would leave 28 cents, or approximately 52 percent of the invoice price, before allocated manufacturing overhead, selling expenses, royalties, and administration.

This is a hypothetical reconstruction rather than a measured Stevens margin. Nevertheless, it illustrates why a manufacturer selling relatively inexpensive toys could remain profitable even when the work involved several specialized operations.

The individual models also had markedly different economics. The Owl, for example, reportedly sold for approximately 33 cents while requiring nine cents of casting labor. Casting alone could therefore consume more than a quarter of the invoice price. The Kicking Mule, invoiced at approximately 60 cents, incurred only about five cents of comparable labor.

To understand the firm's overall profitability, it is necessary to move from the cost of individual articles to the general ledger.

Four Years of Declining Sales

The March 1891–February 1895 general ledger contains annual figures under an account labeled Manufacturers, along with labor and other expenditure accounts. [3]

Sales declined by approximately 15.8 percent between the first and final reported years. This was not an isolated period in American economic history. The Panic of 1893 produced a severe national depression, with bank failures, business contractions, unemployment, and reduced consumer purchasing power.

The Stevens figures show that the company experienced a decline during this unsettled period, although the downturn appears to have begun before the panic.

Labor expenditures also declined, but not in direct proportion to sales.

The ledger's annual sales labels and fiscal expense periods do not explicitly establish identical start and end dates. Nonetheless, pairing them in their reported sequence indicates labor expenditures equivalent to approximately 29–32 percent of sales.

That is a substantial labor commitment, but not necessarily an excessive one for an enterprise combining skilled foundry work, assembly, finishing, and hand decoration.

The records also identify annual amounts in an Expense account of roughly $17,000–$19,000 during these years. These amounts apparently relate to the profit-and-loss accounts, but their precise contents require verification before they can be treated as a modern equivalent of factory overhead or operating expenses.

This distinction is critical. The surviving ledger transcription is an index of accounts and transactions, not a reconstructed income statement. Its Profit & Loss balances cannot simply be interpreted as annual profits.

What Can Be Inferred About Margins?

Gross margin and net margin are different measures. Gross margin ordinarily reflects sales less the cost of manufacturing the goods sold, while net margin reflects the earnings remaining after all applicable expenses.

The Stevens records offer substantial information about the first calculation, but they do not fully establish it.

For the 1891 sales figure of $147,306, the adjacent annual labor expenditure of $45,623 represented approximately 31 percent. If we provisionally apply an annual materials expenditure of approximately $17,100, those two cost categories together would account for about 43 percent of sales, leaving a residual of roughly 57 percent.

That residual is not a verified gross margin. The labor and sales periods need reconciliation, the materials figure requires checking against the underlying classification, and other direct production costs may not be included. Inventory movements can also cause costs incurred during one year to differ from the cost of goods actually sold.

Still, the available figures challenge the assumption that Stevens necessarily operated on negligible manufacturing margins.

The company sold large quantities of inexpensive goods, but much of its advantage lay in its established production infrastructure. Once patterns and machinery had been developed, repeated castings could be produced without bearing the entire original design cost each time. The incremental cost of manufacturing additional pieces could be considerably less than the first cost of bringing a product into production.

The more defensible conclusion is that Stevens had a meaningful spread between recorded labor costs and manufacturing receipts. Whether its formal gross margin was 30 percent, 40 percent, or higher cannot yet be established. Neither can a reliable net-profit percentage.

Dividend payments, however, offer an independent measure of the cash the owners believed the business could distribute.

The Shareholders Collected $42,000 in Dividends

Between March 1891 and March 1894, Stevens recorded six dividend declarations totaling $42,000.

The dividends represented an aggregate $30 per share over approximately three years. Relative to the original $100 par value, shareholders received cash distributions totaling 30 percent of nominal invested capital. That is not a 30 percent investment return, since the market value and acquisition cost of the shares are unknown. [3]

The timing is particularly revealing. Stevens continued paying dividends during the economic contraction. The final $7,000 declaration came in March 1894, after the financial panic had already affected the national economy.

Across the four fiscal years covered by the ledger, the company recorded approximately $540,000 in Manufacturers sales. The $42,000 in dividends was equivalent to about 7.8 percent of that four-year sales total, although the declaration dates do not extend across precisely the same four-year interval.

Dividends are not identical to earnings. They may be distributed from retained profits, accumulated cash, or other available corporate resources. But repeated payments of this size, sustained across several years, suggest that Stevens was more than a marginal operation barely covering its factory costs.

The company was generating—or had accumulated—substantial distributable resources.

Russel Frisbie and the Transfer of Economic Power

The dividend records also reveal an important change in the company's ownership.

When the corporation was organized in 1869, Russel Frisbie held 350 shares, or 25 percent. By the early 1880s, his recorded dividend payments imply a holding of approximately 460 shares. During the 1891–1895 ledger period, the combined dividends paid to him amounted to $14,700.

Since the company declared $30 per share during that period, Frisbie's aggregate payments correspond to 490 shares.

By 1897–1898, later dividend records suggest his holding had increased to approximately 498 shares, equivalent to 35.6 percent of the original 1,400-share capitalization. [3][4][5]

These figures are inferred from dividend payments, rather than confirmed by a surviving stock-transfer register. They nevertheless demonstrate a significant growth in Frisbie's economic interest.

Frisbie had been instrumental in transforming Stevens from a diversified hardware foundry into a manufacturer celebrated for its mechanical toys. He was also a shareholder whose growing stake entitled him to an increasingly substantial portion of the company's distributions.

The 1891–1895 records identify another $11,040.35 paid to Frisbie under salary and expense accounts. Not all of that necessarily represented personal compensation; some may have been reimbursement for business expenses. Together, however, his dividends and management-related payments illustrate the close relationship between ownership and management.

Frisbie was not simply an employee administering the interests of outside investors. He was among the principal economic beneficiaries of the business.

The Stevens, Waters, Coe, and Hulbert families also continued to appear in the dividend accounts. Corporate ownership had dispersed among family members and heirs while Frisbie became the largest identifiable individual shareholder.

Exactly how he acquired the additional shares remains unresolved. The initial Waters inheritance included three blocks of 77 shares, making those holdings possible sources of later transactions. Subsequent dividend records, however, indicate that some Waters descendants held substantial interests long afterward. Probate inventories and stock-transfer records are needed before specific purchases can be attributed to Frisbie.

Was Stevens a Cash Cow?

The phrase is modern, but it describes a recognizable nineteenth-century business arrangement: an established enterprise with relatively predictable operations, limited apparent expansion requirements, and owners who regularly withdrew a meaningful portion of its accumulated resources.

Stevens exhibits several characteristics of such a business.

Its principal manufacturing processes were well established by the 1880s. The company already possessed the foundry, machinery, patterns, commercial relationships, and trained labor necessary to produce its goods. Although products changed continuously, the underlying manufacturing system did not require a completely new factory each time a mechanical bank was introduced.

The dividend records establish regular distributions. The ownership records show substantial payments to individuals who also played important roles in management. And the balance-sheet inventories offer a revealing indication of the firm's capital priorities.

The Tools & Machinery account was valued at $8,176.75 in 1892, $8,000 in 1893, and $8,500 in both 1894 and 1895. It subsequently rose to $9,500 and remained at that recorded amount through February 1898. [3][4]

This was not the accounting pattern one might expect from a company undertaking continuous, expensive enlargement of its plant.

But the evidence does not justify claiming that Stevens neglected reinvestment. The same records show that the value assigned to patents increased from $3,000 in 1892 to $11,500 in 1894, remaining there in 1895. Patterns increased from $16,000 to $17,500 over the same interval.

These are asset valuations, not proof of equivalent cash expenditures. Patent values might reflect purchases, capitalization, accounting adjustments, or other transactions. Nevertheless, they demonstrate that intellectual property and manufacturing patterns remained important corporate assets.

A more precise interpretation is that Stevens appears to have maintained a relatively stable physical manufacturing base while continuing to assign significant value to its products, designs, and patent portfolio.

Its investment strategy may have favored invention and product variation over wholesale modernization of the factory.

In that sense, the cash-cow characterization is plausible, but qualified. The ledgers establish substantial shareholder distributions and limited growth in the recorded machinery valuation. They do not yet establish the total sums invested in equipment, buildings, repairs, new designs, or other capital improvements.

Nor do they establish that dividends consumed most of the firm's annual profits.

A Business That Weathered the Panic

Perhaps the strongest evidence of Stevens's operating resilience is its response to the difficult economic conditions of the 1890s.

Annual sales declined, but the company continued operating, retained its established customers, paid workers, compensated inventors, and distributed dividends. The inventory and machinery accounts indicate a continuing business with substantial productive assets.

The later 1895–1898 ledger records labor expenditures falling from approximately $34,850 in fiscal 1895–1896 to $27,128 in 1897–1898. It also shows declining manufactured-goods inventory valuations, from approximately $40,655 to $27,814. These changes may reflect reduced production, an attempt to conserve working capital, or an adjustment to weaker market conditions. [4]

They suggest that Stevens was capable of contracting its operations when circumstances required.

Its resilience also rested on a broad customer network. Company accounts include George Borgfeldt & Co., Horace Partridge & Co., Butler Brothers, and numerous other wholesalers and retailers. The surviving invoices confirm shipments to F. W. Woolworth stores and business relationships with other toy manufacturers, including Weeden.

Thus the company was not dependent upon the mechanical bank collector's favorite products alone. It served a national wholesale market with merchandise spanning several price levels.

The firm's mechanical banks represented an extraordinary combination of novelty, mechanical design, and decorative appeal. Its simpler products helped make full use of its manufacturing capability.

There is no reason to assume that the bank selling for 54 cents was more profitable than the cap pistol selling for two cents—or the reverse. Each required a different combination of labor, materials, capital equipment, and volume. What the ledgers show is a business that successfully manufactured and distributed both, together with numerous other goods.

The Owners and the Workers

The financial records also permit a comparison between shareholder income and ordinary factory wages.

The 1891 payroll research reports average monthly foundry earnings of approximately $36.48. Workers employed throughout the year frequently received total compensation of several hundred dollars. Skilled personnel earned more, and some pieceworkers may have been able to earn as much as $3 per day, a respectable wage for the period. [2]

By comparison, Frisbie's $14,700 in dividends during the six declarations of 1891–1894 represented substantial wealth. His earlier recorded annual dividend payments sometimes exceeded $5,000, and in one year reached $8,280.

These payments were returns on ownership, not compensation for equivalent hours worked. Frisbie also had decades of experience, managerial responsibility, and capital at risk. Nevertheless, the difference illustrates how the economic rewards of industrial production were distributed.

Foundry workers were paid for their labor, often according to the amount of acceptable work they produced. Inventors received royalties when their designs were manufactured or sold under the applicable arrangements. Shareholders received dividends derived from the company's financial resources.

The continuing flow of dividends demonstrates the substantial economic value created by the operation. The distinction between the workers' wages and the owners' returns provides a useful perspective on nineteenth-century American manufacturing.

What the Records Still Cannot Tell Us

There are limits to the conclusions that can be drawn from these transcriptions.

The exact gross and net margins of J. & E. Stevens remain unestablished. The surviving summaries identify major costs, but do not yet provide a fully reconciled income statement. The treatment of inventory, depreciation, capital expenditures, administrative expenses, bad debts, and other accounts requires further examination.

Nor is the precise ownership history complete. Dividend payments provide strong evidence of share counts, but do not identify every purchase, sale, inheritance, or transfer.

The records also leave unresolved whether the directors consciously preferred shareholder distributions over expansion. That would require corporate minutes, dividend declarations, and evidence of decisions concerning major investments.

Several important primary sources remain available for independent investigation. Connecticut corporation filings may establish the original capital and later amendments. Probate records for John Stevens and Russel Frisbie could identify their stockholdings and the values placed upon them at death. Corporate ledgers in university archives may supply additional evidence of assets and expenditures.

These are not minor details. They could transform an already substantial financial history into a more complete account of corporate governance, family wealth, and industrial investment.

Conclusion: The Business Behind the Mechanical Banks

The surviving financial history of J. & E. Stevens suggests a successful manufacturer rather than an enterprise perpetually struggling to make pennies add up.

The company generated substantial annual sales, maintained a nationwide distribution system, operated with significant but manageable labor costs, and paid generous dividends to its shareholders. Its owners benefited materially from the company's success, with Russel Frisbie emerging as the dominant individual investor while continuing to participate in management.

The dividend history and comparatively stable machinery valuations support the interpretation of Stevens as a mature, income-producing business. Yet the increasing value assigned to patents and patterns cautions against portraying it as technologically stagnant or uninterested in reinvestment.

Perhaps the most important lesson is that the commercial success of Stevens cannot be understood from its celebrated mechanical banks alone. The company made money through an extensive industrial operation in which sophisticated toys, inexpensive novelties, household products, and contract manufacturing shared the same productive resources.

The mechanical banks are what survived in collectors' cabinets. The ledgers reveal the business that made them possible.

For historians of American industry, they also preserve something more consequential: a record of who owned that business, what it cost to operate, and who received the wealth it generated.

References and Sources for Independent Investigation

The following sources include transcriptions of original records, historical studies, and archival collections. The collector-club transcriptions are particularly valuable but should not be confused with the handwritten originals; the transcribers themselves caution that some readings may be uncertain.

[1] Historical organization and ownership

Cromwell, Connecticut, 1650–1990, reproduced in the Still Bank Collectors Club of America's 2012 compilation, with citation to Corporation Record Book, Volume III, pages 18–22.

Contains the 1869 capitalization, original shareholders, division of the Waters interest, and early management history.

2012 J. & E. Stevens research compilation

[2] Original sales invoices and foundry payroll

Still Bank Collectors Club of America and Mechanical Bank Collectors of America, J. & E. Stevens Company, compiled for their joint convention, Timonium, Maryland, June 2012.

Includes transcribed 1891 foundry payroll records, manufacturing labor analysis, sales by product, invoice examples, and customer correspondence.

Read the research compilation

[3] Company accounting ledger, March 1891–February 1895

J. & E. Stevens Company, Accounting Ledger, transcribed by Bill and Wendy Jones, courtesy of Robert Watrous, hosted by Mechanical Bank Collectors of America.

Contains accounts for dividends, shareholders, sales, salaries, wages, expenses, patents, patterns, equipment, and customers.

1891–1895 accounting ledger

[4] Company accounting ledger, March 1895–February 1898

J. & E. Stevens Company, Accounting Ledger, transcribed and hosted by Mechanical Bank Collectors of America.

Contains later dividend payments, manufacturing accounts, machinery valuations, labor expenses, inventory values, and the customer network.

1895–1898 accounting ledger

[5] Company accounting ledger, 1887–1891

J. & E. Stevens Company, Accounting Ledger, Mechanical Bank Collectors of America.

Provides additional shareholder, customer, and financial records preceding the Panic of 1893.

1887–1891 accounting ledger

[6] Manufacturing processes and company history

Shirley S. DeVoe, The Iron Toys of the J. & E. Stevens Company of Cromwell, Conn., Connecticut Historical Society Bulletin, 1971.

An important historical study of foundry techniques, labor, finishing, product development, and factory organization.

Read the historical study

[7] Historical corporate records

Connecticut State Library, Secretary of the State records and Articles of Incorporation Index, 1837–1945.

The relevant research target is the 1869 incorporation record, particularly the historical citation Corporation Record Book, Volume III, pages 18–22.

Connecticut incorporation records

[8] Surviving institutional corporate collection

University of Connecticut, Archives & Special Collections, J. & E. Stevens Company Records, 1888–1898, Collection 2008.0024.

The collection's described financial records offer an independent institutional source for investigating the company's suppliers, customers, and expenditures.

UConn archival collection

[9] Probate records for principal shareholders

Connecticut State Library, Middletown Probate District estate files, particularly the 1881–1915 collection.

Priority targets include John Stevens (died 1892) and Russel Frisbie (died 1898). Inventories, estate settlements, and distributions may establish share ownership and stock valuations. Specific inventories have not yet been examined.

Connecticut probate estate files directory

[10] Bank patents and product chronology

Mechanical Bank Collectors of America, List of J. & E. Stevens Company Banks, chronological register of mechanical bank designs and patents.

Mechanical bank patent and product register

Research note: All monetary figures are historical nominal U.S. dollars unless otherwise identified. Cost and margin calculations described as illustrative should not be interpreted as audited company financial results. The surviving ledger transcriptions contain account classifications and balances whose precise accounting meaning is not always established.